PDA

View Full Version : 3-500Z Filament To Grid Shorts, The Other Cause



AF6LJ
01-22-2012, 09:04 PM
One of the members of our weekend group on seventy five meters owns an SB-220, it works well and is lightly modified as he put it. The first weekend in January he had been doing some cleaning had every peace of gear in it's place save for the SB-220. One evening while in QSO while the ampli9fier was sitting idle he gave the amplifier a gentle push back to it's normal resting place and at that moment. BANG!!
He shut down the amplifier and proceeded to troubleshoot it the next day.
Besides the fact the grid shut was blown off the board one of the tubes had developed a filament to grid short.

This brings up the question; how many of those tube failures are caused by moving or even bumping the amplifier while the filaments are hot?

never know but given the conversations that have been raised on the subject I had to pass this along as just one more data-point in the discussion.

I've always taken it for granted that you don't move an amplifier with hot directly heated cathode tubes in it and I have always rolled my eyes when I hear of people in the old days running those Heath compact KW amplifiers mobile.

VK4TUX
01-22-2012, 09:43 PM
Sue, What do you think happened first, the "bang"or the bent filament?

A 50 ohm 50w ohmite glitch resistor would have saved the tube. A forward biased 6A10 across R3 would have been beneficial along with the usual
meter diode protection etc etc.

If you think a hot filament touching the grid during amp movement caused this event, then please tell me what you think caused the big bang?

I imagine the grid choke and P-supp R are not too healthy either.


Adrian ... vk4tux

AF6LJ
01-22-2012, 09:59 PM
Sue, What do you think happened first, the "bang"or the bent filament?

A 50 ohm 50w ohmite glitch resistor would have saved the tube. A forward biased 6A10 across R3 would have been beneficial along with the usual
meter diode protection etc etc.

If you think a hot filament touching the grid during amp movement caused this event, then please tell me what you think caused the big bang?

I imagine the grid choke and P-supp R are not too healthy either.


Adrian ... vk4tux

First came the bent filament, the big bang was the grid resistor going up in smoke.
The amplifier had a Harbach replacement board in it and had the meter protection.

W8JI
01-22-2012, 11:03 PM
One of the members of our weekend group on seventy five meters owns an SB-220, it works well and is lightly modified as he put it. The first weekend in January he had been doing some cleaning had every peace of gear in it's place save for the SB-220. One evening while in QSO while the ampli9fier was sitting idle he gave the amplifier a gentle push back to it's normal resting place and at that moment. BANG!!
He shut down the amplifier and proceeded to troubleshoot it the next day.
Besides the fact the grid shut was blown off the board one of the tubes had developed a filament to grid short.

This brings up the question; how many of those tube failures are caused by moving or even bumping the amplifier while the filaments are hot?


I would say none, or nearly none.

I regularly move hot-filament amplifiers, pick them up and inch or two and drop them, bang on tubes with a long insulated rod, and so on.

Unless the tube already has an intermittent short or bad weld, nothing bad happens. If it does short, the only thing that happens is the plate current increases to the level of zero bias plate current. Why would anything else happen?


never know but given the conversations that have been raised on the subject I had to pass this along as just one more data-point in the discussion.

I've always taken it for granted that you don't move an amplifier with hot directly heated cathode tubes in it and I have always rolled my eyes when I hear of people in the old days running those Heath compact KW amplifiers mobile.


As someone who regularly has hammered on tubes to locate intermittent shorts, it wouldn't bother me in the least to operate a tube mobile unless it was constant subjected to vibrations of significant magnitude. The filament doesn't get soft, weak, or mushy from being hot.

Ask him what he had in the amplifier for filter cap size and ESR, and what surge limiting resistor he had.

I see shorted grid to filaments dozens of times a year, and they never cause a bang. On the other hand an anode to grid discharge, without a suitable fault current limiting resistance, is known to cause damaged grids.

And naturally this would come down to some thinking an oscillation can cause such an arc. Never mind common sense, that even if we drove the tube with 1000 volts positive on the grid, current could not exceed the emission limitation of the filament unless the tube had gas. :-)

73 Tom

AG6K
01-22-2012, 11:07 PM
One of the members of our weekend group on seventy five meters owns an SB-220, it works well and is lightly modified as he put it. The first weekend in January he had been doing some cleaning had every peace of gear in it's place save for the SB-220. One evening while in QSO while the ampli9fier was sitting idle he gave the amplifier a gentle push back to it's normal resting place and at that moment. BANG!!
He shut down the amplifier and proceeded to troubleshoot it the next day.
Besides the fact the grid shut was blown off the board one of the tubes had developed a filament to grid short.

 RE: the SB220 :
1. Was there a glitch-R in the +HV lead - and if so how many Ωs and Joules?
2. was the shorted tube V2 or V1?
3. was either grid RFC damaged?
4. what was the measured resistance of the 47Ω R-supps?
Would the owner like to have the shorted tube autopsied, photographed, and the JPGs posted? If so, I will pay the postage.



This brings up the question; how many of those tube failures are caused by moving or even bumping the amplifier while the filaments are hot?

 When Karina A. and I were in the filament straightening business we found that straightening a bent 3-500 filament required 5.85v (1910บK) on the filament and 30 to 40-seconds at 11G, so it seems that the force that bends 3-500 filaments is nothing trivial.



never know but given the conversations that have been raised on the subject I had to pass this along as just one more data-point in the discussion.

I've always taken it for granted that you don't move an amplifier with hot directly heated cathode tubes in it and I have always rolled my eyes when I hear of people in the old days running those Heath compact KW amplifiers mobile.

 Catwoman: As long as it's not an aerobatic stunt plane mobile there should be no problem since on a bench test I determined that one month operating a 3-500Z horizontal made little change in grid-filament BDV.
• Here's a puzzler: TL-922s have a well deserved reputation for shorting the inboard tube while simultaneously blowing R-supp, collapsing or otherwise damaging the inboard tube's 1A grid RFC, and making a big band - all of which happens when the ZSAC drops to Zero.

• Rich, ag6k


"Almost nothing is as simple as it first appeard/" -- Mr. Murphy

VK4TUX
01-22-2012, 11:57 PM
Sue, What would the voltage be across the ~30 ohm grid resistor you think to make it cause the bang heard "going up in smoke". ~2.25 - 5.1v would be the voltage available from the cathode in contact.The amp is near zero biased so plate current increases, not sure if this would effect the IR = E across the grid resistor though? The filament is extremely difficult to bend at emission temperature.

I have an open 3-500zg that had the glass broken in transit, and I did a check a month back tapping the base hard with cathode lit 5v @ 14amp . The grid was connected to a fluke 87 with record min/max set, there was no voltage recorded between grid and FT CT connection during the test which involved wacking the base mounted in a socket with a small mallet. I also had a Sampo 60mhz scope connected visually looking for spikes with display retain settings. Nothing seen.

I seems something else happened to your friend besides his moving the amp, which may have initiated the event, but by causing a filament - grid short, I think not..


Adrian ... vk4tux

AF6LJ
01-23-2012, 12:16 AM
I would say none, or nearly none.

I regularly move hot-filament amplifiers, pick them up and inch or two and drop them, bang on tubes with a long insulated rod, and so on.

Unless the tube already has an intermittent short or bad weld, nothing bad happens. If it does short, the only thing that happens is the plate current increases to the level of zero bias plate current. Why would anything else happen?



As someone who regularly has hammered on tubes to locate intermittent shorts, it wouldn't bother me in the least to operate a tube mobile unless it was constant subjected to vibrations of significant magnitude. The filament doesn't get soft, weak, or mushy from being hot.

Ask him what he had in the amplifier for filter cap size and ESR, and what surge limiting resistor he had.

I see shorted grid to filaments dozens of times a year, and they never cause a bang. On the other hand an anode to grid discharge, without a suitable fault current limiting resistance, is known to cause damaged grids.

And naturally this would come down to some thinking an oscillation can cause such an arc. Never mind common sense, that even if we drove the tube with 1000 volts positive on the grid, current could not exceed the emission limitation of the filament unless the tube had gas. :-)

73 Tom

I can answer some of this....

The SB-220 in question was stock for the Harbach replacement HV diode board.
The tubes were original tubes from the early seventies.
As I remember that board has no provisions for a glitch resistor.

Next time I talk I will ask him about the filter caps....
The tubes in question had a lot of hours on them so it wouldn't but out of the relm of possibility for mechanical weakness to be the cause.

The ham in question has an electronics background...

So we shall see I'll ask a few more questions and get back to you.

For the record;
I've seen RCA conduction UHF power amplifier tuves shaken apart in high power (MST series) Motrack radios. and those tubes are metal ceramic with indirectly heated cathodes.

AF6LJ
01-23-2012, 12:19 AM
 RE: the SB220 :
1. Was there a glitch-R in the +HV lead - and if so how many Ωs and Joules?
2. was the shorted tube V2 or V1?
3. was either grid RFC damaged?
4. what was the measured resistance of the 47Ω R-supps?
Would the owner like to have the shorted tube autopsied, photographed, and the JPGs posted? If so, I will pay the postage.



 When Karina A. and I were in the filament straightening business we found that straightening a bent 3-500 filament required 5.85v (1910บK) on the filament and 30 to 40-seconds at 11G, so it seems that the force that bends 3-500 filaments is nothing trivial.



 Catwoman: As long as it's not an aerobatic stunt plane mobile there should be no problem since on a bench test I determined that one month operating a 3-500Z horizontal made little change in grid-filament BDV.
• Here's a puzzler: TL-922s have a well deserved reputation for shorting the inboard tube while simultaneously blowing R-supp, collapsing or otherwise damaging the inboard tube's 1A grid RFC, and making a big band - all of which happens when the ZSAC drops to Zero.

