PDA

View Full Version : SB-220 Problems



AH6RR
04-15-2011, 03:42 AM
I have a SB-220 that has developed a problem. When I first turn it on it works just fine with full output. After 4 or 5 transmissions the input SWR jumps up to over 3:1 and no output. I can turn it off wait for 5 to 10 minutes and it repeats it self.
I disassembled and cleaned and polished the relay but not sure if that was the problem it is the original relay so maybe that is the problem. I hope it is not a tube since they are the original Eimac's and still have full output. Any suggestions on what it might be?

Roland AH6RR

WA7PRC
04-15-2011, 05:23 PM
Hi Roland,

You didn't state whether this happens on one band or all bands. If it's one band, consider that one of the silver mica capacitors in the input circuit may be overheating. If it happens on more than one band, look elsewhere.

A weak spot with 3-500Z tubes is the filament pins. The wire leads have nickel-plated brass sleeves soldered to them. If cooling is insufficient, they will get hot. They overheat, melt the solder, and then things stop working properly. This could manifest itself as a change in input VSWR with heating. In an extreme case, you can overheat the socket contacts like this:

http://farm5.static.flickr.com/4104/4961553319_c219a0d461_t.jpg (http://www.flickr.com/photos/bswadener/4961553319/)
(click for larger image)

Note the good grid contact on the left and the overheated tensioner on the bad filament contact on the right. On the bad contact, the contacts don't quite close.

It's been said that the Heath cooling scheme is pretty good but, there's not much overhead. If the fan motor doesn't turn freely, the bearings need cleaning & reoiling. If the bearings are loose, it's time for a new motor. Grainger sells the Dayton 4M070 (http://www.grainger.com/Grainger/DAYTON-CFrame-Motor-4M070), which will move more air. K5DBX came up with a shroud that minimizes fan noise. See the last post in this QRZ thread (http://forums.qrz.com/showthread.php?252965-Noisy-Fan-on-SB-221). I haven't tried it (yet).

AH6RR
04-15-2011, 06:13 PM
The amp has new tube sockets and I have jut recently did a complete rebuild. It does it on all bands I will check the solder joints again. I am still thinking it is the relay since before I cleaned it I checked the input side and it was kind of flakey but seemed good after the cleaning. The amp has the Harbach fan and runs very cool and the HV is good. I will check the input band switch.
When I rebuilt the amp from the ground up I completely striped the thing to a bare chassis. The amp had sat in a open carport with a blanket over it in the Hilo Rain Forrest where the humidity is very high. It needed a complete rebuild because of mold and dirt the paint was total junk. The good part is it had the Harbach Filter caps, Soft-start and fan it also has a Peter Dalh filament transformer I added the Harbach Rectifier/Metering board. I replaced the tube sockets because the old ones were corroded and looked like junk. All the old wiring was replaced with Silver Teflon wire (except the transformer's) the SO-239's were replaced with new silver/Teflon ones, the grids were directly grounded with silver/copper braid (from RG-214). The input coax was replaced with new RG-8X from the SO-239 to the relay and new Belden 9907 (silver braid RG-58) for both of the input bandswitch coax's. The output coax was replaced with a piece of RG-214. The only thing I did not replace was the relay I did just order an new one today. I used the amp for the CQWPX contest without a problem. The problem developed 12 day's later.
I will completely go over the input circuit when I replace the relay.

Roland AH6RR

WA7PRC
04-15-2011, 07:31 PM
Is the problem on one band or more than one band?

AH6RR
04-15-2011, 09:02 PM
Is the problem on one band or more than one band?

It does it on all bands

WA7PRC
04-16-2011, 07:35 PM
Your amplifier has a thermal issue. Relay contacts don't have thermal issues. Capacitors have series resistance (ESR) and will dissipate power with current passing through them. But, since it's happening on more than one band, the input matching network (and bandswitch) is ruled out. That leaves the coupling capacitors C32 or C21 (between the input matching network and the tubes).

I wonder what you meant by "polishing" the relay contacts. Most relay contacts have a thin plating that can't take much abrasion/burnishing.