• Rich, ag6k


"Almost nothing is as simple as it first appeard/" -- Mr. Murphy

Catwoman :D
I love it. :)

I'll talk to him about the tube.

WG7X
01-23-2012, 12:58 AM
Well... Speaking only in my semi-official capacity as a routine critter catcher, I can state unequivocally that your friend obviously, "accidentally", unlatched the critter catcher located somewhere in the lower bowel region of his amp, there by loosing the dreaded "banger cat" which then exited the amp at supersonic speed, causing the small "sonic boom" which was heard.

The sonic boom in turn caused various wires and thingies to become discombobulated and making with the arcky-sparky and flashy-washy which then rendered the amp unfit for use.

There!

Simple, no?

73 Gary

PS: For the humor impaired::D

W8JI
01-23-2012, 01:30 PM
I thought it was because the movement made the amplifier cross the flux lines of the earth, generating a small current that triggered a parasitic.

This can be cured by positing the amplifier so the front panel width runs north-south, or by contra-winding the suppressors.

Or perhaps it was because the reflection of his diamond ring diverted a photon into the case, where it hit something in the tube and made it break into uncontrolled oscillation. This is why we should never wear rings when working on live circuits, especially in solar storms.

N4BBQ
01-23-2012, 03:15 PM
I thought it was because the movement made the amplifier cross the flux lines of the earth, generating a small current that triggered a parasitic.

This can be cured by positing the amplifier so the front panel width runs north-south, or by contra-winding the suppressors.

Or perhaps it was because the reflection of his diamond ring diverted a photon into the case, where it hit something in the tube and made it break into uncontrolled oscillation. This is why we should never wear rings when working on live circuits, especially in solar storms.

That was darn funny there Tom.

WA6MHZ
01-23-2012, 03:30 PM
What is a good value/wattage for a Glitch resistor on the SB-220 and Drake L7? I probably should install one of those while rebuilding the Power Supply. It just goes in series with the HV lead, correct?

AF6LJ
01-23-2012, 03:31 PM
Regardless of the cause;
I don't believe the movement shock excited the amplifier into oscillation.

The only other possibility I can see would be a loose grid wire touching the anode.

From something I remember or think I do the grids in the amplifier are directly grounded.
I will know more Saturday morning.

If it was a stray grid wire the evidence will be inside the tube.

AF6LJ
01-23-2012, 03:33 PM
I thought it was because the movement made the amplifier cross the flux lines of the earth, generating a small current that triggered a parasitic.

This can be cured by positing the amplifier so the front panel width runs north-south, or by contra-winding the suppressors.

Or perhaps it was because the reflection of his diamond ring diverted a photon into the case, where it hit something in the tube and made it break into uncontrolled oscillation. This is why we should never wear rings when working on live circuits, especially in solar storms.

I always make a point of taking off my earrings while operating or working on electronic gear.

W8JI
01-23-2012, 03:38 PM
What is a good value/wattage for a Glitch resistor on the SB-220 and Drake L7? I probably should install one of those while rebuilding the Power Supply. It just goes in series with the HV lead, correct?

A good resistor value, that will make a significant difference, is 20 ohms.

It has to be capable of handling shorts without arcing internally or exploding. The 175 series P rated RCD resistors are good. Ameritron uses two ten ohm in series, and has virtually no resistor failures in tube arcs.

For larger capacitance values, and higher voltage, it takes a large carbon glo-bar type, like a Stackpole or Carborundum style a few inches long.

Upsizing the electrolytics in the SB220 could cause problems. With stock electrolytics, tube grid damage is not as likely as with higher capacitance low-ESR capacitors.

Some upgrades, like better capacitors, come with hidden downgrades, like increased fault currents.

73 Tom

AG6K
01-23-2012, 05:48 PM
A good resistor value, that will make a significant difference, is 20 ohms.

It has to be capable of handling shorts without arcing internally or exploding. The 175 series P rated RCD resistors are good. Ameritron uses two ten ohm in series, and has virtually no resistor failures in tube arcs.

 Tube arcs? Big-bangs are from arcs made in the atmosphere, not in a near-vacuum.



For larger capacitance values, and higher voltage, it takes a large carbon glo-bar type, like a Stackpole or Carborundum style a few inches long

 agreed - or one can put two Ohmite 120J surge resistors in series to handle more Joules. In a SB220, two Ohmite 15Ω, 120J surge resistors will safely discharge 400uF filter caps and limit peak I to only 100 Amperes. .



Upsizing the electrolytics in the SB220 could cause problems. With stock electrolytics, tube grid damage is not as likely as with higher capacitance low-ESR capacitors.

 it's not the grid that gets bent Tom, it's the filament helices:
84248
p.15 September,1990, QST Magazine



Some upgrades, like better capacitors, come with hidden downgrades, like increased fault currents.

73 Tom

 Good point. However, without a suitable glitch-R, during its intermittent parasite near 110MHz, the original el-cheapo SB-220 filter caps bent many 3-500Z filaments, resulting in fil-grid shorts and T2 meltdowns.
• Rich, ag6k

AG6K
01-23-2012, 06:01 PM
What is a good value/wattage for a Glitch resistor on the SB-220 and Drake L7? I probably should install one of those while rebuilding the Power Supply. It just goes in series with the HV lead, correct?

 Correct For 3-500Zs, limiting peak discharge current to 200A seems to prevent filament bending during glitches that cause a large burst of grid-I. At 3000v this means that 15Ω will suffice. For a SB-220 with a stock C-filter = 25uF, an Ohmite 120J surge resistor will do the job. - note - If u need one, we buy these in bulk for our suppressor retrofit kits and we sell them separately for $1.77.
• Rich, ag6k

VK4TUX
01-23-2012, 11:27 PM
I have an open 3-500zg that had the glass broken in transit, and I did a check a month back tapping the base hard with cathode lit 5v @ 14amp . The grid was connected to a fluke 87 with record min/max set, there was no voltage recorded between grid and FT CT connection during the test which involved wacking the base mounted in a socket with a small mallet. I also had a Sampo 60mhz scope connected visually looking for spikes with display retain settings. Nothing seen.



Picture of RFP 3-500zg tested;

84279

The grid filament clearance is substantial on these.

I'm curious if the floor has carpet where this occurred Sue?

Adrian ... vk4tux

AF6LJ
01-23-2012, 11:30 PM
http://forums.qrz.com/attachment.php?attachmentid=84279&d=1327361059

That's interesting looks like the filament doesn't go all the way up to the top of the grid structure.

VK4TUX
01-23-2012, 11:37 PM
That's interesting looks like the filament doesn't go all the way up to the top of the grid structure.

That's right, and helps to explain the intensity of color (hot-orange) on the bottom third of 3-500zg RFP anodes seen when operating. The color slowly rises as the heavy usage continues to make the whole graphite anode bright. On low-medium use you just see some color at the base.


Adrian ... vk4tux

W8JI
01-24-2012, 01:25 AM
Sue,

The grid of a 3-500Z tube is about 1.5 inches long, and the filament goes to about 1/10th of an inch from the top. Some slightly less.

It would be silly to have a filament halfway up a grid in a power grid tube. The real reasons the lower part of the anode gets hotter first is heat conduction out of the stem keeps the top cooler, there is much larger thermal mass and area to radiate at the top because the top is closed instead of open-ended, and to a lesser extent because the grid and filament sit slightly down from the very anode top for gap clearance.

Here is what they actually look like:


84308

You can very clearly see the filament, as attached while the base is in place, from this side view.

84309


The entire thing about parasitics and filaments or grids is bizarre. The grid-filament clearance is not "substantial" in any 3-500, unless you are a microbe. It actually measures 22 / 1000th of an inch. I don't know where some of this stuff comes from, but high mu tubes have very close filament and grids.

With .022 inch spacing, and a ~1.8 inch length from the support, you can imagine the alignment issues, especially with heat.

73 Tom

AF6LJ
01-24-2012, 01:30 AM
Sue,

The grid of a 3-500Z tube is about 1.5 inches long, and the filament goes to about 1/10th of an inch from the top. Some slightly less.

It would be silly to have a filament halfway up a grid in a power grid tube. The real reasons the lower part of the anode gets hotter first is heat conduction out of the stem keeps the top cooler, there is much larger thermal mass and area to radiate at the top because the top is closed instead of open-ended, and to a lesser extent because the grid and filament sit slightly down from the very anode top for gap clearance.

Here is what they actually look like:


84308

You can very clearly see the filament, as attached while the base is in place, from this side view.

84309


The entire thing about parasitics and filaments or grids is bizarre.

73 Tom

Those pictures are much better the one the other poster had looked like the filament was only two thirds up the inside of the grid.
I was scratching my head.

W8JI
01-24-2012, 01:53 AM
I scratch my head a lot reading things from your homeland. :-)

I broke the grid top off where it folds over, so I could plug the grid-filament with a wire, and it just clears a .022 inch rod. That means the filament support rod in the middle has to hold alignment over the distance down to the base insulator of roughly about 1.8 inches long, and not let two things .022 inch apart touch.

I'd hardly call that "substantial spacing".'

A parasitic blew the glass clean off this tube. I confirmed that because my plasma cutter's head shows a dip on my dip meter at c.90 MHz. :rolleyes:

73 Tom

AF6LJ
01-24-2012, 02:00 AM
I scratch my head a lot reading things from your homeland. :-)

I broke the grid top off where it folds over, so I could plug the grid-filament with a wire, and it just clears a .022 inch rod. That means the filament support rod in the middle has to hold alignment over the distance down to the base insulator of roughly about 1.8 inches long, and not let two things .022 inch apart touch.

I'd hardly call that "substantial spacing".'

A parasitic blew the glass clean off this tube. I confirmed that because my plasma cutter's head shows a dip on my dip meter at c.90 MHz. :rolleyes:

73 Tom

That is not much clearance.
I could easily see a sudden jolt of a few Gs could cause a fatigued filament and support to short.

VK4TUX
01-24-2012, 02:57 AM
Those pictures are much better the one the other poster had looked like the filament was only two thirds up the inside of the grid.
I was scratching my head.

The fact it is a different type tube does not matter? Eimac 3-500z tantulum does not equal 3-500zg RFP RFP = RF Parts. The filament is as it looks; 5/8 the height of the grid cage, and the area with the largest emission concentration.

So you see Sue the pictures are different because the tubes are in fact different.

Adrian ... vk4tux

AF6LJ
01-24-2012, 03:07 AM
The fact it is a different type tube does not matter? Eimac 3-500z tantulum does not equal 3-500zg RFP RFP = RF Parts. The filament is as it looks; 5/8 the height of the grid cage, and the area with the largest emission concentration.

So you see Sue the pictures are different because the tubes are in fact different.

Adrian ... vk4tux
Actually it does matter you misrepresented the and implied it was a 3-500
We are done I don't deal with liars.
Welcome to my ignore list.

VK4TUX
01-24-2012, 03:17 AM
Actually it does matter you misrepresented the and implied it was a 3-500
We are done I don't deal with liars.
Welcome to my ignore list.

Saying "I have an open 3-500zg that had the glass broken in transit" quoted before and regarding the picture posted is not a lie as that's what it is.

The Eimac is not a zg it is a z, Even the 1967-1969 Eimac graphite 3-500z was called a 3-500z with no g involved.

So you're very quick to call me a liar, however I think you realise now what you dont wish to accept, which has you struggling with your temper. You can put me on your ignore list, because you dont like the truth rather than lies.

I thought many would like to see what a RFP 3-500zg looks like inside. Feel free to break one open and check it out if you think my picture is not genuine.


Adrian ... vk4tux

AF6LJ
01-24-2012, 03:28 AM
Saying "I have an open 3-500zg that had the glass broken in transit" quoted before and regarding the picture posted is not a lie as that's what it is.

The Eimac is not a zg it is a z, Even the 1967-1969 Eimac graphite 3-500z was called a 3-500z with no g involved.

So you're very quick to call me a liar, however I think you realise now what you dont wish to accept, which has you struggling with your temper. You can put me on your ignore list, because you dont like the truth rather than lies.

I thought many would like to see what a RFP 3-500zg looks like inside. Feel free to break one open and check it out if you think my picture is not genuine.


Adrian ... vk4tux
Okay.
..........

VK4TUX
01-24-2012, 03:30 AM
One of the members of our weekend group on seventy five meters owns an SB-220, it works well and is lightly modified as he put it. The first weekend in January he had been doing some cleaning had every peace of gear in it's place save for the SB-220. One evening while in QSO while the ampli9fier was sitting idle he gave the amplifier a gentle push back to it's normal resting place and at that moment. BANG!!
He shut down the amplifier and proceeded to troubleshoot it the next day.
Besides the fact the grid shut was blown off the board one of the tubes had developed a filament to grid short.

This brings up the question; how many of those tube failures are caused by moving or even bumping the amplifier while the filaments are hot?

never know but given the conversations that have been raised on the subject I had to pass this along as just one more data-point in the discussion.

I've always taken it for granted that you don't move an amplifier with hot directly heated cathode tubes in it and I have always rolled my eyes when I hear of people in the old days running those Heath compact KW amplifiers mobile.

A couple of things here. Do you know what tubes are inside this amp? Do you know if the tube sockets have grid resistors or chokes fitted?

Usually they have 1mH grid chokes, but you say resistors?

If grid resistors were fitted usually the correct ones which would have protected the tube from developing a grid - filament short would have protected the tube, but you say it has a bent filament with a grid-filament short?

You should really get your facts right before you come on here pretending to know something.

Adrian ... vk4tux ..

AF6LJ
01-24-2012, 03:47 AM
A couple of things here. Do you know what tubes are inside this amp? Do you know if the tube sockets have grid resistors or chokes fitted?

Usually they have 1mH grid chokes, but you say resistors?

If grid resistors were fitted usually the correct ones which would have protected the tube from developing a grid - filament short would have protected the tube, but you say it has a bent filament with a grid-filament short?

You should really get your facts right before you come on here pretending to know something.

Adrian ... vk4tux ..
I know what is inside of a stock SB -220
Fixed one for my EX the Freebander several years ago.
We had the Nichrome discussion (more like a shouting match).
He went off to pout...
I replaced the stock parasitic choke resistors and directly grounded the grids.
No more barking on forty meters.

So while I know what a stock 220 as I said a couple of posts down from that one I'll have more information after the weekend.

As far as I am concerned it's foolish to do anything bur directly ground the grids of those tubes.

VK4TUX
01-24-2012, 04:14 AM
Ok , well it's nice to see you have become less hostile with me here. I'd like to know what is in the grid circuit of your friends sb-220 that went "BANG"? The Topic post amp etc, and what type tubes are in there of which one failed > shorted.
Perhaps you will find that out later. My main sb-220 (of three) has directly grounded grids too, and I also like the resulting smoothness of tuning and stability. The extra gain is nulled by cathode resistors I fitted.


Adrian ... vk4tux

AF6LJ
01-24-2012, 04:22 AM
Ok , well it's nice to see you have become less hostile with me here. I'd like to know what is in the grid circuit of your friends sb-220 that went "BANG"? The Topic post amp etc, and what type tubes are in there of which one failed > shorted.
Perhaps you will find that out later. My main sb-220 (of three) has directly grounded grids too, and I also like the resulting smoothness of tuning and stability. The extra gain is nulled by cathode resistors I fitted.


Adrian ... vk4tux

That makes more sense to do.
Even the Eimac application note for that tube showed the grids directly grounded.
It has never made any sense to me why Heath Kenwood and others did what they did.

As soon as I know You, Tom, Rich, and Co. will know.

AG6K
01-24-2012, 10:08 AM
Those pictures are much better the one the other poster had looked like the filament was only two thirds up the inside of the grid.
I was scratching my head.

 The filament helices in that tube had apparently been shattered by the same EMF that bent them just before they broke. This is rare in 3-400s and 3-500s. In 811As it isn't. . . Apparently 4A Th-W filament wire is more fragile than than 14.7A. . Perhaps the presence of a ceramic spacer in that particular production run of 3-400s was a factor in the filament shattering?
• Rich, ag6k

AG6K
01-24-2012, 11:42 AM
Sue,

The grid of a 3-500Z tube is about 1.5 inches long, and the filament goes to about 1/10th of an inch from the top. Some slightly less.
It would be silly to have a filament halfway up a grid in a power grid tube.


 3-400Z s/n Q02888 has a bent and shattered filament. The broken off parts of each helix landed in the base. The bent parts are inside the grid cage



The real reasons the lower part of the anode gets hotter first is heat conduction out of the stem keeps the top cooler, there is much larger thermal mass and area to radiate at the top because the top is closed instead of open-ended, and to a lesser extent because the grid and filament sit slightly down from the very anode top for gap clearance.

Here is what they actually look like:


84308

You can very clearly see the filament, as attached while the base is in place, from this side view.

84309


The entire thing about parasitics and filaments or grids is bizarre.

 The first time I was in a welding shop and I saw 00-awg Cu cables jumping around on the floor from a 400A arc-welder I thought it was pretty bizarre Tom. . .
• So how much peak discharge-I can we expect from modern electrolytics? Example: Panasonic's long-life 105บบC, low ESR capacitors. The 450wv, 560uF rated units have an ESR of <0.236Ω. Ohm's Law tells us that the peak discharge-I should be 450v/0.236Ω = 1906A. If 400A can move 00-awg Cu cable around on the floor, what could 1900A likely due to a hot Th-W filament in a 3-500 during a glitch?
• It is my opinion that it is bananas to run a 3-400 or a 3-500 without a glitch-R in the HV+ lead that will limit peak filter-C discharge-I to 200A or less. As I see it, it would make economic sense for tube suppliers like RF Parts to state that 3-500s that are used in an amplifier that does not have a suitable glitch-R to limit potentially fatal filter-C peak discharge-I are not covered by warranty and that a tube that is returned will not be replaced gratis if it fails the filament-grid BDV test that it passed prior to originally being being shipped. . .

[QUOTE]
The grid-filament clearance is not "substantial" in any 3-500, unless you are a microbe

 I have tested 3-500Zs that measured 8000 BDV between the filament and the grid - although the average is a bit less. However, in practice a 3-500 that tests at >5000 BDV between the filament and grid is apparently ok to put into service.
• Rich, ag6k


.

It actually measures 22 / 1000th of an inch. I don't know where some of this stuff comes from, but high mu tubes have very close filament and grids.

With .022 inch spacing, and a ~1.8 inch length from the support, you can imagine the alignment issues, especially with heat.

73 Tom

W8JI
01-24-2012, 11:49 AM
The fact it is a different type tube does not matter? Eimac 3-500z tantulum does not equal 3-500zg RFP RFP = RF Parts. The filament is as it looks; 5/8 the height of the grid cage, and the area with the largest emission concentration.

So you see Sue the pictures are different because the tubes are in fact different.