AH6RR
04-17-2011, 02:46 AM
What I did is disassemble the relay and used a Dremel tool polishing kit ( Soft pads and metal polish) to buff the contacts to a nice shine. I will check the input caps.

W8JI
04-18-2011, 02:54 PM
I have a SB-220 that has developed a problem. When I first turn it on it works just fine with full output. After 4 or 5 transmissions the input SWR jumps up to over 3:1 and no output. I can turn it off wait for 5 to 10 minutes and it repeats it self.
I disassembled and cleaned and polished the relay but not sure if that was the problem it is the original relay so maybe that is the problem. I hope it is not a tube since they are the original Eimac's and still have full output. Any suggestions on what it might be?

Roland AH6RR
\

I don't know why so many people polish or clean relay contacts for problems that cannot be the relay. Relay contact "dirt" only shows up on receive. Dirty contacts in anything, including switches and relays, would never be an "operate for a while and then high SWR" type problem. We need a big billboard somewhere discouraging cleaning of contacts. :-)

If you have a time failure problem with SWR increasing on all bands, it cannot be input capacitors. Most likely it is a bad tube going grid/filament short.

If the tube is shorted you will notice idle current pop up at the same time as SWR, almost always showing high idle current without the TR relay being closed.

There are a few other potential causes, like a bad connection on the input coax to the tube or perhaps an open zener or wire from the transformer CT to the bias and metering system.

The thing that would pinpoint your problem is what the plate and grid current meters do when the amp acts up. By looking at the plate and grid currents during the failure and noting what they are with the amp keyed and unkeyed, it is possible to virtually 100% know what the problem is.

My best guess with the very limited information given is a bad tube. A better guess would require knowing what the plate and grid meters read at the time of failure, but I can say it is virtually 100% NOT likely to be a capacitor. Don't waste time there.

73 Tom

AH6RR
04-18-2011, 08:59 PM
\

I don't know why so many people polish or clean relay contacts for problems that cannot be the relay. Relay contact "dirt" only shows up on receive. Dirty contacts in anything, including switches and relays, would never be an "operate for a while and then high SWR" type problem. We need a big billboard somewhere discouraging cleaning of contacts. :-)

If you have a time failure problem with SWR increasing on all bands, it cannot be input capacitors. Most likely it is a bad tube going grid/filament short.

If the tube is shorted you will notice idle current pop up at the same time as SWR, almost always showing high idle current without the TR relay being closed.

There are a few other potential causes, like a bad connection on the input coax to the tube or perhaps an open zener or wire from the transformer CT to the bias and metering system.

The thing that would pinpoint your problem is what the plate and grid current meters do when the amp acts up. By looking at the plate and grid currents during the failure and noting what they are with the amp keyed and unkeyed, it is possible to virtually 100% know what the problem is.

My best guess with the very limited information given is a bad tube. A better guess would require knowing what the plate and grid meters read at the time of failure, but I can say it is virtually 100% NOT likely to be a capacitor. Don't waste time there.

73 Tom

Before the SWR starts to increase I get 150mils of plate current and 0 mils grid current with no modulation and 450mils plate and 75 mils grid with modulation and 1400W out. Then SWR starts to rise for very short (less than a second) time, then rises to full scale and the Plate current is 0 mils and the Grid is 0 mils with or without modulation or keyed or un-keyed. When I let it sit for a couple of minutes letting the fan cool it down while running it will work again for a few more minutes then the SWR starts to rise again. I at first thought it was a tube also but I put them in another amp and they are fine in that amp. I will the diodes in the bias circuit on the Harbach board and also the centertap. It does look to be thermal related but I'll be danged if I can find it.
Roland AH6RR

W8JI
04-19-2011, 03:48 PM
Then SWR starts to rise for very short (less than a second) time, then rises to full scale and the Plate current is 0 mils and the Grid is 0 mils with or without modulation or keyed or un-keyed.


There you go!! That's the critical information that was missing.

The lack of quiescent current shows it is NOT tubes, and is not anything except an open in the path from the filament transformer center tap through the bias to the HV PS negative buss.

You either have a CT on the transformer that is opening, a poor solder joint in that path someplace, a bias diode that is opening, or maybe (unlikely) a relay bias contact that is opening. One way to find this is with a little visual inspection and substitution.