Adrian ... vk4tux


The RF Parts 3-500ZG filament stops ~0.15 inches below the top grid end. The grid is ~1.6 inches long. The spacing is ~.022 inch.

5/8th of the way would be a 1 inch tall filament. The filament isn't even remotely close to that height.


Here is an RF Parts 3-500ZG tube grid. The attachment point is just above the very top wire retaining ring. Those horizontal rings are about 0.15 inches apart:

84331

Every 3-500 tube is about the same inside dimensionally.

It would be silly to extend a grid far beyond the height of the filament and emission area, and impossible to have "significant spacing" from grid to filament in a high-mu triode.

73 Tom

AG6K
01-24-2012, 12:06 PM
That makes more sense to do.
Even the Eimac application note for that tube showed the grids directly grounded.
It has never made any sense to me why Heath Kenwood and others did what they did.

As soon as I know You, Tom, Rich, and Co. will know.

 The idea to gnd the grids through capacitors came from Bill Orr, W6SAI - a guy who wrote to QST Magazine complaining that my articles had many technical errors - but he never got around to poinnting one out. QST Staff sent me a photocopy of Orr's letter if anyone would like to see it.
• Grounding grids through capacitors affects grid-resonance about 1% so it probably has little effect on anything. The only plus I can see is that grid caps very slightly widen the self-neutralizing freq. window of the 3-500 from roughly 0 - 89MHz to 0 to 90MHz by canceling part of the grid-L. However, to much improve VHF stabilization, VHF amplification needs to be reduced with a L/R suppressor in the anode.
• Rich, ag6k

AG6K
01-24-2012, 12:24 PM
Ok , well it's nice to see you have become less hostile with me here.

 Adrian == I take hostility as an indication that I'm likely headed in the right direction. I pretty much figure I'm on the right path when a moderator/censor exhibits wrath for pointing out a technical gaffe, or posting an "expert"-embarrassing photograph.
• Rich, ag6k




I'd like to know what is in the grid circuit of your friends sb-220 that went "BANG"? The Topic post amp etc, and what type tubes are in there of which one failed > shorted.
Perhaps you will find that out later. My main sb-220 (of three) has directly grounded grids too, and I also like the resulting smoothness of tuning and stability. The extra gain is nulled by cathode resistors I fitted.


Adrian ... vk4tux

W8JI
01-24-2012, 12:42 PM
 The idea to gnd the grids through capacitors came from Bill Orr, W6SAI - a guy who wrote to QST Magazine complaining that my articles had many technical errors - but he never got around to poinnting one out. QST Staff sent me a photocopy of Orr's letter if anyone would like to see it.

Dozens of people wrote to QST about your articles Rich, let's not single Orr out as a bad guy. Orr made mistakes, as we all do, but a few mistakes does not mean his input has no value.

The ARRL was more than fair. They worked hard to find anyone anywhere in the world with experience or credentials who agreed with you, but came up with nothing. The ARRL had a stack of papers a foot tall from dozens of people who responded, none of whom agreed with your articles. I would have thought if there was any technical merit, there would have been a stack at least equally high agreeing with you from reliable sources, yet there were none!

Rather than pick on Orr, why not point out some credible sources who agreed with you? Better yet, this was a quarter-century ago. Why not move on to something new? If in over 25 years no one with any experience or credentials has agreed with your articles, it is unlikely to change in the next 25 years.

Showing pictures of filaments relocated so they are half-way up in tube grids doesn't help your cause, and embarrasses the few who do support your claims by relying on those pictures as factual, complete, or accurate.

73 Tom

VK4TUX
01-24-2012, 12:55 PM
The RF Parts 3-500ZG filament stops ~0.15 inches below the top grid end. The grid is ~1.6 inches long. The spacing is ~.022 inch.

5/8th of the way would be a 1 inch tall filament. The filament isn't even remotely close to that height.


Here is an RF Parts 3-500ZG tube grid. The attachment point is just above the very top wire retaining ring. Those horizontal rings are about 0.15 inches apart:

84331

Every 3-500 tube is about the same inside dimensionally.

It would be silly to extend a grid far beyond the height of the filament and emission area, and impossible to have "significant spacing" from grid to filament in a high-mu triode.

73 Tom

Tom & Sue I rechecked the 3-500zg after seeing your picture Tom, and the filament has dropped down without me realizing that's not normal. So thank you for the education there and I apologize for unintentionally misleading anyone. It's irrelevant to Sue's original post, but good to learn what is normal structure within the tube. Thankyou.

8433384334

Adrian ... vk4tux

W8JI
01-24-2012, 01:06 PM
Adrian,

Everyone makes mistakes, so don't worry about it. A mistake once in a while doesn't mean anything beyond a mistake.

Ten or twenty mistakes in one article is another matter. ;)

73 Tom

VK4TUX
01-24-2012, 01:09 PM
Showing pictures of filaments relocated so they are half-way up in tube grids doesn't help your cause, and embarrasses the few who do support your claims by relying on those pictures as factual, complete, or accurate.


What does what I posted have to do with Rich? I posted a pic I took a while back without realizing the filament had dropped, and I found out.
So that's my deal and nothing to do with why filaments get bent, vhf parasitic s etc or Rich. Be fair!

Adrian ... vk4tux

AF6LJ
01-24-2012, 01:39 PM
Tom & Sue I rechecked the 3-500zg after seeing your picture Tom, and the filament has dropped down without me realizing that's not normal. So thank you for the education there and I apologize for unintentionally misleading anyone. It's irrelevant to Sue's original post, but good to learn what is normal structure within the tube. Thankyou.

8433384334

Adrian ... vk4tux
That's cool I apologize for getting bent out of shape over it.

AF6LJ
01-24-2012, 01:44 PM
 Adrian == I take hostility as an indication that I'm likely headed in the right direction. I pretty much figure I'm on the right path when a moderator/censor exhibits wrath for pointing out a technical gaffe, or posting an "expert"-embarrassing photograph.
• Rich, ag6k

Everybody makes mistakes; it's how we deal with them that sets eachother apart and determines the quality of a person.

I got angery over what I thought was a misrepresentation of a 3-500Z.
Adrian and I cleared up the situation and each has apologized.

AF6LJ
01-24-2012, 01:56 PM
 The idea to gnd the grids through capacitors came from Bill Orr, W6SAI - a guy who wrote to QST Magazine complaining that my articles had many technical errors - but he never got around to poinnting one out. QST Staff sent me a photocopy of Orr's letter if anyone would like to see it.
• Grounding grids through capacitors affects grid-resonance about 1% so it probably has little effect on anything. The only plus I can see is that grid caps very slightly widen the self-neutralizing freq. window of the 3-500 from roughly 0 - 89MHz to 0 to 90MHz by canceling part of the grid-L. However, to much improve VHF stabilization, VHF amplification needs to be reduced with a L/R suppressor in the anode.
• Rich, ag6k

It just seems to be contrary to common sense to do that...
Orr has produce3s some interesting stuff, not the least of which was his super cathode drive which works well for sweep tubes. Orr has made a number of contributions to Amateur Radio over the decades and like all of us is entitled to make a mistake now and then.

AG6K
01-24-2012, 02:33 PM
It just seems to be contrary to common sense to do that...
Orr has produce3s some interesting stuff, not the least of which was his super cathode drive which works well for sweep tubes. Orr has made a number of contributions to Amateur Radio over the decades and like all of us is entitled to make a mistake now and then.

 Orr's reported history regarding a UCLA fraternity would make interesting reading if there was an Amateur Radio version of "The National Enquirer".

• Do you know how many articles by W6SAI were published in QST?

• Rich, ag6k

AF6LJ
01-24-2012, 02:42 PM
 Orr's reported history regarding a UCLA fraternity would make interesting reading if there was an Amateur Radio version of "The National Enquirer".

• Do you know how many articles by W6SAI were published in QST?

• Rich, ag6k

I know there were at least one I can remember off the top of my head.T
hat's where I first saw the super cathode drive article.

AG6K
01-24-2012, 03:31 PM
Dozens of people wrote to QST about your articles Rich, let's not single Orr out as a bad guy. Orr made mistakes, as we all do, but a few mistakes does not mean his input has no value.

The ARRL was more than fair.

 What was fair about refusing to publish my rebuttal to: Technical Correspondence, page 71, September 1994 QST, "Revisiting 'The Nearly Perfect Amplifier' " --- wherein you included yourself in your list of "our recognized amplifier experts". Was it just a coincidence that MFJ-Ameritron paid for 10 full pages of advertising in the following month's QST?



They worked hard to find anyone anywhere in the world with experience or credentials who agreed with you, but came up with nothing.


 then why did they ask that I write more articles, examples of which included "Parasitics Revisited" and "The Nearly Perfect Amplifier" ?


The ARRL had a stack of papers a foot tall


 A foot high stack of ordinary 3-mil paper would be c. 4000 pages. This seems like a bit of a stretch for an article that they told me they were pleased with before they sent the issue to the printer.


from dozens of people who responded, none of whom agreed with your articles. I would have thought if there was any technical merit, there would have been a stack at least equally high agreeing with you from reliable sources, yet there were none!

Rather than pick on Orr, why not point out some credible sources who agreed with you? Better yet, this was a quarter-century ago. Why not move on to something new?

 Because even today I am still not okay with seeing my fellow Amateur Radio operators needlessly toss their $$ away.



If in over 25 years no one with any experience or credentials has agreed with your articles, it is unlikely to change in the next 25 years.

Showing pictures of filaments relocated so they are half-way up in tube grids doesn't help your cause,

 In the s/n Q02888 photo, both filament helices were bent and partly intact, but they had not moved. The top part of each filament helix is not seen because it had shattered into bits and dropped into the base


and embarrasses the few who do support your claims by relying on those pictures as factual, complete, or accurate.