Meter readings really help a lot in pinning down problems. Without any doubt at all, the cathode current path from the filament transformer CT to the metering/ HV power supply negative rail is going open. Most likely in that path is the bias diode going open, and second most likely a bad solder joint. Just hope it is not inside the filament transformer! :-)

73 Tom

AH6RR
04-19-2011, 04:48 PM
Thanks for the reply Tom I hope it is not the filament transformer either since this one is somewhat new (last 5 years) Peter Dalh. I am going to put it on the bench today and do a thorough going over on the Bias circuit.

Roland

AH6RR
04-20-2011, 06:58 AM
After checking the Bias circuit out I did find that the TX contact on the relay showed signs of poor contact even after I had cleaned it it was black and does have a pit in the middle that was there when I got the amp. I cleaned it up and re-soldered all the joints in the Bias circuit. I checked the Filament transformer and it seems fine. I the Bias diodes are good. I have been using it for 40+ minutes now and it is working as it should. I'll keep my fingers crossed. I did order a new relay just in case.

Thanks
Roland AH6RR

K9FV
04-24-2011, 02:55 AM
Darn Tom is good at diagnosing amps issues! Sounds like he hit the nail on the head with the bias?

Congrats on a job well done Roland -

73 de Ken H>

KM1H
04-24-2011, 05:41 PM
Since the usual diode failure mode is a short that eliminated it as a culprit. Other common problems in that circuit is a weak relay contact (they lose their spring tension as they get old, just like us) or a leaky bias electrolytic.

Replacing the relay is the only sure fix when they act up.

Carl

AH6RR
04-24-2011, 06:33 PM
63293
Since the usual diode failure mode is a short that eliminated it as a culprit. Other common problems in that circuit is a weak relay contact (they lose their spring tension as they get old, just like us) or a leaky bias electrolytic.

Replacing the relay is the only sure fix when they act up.

Carl

Carl,
I agree with you the relay has to be at least 30+ years old and since that is one of the only things I did not replace in the amp might as well take care of that too. The bias electrolytic is new as are most of the other parts in the amp. After the complete rebuild I used it in the CQWPX for around 25 hours total and it worked fine with a couple of hick-ups nothing major and all were related to the relay(loss of receive). But all it took then was hit the foot swicth a couple of times. Before putting the amp back on the air I did the paper slide across the relay contacts to clean them and that seemed OK. When the problem in this thread showed up I did disassemble the relay and buffed the contacts I did notice at that time that the bias contact had a pit in the middle of it on the TX side. Then when it started again and the relay was taken apart I could see the black arcing results of poor contact cleaned that and added a small amount of Silver Solder to fill the pit on the contact and it is working again. But it will get a new relay. I have added a picture of the amp I know it is not HeathKit green but I like it better then the green.

WA7PRC
04-25-2011, 02:43 AM
Good job, Roland. I've never known relay contacts to be temperature sensitive but apparently, I was wrong.

But, I could've done without seeing a non-green SB-220. :p
I'm working on building a monoband amplifier into a SB-220 outer case. It will mimic the Heath color scheme. Yes, to match my SB-220, it will be green. :D

W8JI
04-25-2011, 10:44 AM
Roland,

In another section I relied about the AL811H relay in detail, which also fits the SB220 relay exactly. Here is the reply, to that I would like to add "weak spring" as a cause that I have never once seen in the past 30 years or more, although I suppose a weak spring might occur as a relay manufacturer defect. I've just personally never seen a weak spring failure. here is my relay response:

I have never in my life seen a relay "stick" from the pole being magnetized. If that is a problem with a relay, it is a problem that would be there from the minute a relay is used because they would have to use the wrong iron in the relay. The only place I have seen this mentioned is on a west coast amplifier website and now here. That idea flies in the face of any reasonable relay design, and in the ten's of thousands of amplifiers and hundreds of thousands of relays used in antenna switches has never once occurred, not one single time.