73 Tom

 who's embarrassed ?

• Rich, ag6k

AG6K
01-24-2012, 03:51 PM
I know what is inside of a stock SB -220
Fixed one for my EX the Freebander several years ago.
We had the Nichrome discussion (more like a shouting match).
He went off to pout...
I replaced the stock parasitic choke resistors and directly grounded the grids.
No more barking on forty meters.

So while I know what a stock 220 as I said a couple of posts down from that one I'll have more information after the weekend.

As far as I am concerned it's foolish to do anything bur directly ground the grids of those tubes.

 ... which is what I assumed until i started checking grid-resonance freq. with a dipmeter. After reading Fred Terman's "Radio Engineer's Handbook", Section 2, Circuit Elements, Inductance and Mutual Inductance, I realized that no matter how wide or how short a copper - or even a silver strap - is, it definitely has inductance. IOW, a literal Direct Ground is not to be found.

• Rich, ag6k

K9STH
01-24-2012, 08:36 PM
The truth be known, although Bill Orr, W6SAI, wrote a number of articles for QST, he had considerably more articles published in CQ Magazine. This besides a number of handbooks on topics concerning amateur radio.

I have to check to make sure, but I believe that Orr also wrote a few for 73 Magazine.

Glen, K9STH

G0HZU
01-24-2012, 11:14 PM
 ... which is what I assumed until i started checking grid-resonance freq. with a dipmeter. After reading Fred Terman's "Radio Engineer's Handbook", Section 2, Circuit Elements, Inductance and Mutual Inductance, I realized that no matter how wide or how short a copper - or even a silver strap - is, it definitely has inductance. IOW, a literal Direct Ground is not to be found.

• Rich, ag6k


That's why I'm puzzled by the people who think that grounding with copper tape is grounding the grid 'better' than the 200pF caps at med VHF frequencies. eg 70-110MHz.

The caps will outperform the copper tape over this range as the capacitive reactance of the caps cancels some of the inductive reactance in the grid connection.

Also, I think your dip meter method of measuring 'grid resonance' and also Tom's method of measuring 'grid resonance' are both flawed.

But that's just my opinion based on the physics of the tube and looking at the datasheet.

Can't somebody measure this stuff directly? I'm afraid I have no idea as to how hazardous the materials are in these tubes so maybe physical tests aren't practical or wise.

VK4TUX
01-24-2012, 11:31 PM
That's why I'm puzzled by the people who think that grounding with copper tape is grounding the grid 'better' than the 200pF caps at med VHF frequencies. eg 70-110MHz.

The caps will outperform the copper tape over this range as the capacitive reactance of the caps cancels some of the inductive reactance in the grid connection.

Also, I think your dip meter method of measuring 'grid resonance' and also Tom's method of measuring 'grid resonance' are both flawed.

But that's just my opinion based on the physics of the tube and looking at the datasheet.

Can't somebody measure this stuff directly? I'm afraid I have no idea as to how hazardous the materials are in these tubes so maybe physical tests aren't practical or wise.

Directly strapping to the chassis seems to provide better gain > amplication of the input signal, and a noticable smoother tuning process.

I dont dispute the strap inductance, however does your model show this additional gain achieved?

Adrian... vk4tux

VK4TUX
01-24-2012, 11:35 PM
84375
Looking at the base it seems the centre pin filament support is crimped to one side cathode electrode input and has slipped down through the crimp ?

Adrian ... vk4tux

W8JI
01-25-2012, 12:15 AM
84375
Looking at the base it seems the centre pin filament support is crimped to one side cathode electrode input and has slipped down through the crimp ?

Adrian ... vk4tux

They are welded, not crimped. That should never happen, so unless the tube was subject to many dozens or hundreds of G's from a drop on pins which would have shattered the grid, glass, and everything else, or unless the weld was pried on and broken after the tube was apart, it was a poor weld.

AG6K
01-25-2012, 02:20 AM
That's why I'm puzzled by the people who think that grounding with copper tape is grounding the grid 'better' than the 200pF caps at med VHF frequencies. eg 70-110MHz.

The caps will outperform the copper tape over this range as the capacitive reactance of the caps cancels some of the inductive reactance in the grid connection.

 True enough, but at 3.5MHz more grid bypass C is needed.



Also, I think your dip meter method of measuring 'grid resonance' and also Tom's method of measuring 'grid resonance' are both flawed.

 why is a dipmeter measurement of freq. flawed?



But that's just my opinion based on the physics of the tube and looking at the datasheet.

Can't somebody measure this stuff directly?

 What's more direct than a dipmeter -- and what could br less invasive?


I'm afraid I have no idea as to how hazardous the materials are in these tubes so maybe physical tests aren't practical or wise.

 Thorium (Th) is mildly radioactive. Th can be alloyed with U233 to generate electricity.
• Rich, ag6k

G0HZU
01-25-2012, 05:54 PM
I think that by putting the dip meter at a copper grid strap you will be measuring for a resonant dip that will be affected by whatever reactance is in the anode path to chassis and also whatever is in the cathode path to chassis.

I suppose you can call this 'grid resonance' but to me you are just measuring a dip that can be affected by factors external to the tube. So I'm not sure you can attribute this dip to any kind of internal 'grid resonance'.

AF6LJ
01-25-2012, 06:28 PM
I think that by putting the dip meter at a copper grid strap you will be measuring for a resonant dip that will be affected by whatever reactance is in the anode path to chassis and also whatever is in the cathode path to chassis.

I suppose you can call this 'grid resonance' but to me you are just measuring a dip that can be affected by factors external to the tube. So I'm not sure you can attribute this dip to any kind of internal 'grid resonance'.



You won't be able to couple enough energy into the grid itself and into the tube by coupling through a copper strap between the grid connections and ground.

AG6K
01-25-2012, 06:37 PM
I think that by putting the dip meter at a copper grid strap you will be measuring for a resonant dip that will be affected by whatever reactance is in the anode path to chassis and also whatever is in the cathode path to chassis.[

 Valid points but the affect will be minimal. The anode to gnd path is 5pF in series with C-Tune's 30 to 350pF, which varies C from 4.3pF to 4.9pF. which moves the grid resonance about 4%. As for the cathode path to chassis gnd, the relatively larger values of C2 in in the input Pi-network would probably not much move grid resonance from band to band.

QUOTE]
I suppose you can call this 'grid resonance' but to me you are just measuring a dip that can be affected by factors external to the tube. So I'm not sure you can attribute this dip to any kind of internal 'grid resonance'.[/QUOTE]

 No matter which contributes how much on which band, grid-resonance is important in a g-g amp because below this frequency the g-g circuit is self-neutralizing --- but above this frequency all bets are off since there is no method to achieve stabilization other than to artificially decrease VHF gain with a L/R VHF amplification suppressor. .
• Rich, ag6k

AG6K
01-25-2012, 06:41 PM
You won't be able to couple enough energy into the grid itself and into the tube by coupling through a copper strap between the grid connections and ground.

 I see a sharp dip c. 90MHz on 3-400Z and 3-500Z grids. Carl H., KM1H says 811As and 572B grids dip c. 75MHz.
• Rich, ag6k

AF6LJ
01-25-2012, 08:44 PM
 I see a sharp dip c. 90MHz on 3-400Z and 3-500Z grids. Carl H., KM1H says 811As and 572B grids dip c. 75MHz.
• Rich, ag6k

Yes but...
Do you actually see a sharp dip by coupling to the ground strap from the grid pin of a 3-500Z to ground?
I don't think so.

AG6K
01-26-2012, 02:15 AM
Yes but...
Do you actually see a sharp dip by coupling to the ground strap from the grid pin of a 3-500Z to ground?
I don't think so.

 Yes Catwoman. . . . It has L and C. Do 1/4 wave verticals resonante ?
• Rich, ag8k

AF6LJ
01-26-2012, 02:21 AM
 Yes Catwoman. . . . It has L and C. Do 1/4 wave verticals resonante ?
• Rich, ag8k
yes dear I know the strap is self resonant and you will see it if your dip meter will go high enough.
What I was saying....
in response to

http://files.qrz.com/static/images/misc/quote_icon.png Originally Posted by G0HZU http://files.qrz.com/static/images/buttons/viewpost-right.png (http://forums.qrz.com/showthread.php?p=2448343#post2448343)
I think that by putting the dip meter at a copper grid strap you will be measuring for a resonant dip that will be affected by whatever reactance is in the anode path to chassis and also whatever is in the cathode path to chassis.
"
I don't think enough energy can be coupled into that strap to see any other resonance other than the strap.

I forgot....
Yes quarter wave verticals stubs or an I beam falling through the air will resonate. :)

AF6LJ
01-26-2012, 02:23 AM
 I see a sharp dip c. 90MHz on 3-400Z and 3-500Z grids. Carl H., KM1H says 811As and 572B grids dip c. 75MHz.
• Rich, ag6k

I'm sure you do.
I don't expect to see that resonance by coupling into the ground strap grounding the grid in question....

AG6K
01-26-2012, 12:10 PM
I'm sure you do.
I don't expect to see that resonance by coupling into the ground strap grounding the grid in question....

 Anyone with a dipmeter and a little curiosity can see it. However, there are other dips there between the three ground straps, so temporarily disconnecting all but one ground strap will narrow things down to the grid itself. In a G-G amplifier, grid resonance is an important thing to know because it's above this frequency where squirrels can become legion -- provided that the tubes have enough gain in that region.
• Rich, ag6k

W8JI
01-26-2012, 01:17 PM
Anyone can find dips and spurious resonances all over the place in any multiband HF PA.

NONE of that means it will have parasitic, and it is easy to show there is not enough energy in a parasitic to damage a filament. It can only damage a grid or anode if the oscillation occurs over a long enough steady time to allow heating.

That's the way things behave in the real world.

73 Tom

AG6K
01-27-2012, 01:32 AM
I'm sure you do.
I don't expect to see that resonance by coupling into the ground strap grounding the grid in question....

 The grid grounding conductor is the only place where one can see the grid resonance dip.
Rich, ag6k

AF6LJ
01-27-2012, 01:35 AM
 The grid grounding conductor is the only place where one can see the grid resonance dip.
Rich, ag6k
I would like to see how that is done sometime....
Although I suppose it's much easier with a stock SB-220 or TL-922.

AG6K
01-27-2012, 03:29 PM
Anyone can find dips and spurious resonances all over the place in any multiband HF PA.

NONE of that means it will have parasitic, and it is easy to show there is not enough energy in a parasitic to damage a filament.

 Since AC current exerts zero net EMF on conductors, it is not RF (AC) energy that pushes the 1800บK, 3-500Z filament sideways, it is a brief burst of almost unbelievably high DC current. The high burst of DC grid-I – all of which comes from the filter-C through the cathode/filament, so the filament is subjected to high EMF. The condition that allows a through the roof burst of grid-I with a LPF tank is a VHF oscillation.
• So how do we know that the grid-I was sky high during big-bang events? Hmmm. . . Grid chokes that are wound with #30 Cu wire melt Cu and burn out. . . . . What is the fusing-I rating of #30 Cu?
• How do we know VHF energy was passing through the neighborhood? Hmmm . . . VHF-suppressor R-supps that are virtually shorted out at HF by 0.06ตH or so get zapped during big-bang events.
• Rocket science this is not boys and girl.



It can only damage a grid or anode if the oscillation occurs over a long enough steady time to allow heating.


• Grids and anodes in 3-500s are not what gets damaged Tom. The filament is already hot and it's the filament that gets bent/damaged by the not small EMF from a current-source that we know has a >1kA-peak I capability with modern low ESR filter caps.




That's the way things behave in the real world.

73 Tom

 In the "real world" does AC or DC exert a net EMF?
• Rich, ag6k

G0HZU
01-27-2012, 05:00 PM
Remember that I've never been near one of these amplifiers but even a basic analysis will show that if the amplifier oscillates strongly at VHF (i.e. lots of voltage swing at the anode) then there could be sufficient RF voltage across the grid choke to cause it to seriously overheat. This would be most likely if the resistor in the parasitic suppressor blew open circuit initially, then allowing LOTS of anode swing after the initial oscillation event that took out the suppressor resistor.

I would expect the 1mH grid choke to look like a very small capacitor in series with a highish resistance that varies across VHF. So if that resistance was ballpark 500R at the relevant VHF frequency you could get quite a few watts dissipating in the grid choke.


For example, if you had 120V rms across the choke and it looked like 2pF in series with 500R at 120MHz ,then the parallel equivalent resistance is 1400R meaning you get about 10W dissipated in the choke. That's going to kill it very quickly...

The other thing is that the 200pF grid caps will be subjected to very high stress as well and I'd expect these could break down if the VHF hoot was at the worst case frequency for the caps.

Do these caps ever show signs of damage or changes in characteristics over time?

G0HZU
01-27-2012, 05:10 PM
What would worry me would be what happens once the grid caps or the grid choke blow open circuit.

I don't know enough about tube theory but isn't there then the case for something nasty to happen in terms of a destructive transient due to uncontrolled cathode to grid voltages?

AF6LJ
01-27-2012, 06:46 PM
What would worry me would be what happens once the grid caps or the grid choke blow open circuit.

I don't know enough about tube theory but isn't there then the case for something nasty to happen in terms of a destructive transient due to uncontrolled cathode to grid voltages?

What happens when the grid circuit of a vacuum tube is open circuited it floats to a negative value and the tube is cut off. Without a DC return enough negative bias will develop to not only stop amplification but oscillation should also should be impossible.

G0HZU
01-27-2012, 07:23 PM
Hi Sue
Thanks for the info :)

AG6K
01-27-2012, 10:23 PM
Remember that I've never been near one of these amplifiers but even a basic analysis will show that if the amplifier oscillates strongly at VHF (i.e. lots of voltage swing at the anode) then there could be sufficient RF voltage across the grid choke to cause it to seriously overheat. This would be most likely if the resistor in the parasitic suppressor blew open circuit initially, then allowing LOTS of anode swing after the initial oscillation event that took out the suppressor resistor.

I would expect the 1mH grid choke to look like a very small capacitor in series with a highish resistance that varies across VHF. So if that resistance was ballpark 500R at the relevant VHF frequency you could get quite a few watts dissipating in the grid choke.


For example, if you had 120V rms across the choke


 In a factory-stock SB-220 or TL-922 there's 600pF from the grid and the grid-choke to gnd. At 120MHz, 600pF has 2.4Ω of XC. This means there would be 120v/2.4Ω = 50Amperes through the capacitors @120MHz. What's wrong with this picture?
Rich, ag6k.



and it looked like 2pF in series with 500R at 120MHz ,then the parallel equivalent resistance is 1400R meaning you get about 10W dissipated in the choke. That's going to kill it very quickly...

The other thing is that the 200pF grid caps will be subjected to very high stress as well and I'd expect these could break down if the VHF hoot was at the worst case frequency for the caps.

Do these caps ever show signs of damage or changes in characteristics over time?

AG6K
01-27-2012, 10:30 PM
yes dear I know the strap is self resonant and you will see it if your dip meter will go high enough.
What I was saying....
in response to

http://files.qrz.com/static/images/misc/quote_icon.png Originally Posted by G0HZU http://files.qrz.com/static/images/buttons/viewpost-right.png (http://forums.qrz.com/showthread.php?p=2448343#post2448343)
I think that by putting the dip meter at a copper grid strap you will be measuring for a resonant dip that will be affected by whatever reactance is in the anode path to chassis and also whatever is in the cathode path to chassis.
"
I don't think enough energy can be coupled into that strap to see any other resonance other than the strap.

I forgot....
Yes quarter wave verticals stubs or an I beam falling through the air will resonate. :)

 Correctomundo. A floating I-beam resonates at a half-wave, not at a quarter-wave - so the dipper needs to have a low-C output connection through a few pF cap in order to measure the dip at the end - or one could measure the dip with a plug-in coil at the center of the I-beam.
• Rich, ag6k

AF6LJ
01-27-2012, 10:51 PM
 Correctomundo. A floating I-beam resonates at a half-wave, not at a quarter-wave - so the dipper needs to have a low-C output connection through a few pF cap in order to measure the dip at the end - or one could measure the dip with a plug-in coil at the center of the I-beam.
• Rich, ag6k
I know that :)

I'll have more information on the amplifier tomorrow morning.
I should have answers for the questions up by ten or so if not before.

As for the floating I beam;
If an I beam is resonating and there is nobody around to measure it,...
is it really resonant. ;)
Have a good day Rich.

G0HZU
01-28-2012, 12:32 PM
 In a factory-stock SB-220 or TL-922 there's 600pF from the grid and the grid-choke to gnd. At 120MHz, 600pF has 2.4Ω of XC. This means there would be 120v/2.4Ω = 50Amperes through the capacitors @120MHz. What's wrong with this picture?
Rich, ag6k.

The picture taker did a simple low frequency analysis that summed three 200pF capacitors together and assumed the same analysis applied to real components and layout at VHF?

W8JI
01-28-2012, 03:39 PM
The picture taker did a simple low frequency analysis that summed three 200pF capacitors together and assumed the same analysis applied to real components and layout at VHF?

That's absolutely correct, and the picture taker does that with everything. Over simplifying the system and ignoring limits, like emission current limits, allows some very wild and impossible theories like bending elements from oscillations.

While the multilayer stack inside the capacitor case is pretty broad in bandwidth, the leads and multiple paths add another dimension.

73 Tom

AF6LJ
01-28-2012, 05:27 PM
To answer people's questions regarding the SB-220.....
The grids of said 3-500Zs are connected to ground through 27 Ohm quarter watt resistors which are bypassed to ground.
The amplifier has a Harbach keying interface board, soft start, and high voltage board.
The caps are original.
The tube shorted from the sudden stop not from the initial act of moving the amplifier.
Sorry Rich I forgot to ask about the tube but I shall do that in the morning....
And I forgot to ask which tube it was....
The only items that were replaced were the tube and the twenty seven ohm resistor.

AG6K
01-28-2012, 06:47 PM
The picture taker did a simple low frequency analysis that summed three 200pF capacitors together and assumed the same analysis applied to real components and layout at VHF?

 I did not compose the picture. The three 200pF capacitors are not in parallel?
• Rich, AG6K

AG6K
01-28-2012, 06:53 PM
That's absolutely correct, and the picture taker does that with everything. Over simplifying the system and ignoring limits, like emission current limits, allows some very wild and impossible theories like bending elements from oscillations

 The osc. does not bend the filament/cathode, it's the accompanying D.C. current pulse. .

.

While the multilayer stack inside the capacitor case is pretty broad in bandwidth, the leads and multiple paths add another dimension.

73 Tom

 Where's the error Tom?
• Rich, ag6k.