The largest single problem with amplifier relays is caused by dry switching. This is where a relay switches with virtually no contact current. Large relays suitable for transmitter power have too large a contact for zero current contact loads, like the receive mode where no current passes through the relay contacts. The lack of current lets the relay contact build up a very light non-conductive film, often just a few molecules thick. This has been a problem since the days the first relays were used. It is by far the single most common amplifier relay issue, and is also an issue in antenna switches.

Over the years I've looked at dozens of ways to run a wetting current, but none have been satisfactory for many reasons.

So if your problem is receiver dropout when going from TX to RX, it absolutely is not anything "sticking" or the unheard of in practice claim the relay pole magnetizes. It is almost 100% certainly caused by the contacts being large and running at zero current.

There are two cures for this. The one I recommend is using a 12 volt supply of 2 amperes or more and connecting it through a 10 ohm power resistor or 12 volt one to three ampere light bulb to the exciter (radio) input connector of the amplifier. Ground the center pin (or return to power supply) the center pin of the antenna connector. Then turn the amp on and cycle the relay a few dozen times. This will completely burn off any film on the contacts without damaging plating.

The second cure, if your amp has an open frame relay, is to wet a piece of solid glossy paper with cleaner and draw it between the closed contacts. DO NOT soak the relay ever. DO NOT ever file or burnish the contacts. The contacts have gold flash on them just a mil or less thick at the thickest points, and you will remove the gold flash rather quickly. Once that is done the next incident will be faster.

Later amps have an enclosed relay system, so the only choice is switching under small currents of 1 amp or so for 20-50 cycles to clean contacts.

A second less common failure is contact welding. Contact welding can occur from a tube arc.

None of this has anything to do with component or relay cost, because $100 relays have the same problem and the larger the relay the worse the problem. The only relays that do NOT have this dry contact issue are mercury wetted relays and vacuum relays. It is an age old problem that any good tech should understand and be familiar with, because everyone from telephone companies to stop light control manufacturers has had this problem with low current contacts. The larger the relay, the worse the problem.


Roland, I've kept up with service on amplifiers since 1982 or 83 with the largest data base ever on service history. The lack of receive return is a very common, as failures go, on every large relay used to dry switch at zero or nearly zero current. It's a problem with every single high current relay used, with the exception of mercury wetted or vacuum relays. It's been a known issue since at least the early 1900's.

Your bias return issue with the relay is rare, but not entirely unheard of. It is generally caused by a tube arc that damages the relay contact, unless someone actually ruined the relay mechanically. By now, your amplifier's relay is long past being ruined. :-)

Be sure you sequence the contacts in the SB220 to avoid blowing up band switches. See this page down near the bottom on mechanically ganged relay contacts:


http://www.w8ji.com/relay.htm

73 Tom

AH6RR
04-25-2011, 06:28 PM
Tom,
I have a question on the relay timing on your web page.

"1.) Reference the pictorial above, use needle nose pliers to bend the unused normally OPEN contact at the amplifier tank circuit connection UPWARDS. Do this by positioning one jaw on the top of the bakelite frame and the other under the outer tip of "A" and squeezing until "A" is just slightly bent upwards towards the moving contact. DO THIS ONLY ON THE OUTPUT CONTACT! This will force the TANK CIRCUIT of the amplifier to connect to the antenna BEFORE the input circuit connects. This change prevents arcs in tank components caused by improper relay timing."

This I understand it makes perfect sense to me.

"2.) Locate the unused normally closed contact for bias switching. Using the same technique above, bend "B" (the normally closed contact) upwards until B no longer contacts the moving contact. This change puts 1/3 more contact pressure on the amplifier bypass contacts, and decreases receiver dropout when the amplifier is on bypass."

Now this one has me puzzled the SB-220 uses both the contacts on the bias section of the relay if I am not mistaken. When in the RX position the 120VDC goes to the center tap of the filament transformer. In TX the relay connects the center tap of the filament transformer to the B- supply and turns the tubes on correct? So the question I have is does the 120VDC need to connect to the center tap in the RX position?