G0HZU
01-28-2012, 09:38 PM
If you look at how long the 200p cap leads are (were) in this SB-220 they are over an inch long.

http://www.w6kan.com/sb221under.JPG


I was quite shocked at how long and skinny the cap connections were. So there could be 30nH in the cap leads? This inductance has 23 ohms
reactance at 120MHz. The 200pF part of the cap has about -7 ohms reactance so the net reactance of each leaded 200pF cap is about 16 ohms (inductive) at
120MHz.

There's even more inductance in series with each cap if you include the socket tabs and the socket pins and the tube internal grid wire.

AF6LJ
01-28-2012, 09:51 PM
If you look at how long the 200p cap leads are (were) in this SB-220 they are over an inch long.

http://www.w6kan.com/sb221under.JPG


I was quite shocked at how long and skinny the cap connections were. So there could be 30nH in the cap leads? This inductance has 23 ohms
reactance at 120MHz. The 200pF part of the cap has about -7 ohms reactance so the net reactance of each leaded 200pF cap is about 16 ohms (inductive) at
120MHz.

There's even more inductance in series with each cap if you include the socket tabs and the socket pins and the tube internal grid wire.

The SB-220 My Ex had didn't look like that at all.
The caps were soldered to the tube pins and ground lugsunder the socket mounting screws. The leads were quite short. There also wasn't all that buss wire running around.....

AF6LJ
01-28-2012, 11:12 PM
I would also add that amplifier is a poor example of proper constructions practice.

G0HZU
01-29-2012, 12:12 PM
Assuming these amps were built as kits then I suppose there will be variability in the way the caps are fitted. If some have long leads on the caps and some short then it will means some amps will be more unstable than others in the 'usual' range of frequencies they go unstable.

I found that amp (and one or two others) by googling images of the amp and the caps leads were quite long on all of them.

I am spending far too much time looking at this issue :) and really should be doing other things but I've found Tom's web pages and also pages written by Rich and whilst there LOTS of good stuff on their pages I do think both are a bit dated and vague in their instability analysis and there's some things I don't agree with.

It looks like Tom has an HP4396A/B VNA and this VNA would be my VNA of choice for analysing these amplifiers because it supports an external coupler and it is also capable of exporting S parameter data. I haven't used one of these old combo VNA/SA for many years but they are very versatile instruments. I can remember the company buying one of the A versions in the early 1990s and they were pretty awesome. The A had the CRT display and I can remember fighting to be the one to get to use it as much as possible!

I can't remember but you might be able to export the S parameter data direct to floppy but otherwise you would have to do it the same way I do with my old HP8714B via the GPIB interface to a PC.

From memory, the HP4396A (and B?) are a bit fiddly to set up even to export screen image plots to a floppy and somewhere I have a document outlining how to do it (doc must be 15 years old!)

AG6K
01-29-2012, 12:34 PM
If you look at how long the 200p cap leads are (were) in this SB-220 they are over an inch long.

http://www.w6kan.com/sb221under.JPG


I was quite shocked at how long and skinny the cap connections were. So there could be 30nH in the cap leads? This inductance has 23 ohms
reactance at 120MHz. The 200pF part of the cap has about -7 ohms reactance so the net reactance of each leaded 200pF cap is about 16 ohms (inductive) at
120MHz.

There's even more inductance in series with each cap if you include the socket tabs and the socket pins and the tube internal grid wire.

 Good points. However, trying to move grid resonance higher in freq. to hopefully widen the G-G's self-neutralizing area enough to matter is an exercise in futility since going with a grid bypass C that is small enough to substantially move the grid-resonance up in freq. - making it more VHF-stable, makes the amplifier lose too much gain at HF. . . Since we can't do anything to change C-fb, the only practical way to improve VHF stability is to additionally lower VHF- amplification by going with an even lower-Q VHF-suppressor -- and live with the reality that we are going to lose an additional 2% or so at 28MHz. (0.1db or 1/60 of 1 S-Unit)
• Rich, ag6k

G0HZU
01-29-2012, 01:28 PM
Hi Rich
I am still struggling a bit with the (commonly held?) theories about 'grid resonance' and it's impact on stability.

We all know that instability is an issue if the internal grid and the chassis ground have (typically) inductance between them.
But what I'm struggling with are your test methods and also Tom's test methods and also the subsequent conclusions about the results.

However, I do have another question. How close is the cathode filament to the grid structure (gap size?) on a 3-500Z and what is the breakdown voltage for this gap? I know it's in a vacuum but I was wondering what the BDV would be (especially at RF).

Jeremy
G0HZU

AG6K
01-29-2012, 02:04 PM
 RE: Autopsy of virtually-unused Eimac 3-500Z serial #s AMK096 and AMK085 (date-codes 8504) - which have bent filaments. Yesterday Helper Girl and I spent most of the day getting the water pump recall taken care of for her Prius and repairing the 8000v hi-pot tester, so we did not autopsy any tubes. However, we discussed trying to centrifuge the two Eimac tubes with bent filaments since they are otherwise in new condition. If the gas-test goes well today we are going to try centrifuging the tubes to straighten out their bent filaments. If we are able to get the Filament/grid BDV safely above 5000v by centrifuging them, we are going to try them at 7128KHz.
• Rich, ag6k

AG6K
01-29-2012, 02:08 PM
 Photos of virtually-unused Eimac 3-500Zs with bent filaments:
8467584676
• Rich, ag6k

AF6LJ
01-29-2012, 02:31 PM
 RE: Autopsy of virtually-unused Eimac 3-500Z serial #s AMK096 and AMK085 (date-codes 8504) - which have bent filaments. Yesterday Helper Girl and I spent most of the day getting the water pump recall taken care of for her Prius and repairing the 8000v hi-pot tester, so we did not autopsy any tubes. However, we discussed trying to centrifuge the two Eimac tubes with bent filaments since they are otherwise in new condition. If the gas-test goes well today we are going to try centrifuging the tubes to straighten out their bent filaments. If we are able to get the Filament/grid BDV safely above 5000v by centrifuging them, we are going to try them at 7128KHz.
• Rich, ag6kI'm looking forward to hearing the results.

AG6K
01-29-2012, 03:06 PM
I'm looking forward to hearing the results.

 However Catwoman, in order to get to the promised land with a useful 3-500 in hand, the light from the hi-pot's fil/grid arc has to be visible between the fins in the anode cooler and the 27-yr old seals have to be wonderful.

AF6LJ
01-29-2012, 03:35 PM
 However Catwoman, in order to get to the promised land with a useful 3-500 in hand, the light from the hi-pot's fil/grid arc has to be visible between the fins in the anode cooler and the 27-yr old seals have to be wonderful.
Stands to reason...
Besides without an arc how is one to know which way to orientate the tube in the centrifuge which the Hi-Pot will also tell you.

W8JI
01-30-2012, 03:35 PM
Stands to reason...
Besides without an arc how is one to know which way to orientate the tube in the centrifuge which the Hi-Pot will also tell you.

When an off-center or off-balance object is spun on axis, it gets worse.

When placed in a centrifuge, you'd have to know which way something was off and how much to spin it. I think more than likely, if a tube ever really does get fixed, there is just something loose in the tube (like a broken piece of grid) that gets dislodged.

Most case, you can just gently tap on tubes and make the shorts come and go.

AG6K
01-30-2012, 07:59 PM
When an off-center or off-balance object is spun on axis, it gets worse.

 Correct. The tube is held by rubber bands in a padded cradle with the filament bulge toward the axis of the centrifuge. The cradle is fixed the to the arm of the centrifuge. When the arm spins, G-force pushes the filament bulge the other way to straighten it out



When placed in a centrifuge, you'd have to know which way something was off


 Which is the trickiest part of the process Tom. Since the anode obstructs the view, it takes someone with extraordinary low-light vision to determine the direction of the indirect light coming from the arc between the filament and the grid. . . Since graphite-anode 3-500s do not have hollow cooling fins that allows one to view the grid & filament, we have never attempted to straighten the bent filament in one.


and how much to spin it. I think more than likely, if a tube ever really does get fixed, there is just something loose in the tube (like a broken piece of grid) that gets dislodged.

 The molybdenum grid is tough. We have never damaged a grid at 11G.



Most case, you can just gently tap on tubes and make the shorts come and go.

 True, but in order to be useful there needs to be at least 5000v of fil/grid clearance -and that means one or more trips to the centrifuge.
• Rich, ag6k

G0HZU
02-01-2012, 03:53 PM
Which is the trickiest part of the process Tom. Since the anode obstructs the view, it takes someone with extraordinary low-light vision to determine the direction of the indirect light coming from the arc between the filament and the grid. . . Since graphite-anode 3-500s do not have hollow cooling fins that allows one to view the grid & filament, we have never attempted to straighten the bent filament in one.



Apologies in advance if this suggestion is wacky but could you fit current sensors in line with each of the three grid pins and capture the (arc discharge) current pulses (on a 3 channel DSO?) and maybe 'DF' where it arced based on the current sharing between the three grid pin results?

You wouldn't need to see inside the tube with this test?

You would be looking for very tiny differences but you could maybe get a reading with the right test setup.

G0HZU
02-01-2012, 04:38 PM
I guess the above method would get less accurate the higher up the grid it arced. The other wacky method would be to see if you could stick external sensors around the tube exterior and see if there is an assymetry in any external field set up by the arc. But again, the difference would be very tiny and hard to quantify. (hard because the grid acts a bit like a Faraday cage?)

AF6LJ
02-01-2012, 06:38 PM
Apologies in advance if this suggestion is wacky but could you fit current sensors in line with each of the three grid pins and capture the (arc discharge) current pulses (on a 3 channel DSO?) and maybe 'DF' where it arced based on the current sharing between the three grid pin results?

You wouldn't need to see inside the tube with this test?

You would be looking for very tiny differences but you could maybe get a reading with the right test setup.

I don't think that would work since the grid is terminated in a circular peace of stamped metal the pins are welded to.

AG6K
02-01-2012, 07:43 PM
I don't think that would work since the grid is terminated in a circular peace of stamped metal the pins are welded to.

 Good point Sue. Another way to locate the bulge is the gravity-utilitizing technique invented by Jennings for checking the concentricity of the Cu plates in vacuum variable capacitors: Connect the C in parallel with the L of a variable frequency oscillator / freq.-counter, with the tuning shaft horizontal. Rotate the C. When the freq. reading is lowest, the bulge side is down --- or when the freq. reading is highest, the bulge side is up. . . Jo E. Jennings, W6EGV, was definitely a sharp guy.
•• Rich, ag6k

G0HZU
02-01-2012, 08:47 PM
I don't think that would work since the grid is terminated in a circular peace of stamped metal the pins are welded to.

Oh I see. I don't really know much about the internal construction so maybe that was a bad idea...

I suppose you could use the Jennings method in the centrifuge if you fitted a JFET osc to a socket that used the tube filament/grid as part of the tuned circuit.

Then run it with a tiny battery and read the frequency on a nearby spectrum analyser (or maybe even look at FM on a demod meter) ?

Run the centrifuge at medium speed and try various rotation angles of the tube and pick the one that gives the biggest expected shift in freq.
Then turn up the angular velocity to get max g

You might even be able to monitor progress with this method? :)

AG6K
02-02-2012, 01:10 AM
Oh I see. I don't really know much about the internal construction so maybe that was a bad idea...

I suppose you could use the Jennings method in the centrifuge if you fitted a JFET osc to a socket that used the tube filament/grid as part of the tuned circuit

 The tube has to be at rest to determine the filament/grid spacing.

.

Then run it with a tiny battery and read the frequency on a nearby spectrum analyser (or maybe even look at FM on a demod meter) ?

Run the centrifuge at medium speed and try various rotation angles of the tube and pick the one that gives the biggest expected shift in freq.
Then turn up the angular velocity to get max g

You might even be able to monitor progress with this method? :)

VK4TUX
02-02-2012, 01:47 AM
 The tube has to be at rest to determine the filament/grid spacing.

.

A system using an electronic balance system similar in principle to the method tyres are balanced could be used.
If a identical new tube is used in in a tight pin socket, and then balanced with weights on the socket plate for perfect balance spun on the centre axis of the filament helix then this could provide the test platform.

Once the perfectly weighted socket is setup, the suspect tube is fitted in place of the new perfect tube, and the electronic balance machine weight data/location required for balance would tell you what you need to know.

Adrian ... vk4tux

AG6K
02-02-2012, 02:01 PM
A system using an electronic balance system similar in principle to the method tyres are balanced could be used.
If a identical new tube is used in in a tight pin socket, and then balanced with weights on the socket plate for perfect balance spun on the centre axis of the filament helix then this could provide the test platform.

Once the perfectly weighted socket is setup, the suspect tube is fitted in place of the new perfect tube, and the electronic balance machine weight data/location required for balance would tell you what you need to know.

Adrian ... vk4tux

 A clever method - but one would likely need a milligram-resolution scale to do it Adrian. Maybe an x-ray machine would do the job?

G0HZU
02-02-2012, 04:25 PM
 The tube has to be at rest to determine the filament/grid spacing.

.

Can you explain why it has to be at rest?

The Jennings method would attempt to tell you which orientation make the gap worse. You aren't trying to determine the actual spacing with this test. i.e. you are just finding the correct orientation for the centrifuge.

However, by slowing down and speeding up you could see if there was a permanent shift in frequency indicating that something had permanently 'moved'

I guess you have used the centrifuge successfully but I'm nonetheless a bit surprised it is able to straighten the filament. You won't get much force with 11g on such a small structure (a few grams in mass?)

Are the filaments quite pliable?

AG6K
02-03-2012, 08:33 AM
Can you explain why it has to be at rest?

 Since the filament is a double-helix suspended from the ends, my guess is that the grid/filament spacing is affected by 11G-lateral.



The Jennings method would attempt to tell you which orientation make the gap worse. You aren't trying to determine the actual spacing with this test. i.e. you are just finding the correct orientation for the centrifuge.

 Correct. Once the direction of the bend is determined, there is no reason to backtrack. At this point I'm thinking X-ray for finding the direction of the bend.



However, by slowing down and speeding up you could see if there was a permanent shift in frequency indicating that something had permanently 'moved'

 at 1900บK and 11G most metals can be unbent.



I guess you have used the centrifuge successfully but I'm nonetheless a bit surprised it is able to straighten the filament. You won't get much force with 11g on such a small structure (a few grams in mass?)


 Sometimes we overshot the mark and had to do a 180 in the centrifuge to bend it back to straight.



Are the filaments quite pliable?

 At 1900บK they can be bent but I would never rate Th-W as pliable since W alloys are apparently pretty tough.

AG6K
02-03-2012, 08:39 AM
 I'm presently seeking an out-of-warranty 3-500ZG with a bent filament (filament-grid short) to see if I can determine the direction of the bend with an x-ray machine and possibly straighten the bent filament with a centrifuge.

AG6K
02-12-2012, 01:47 PM
 Bad news: We 3 could not see which way the filament is bent in either 3-500Z s/n AMK 096 of s/n AMK 085. I subsequently measured the filament/grid C at c. 12pf in each tube. Since normal filament/grid C is c. 8.5pF, 12pF confirms that the filament helices are not in the center of the grid cage. Running these tubes in a factory-stock SB-220 or TL-922 would be risky since the filaments get even closer to the grounded-grids when they are hot, increasing the chance of a filament to grounded-grid short also shorting out the+110v PS - which would burn out unfused T-2, the combo filament- xfmr/+110vdc PS-xfmr. The next step is to try running the tubes with bent filaments in a SB-220 that has been modified from the original V-cutoff bias to R-cutoff bias.

AG6K
02-13-2012, 11:19 PM
 UPDATE: The X-ray image was sharp enough but the view of the bent 3-500Z filament was blocked by the shadow of the chrome-plated brass sleeve over the anode lead. The set screw for the brass sleeve is not a Torx, not an Allen and not a Bristol. Does anyone know where to buy a wrench that will fit? Tnx

AF6LJ
02-13-2012, 11:34 PM
 UPDATE: The X-ray image was sharp enough but the view of the bent 3-500Z filament was blocked by the shadow of the chrome-plated brass sleeve over the anode lead. The set screw for the brass sleeve is not a Torx, not an Allen and not a Bristol. Does anyone know where to buy a wrench that will fit? Tnx
I would think it would have to be one of the above...
Has anybody looked at the socket (set screw head) with a powerful magnifying glass?

N3JBH
02-14-2012, 12:45 AM
 UPDATE: The X-ray image was sharp enough but the view of the bent 3-500Z filament was blocked by the shadow of the chrome-plated brass sleeve over the anode lead. The set screw for the brass sleeve is not a Torx, not an Allen and not a Bristol. Does anyone know where to buy a wrench that will fit? Tnx

Rich there is several types of tools used to drive various bolt's , screws etc. Could you post a picture of it and possibly get some measurement's?

AG6K
02-24-2012, 12:28 AM
Rich there is several types of tools used to drive various bolt's , screws etc. Could you post a picture of it and possibly get some measurement's?

 it takes a Bristol wrench. I have the right size.

AG6K
02-24-2012, 12:32 AM
 X-ray of a 3-500Z with filament to grid short:86909

AF6LJ
02-24-2012, 12:37 AM
 X-ray of a 3-500Z with filament to grid short:86909
You can actually see it....
Wow...
Glad you found out what wrench was needed to remove the anode sleeve.

AG6K
02-24-2012, 12:43 AM
 A better x-ray:
86919

AF6LJ
02-24-2012, 12:52 AM
 A better x-ray:
86919
That is a better X-Ray, Have you tried to straighten the filament in this one yet?

N3JBH
02-24-2012, 02:33 AM
Rich @ 2960.33 degree Fahrenheit there is not a lot of metals that wont move

AG6K
02-27-2012, 01:01 AM
Rich @ 2960.33 degree Fahrenheit there is not a lot of metals that wont move

 agreed Jeff.

AG6K
02-27-2012, 01:11 AM
That is a better X-Ray, Have you tried to straighten the filament in this one yet?

 We tried today. The X-ray was no help. We found the direction of the bend by increasing the current in the filament-grid arc. After centrifuging the filament-grid BDV increased about 30%. Next time I raise the filament V. and find someone who can spin it faster.

W1QJ
02-27-2012, 12:29 PM
Rich, are any of these the tubes I sent you? Lou

AG6K
02-29-2012, 06:14 PM
Rich, are any of these the tubes I sent you? Lou


 No Lou. We have been experimenting with the 3 we had on hand. Things have been going pretty slow lately since helper girl #3 has started cosmetology school and helper girl #4 - a.k.a. Wonderwoman - is expecting.

AF6LJ
02-29-2012, 06:17 PM
 We tried today. The X-ray was no help. We found the direction of the bend by increasing the current in the filament-grid arc. After centrifuging the filament-grid BDV increased about 30%. Next time I raise the filament V. and find someone who can spin it faster.

Might be tine to build a centrifuge, there are plenty of washing machines out there that could be converted for this application.

AG6K
02-29-2012, 09:37 PM
That is a better X-Ray, Have you tried to straighten the filament in this one yet?

 Yes. It improved the filament/grid BDV about a third so it seems we are moving it in the right direction.

ad: KU4HP-1