I do understand that the relays do not become magnetized I started working with them in the 70's when I worked with the Texas Highway Department installing traffic lights and the Oilfeild building refineries cotrol rooms (1000's of relays) in the 70's and 80's. I do not know the full history of this amp as it was from a SK. I do know that at one point it had a filament transformer meltdown it is one of the reasons I striped it down to bare chassis to clean the mess it also has some some what new Eimac tubes (Bright red lettering) the 40M contact was roasted on the tank band switch that I replaced. Like I said earlier the relay was pitted in the TX Bias contact. So at one time there must have been a tube failure that caused those. And the problem I was having was the poor contact on Bias contacts and the amp has been working fine since I filled the pit in the relay contact. But I will replace it with a new relay to be on the safe side.

Thanks Tom your input is most welcome here as the wealth of knowledge you have on amps is far more than I will will ever learn in 3 lifetimes.

73
Roland AH6RR

W8JI
04-25-2011, 07:57 PM
Now this one has me puzzled the SB-220 uses both the contacts on the bias section of the relay if I am not mistaken. When in the RX position the 120VDC goes to the center tap of the filament transformer. In TX the relay connects the center tap of the filament transformer to the B- supply and turns the tubes on correct? So the question I have is does the 120VDC need to connect to the center tap in the RX position?

The amp doesn't even need 120 volts bias. You can just open the filament CT. The connection to terminal 3 on the relay can be eliminated. There is no logical reason in the world to force the filament to 120 volts positive.

As a matter of fact while you are at it, you might want to just use a 12 volt relay and add a little 12 VDC power supply for the relay. Relays will be more plentiful, and the amp will be safer to use, and it will not affect performance in the least.



I do not know the full history of this amp as it was from a SK. I do know that at one point it had a filament transformer meltdown it is one of the reasons I striped it down to bare chassis to clean the mess it also has some some what new Eimac tubes (Bright red lettering) the 40M contact was roasted on the tank band switch that I replaced. Like I said earlier the relay was pitted in the TX Bias contact. So at one time there must have been a tube failure that caused those. And the problem I was having was the poor contact on Bias contacts and the amp has been working fine since I filled the pit in the relay contact. But I will replace it with a new relay to be on the safe side.

I'd rip the 120V system out and add a 12V system. Then all you need do is open the CT.

read this page:

http://www.w8ji.com/bias.htm

73 Tom

KM1H
04-25-2011, 08:01 PM
The "weak spring" comment I made also includes the contact arms which no longer return to the NC position with the same force as new.

Ive replaced dozens of SB-220 family relays that were just "lazy" especially in VOX SSB and CW operation. One thing I dont do in my repair service is pad the bill with unecessary parts. Nor do I like returns because I tried to take the fast way out.

This is based on doing amplifier repairs for almost 50 years.

Carl

W8JI
04-25-2011, 09:22 PM
The "weak spring" comment I made also includes the contact arms which no longer return to the NC position with the same force as new.

Ive replaced dozens of SB-220 family relays that were just "lazy" especially in VOX SSB and CW operation. One thing I dont do in my repair service is pad the bill with unecessary parts. Nor do I like returns because I tried to take the fast way out.

This is based on doing amplifier repairs for almost 50 years.

Carl

I've never seen that in any of the amps or antenna relays returned for service or production failures, and loss of tension in beryllium contact carrier bars is certain not ever an issue unless the bar has been seriously overheated. That might be the case when Moon Dog 007 runs 3 kW of 27 MHz AM on a KA11 size relay, but it never happens with properly sized relays in normal service.

The same is true for loss of spring pressure.

Ameritron sold about 1000-1500 relays a month for the past 25-30 years, and there isn't a single case of spring or contact bar weakening. In order it is loss of contact conductivity on low current switching systems, much less it is arc-damaged contacts, and very rarely (usually lightning) open coils. Virtually all problems are caused by either lack of wetting current, and much rarer by tube flash-overs that arc and weld or pit the relay contacts. An even more distant third is open relay coils. Weakened springs and weaken metals and magnetized poles don't even make any service list.

Check these links out:


http://www.stabilant.com/appnt31h.htm

http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3236.pdf

http://ieeexplore.ieee.org/Xplore/login.jsp?url=http%3A%2F%2Fieeexplore.ieee.org%2Fi el1%2F33%2F5742%2F00219407.pdf%3Farnumber%3D219407&authDecision=-203

and the list goes on....

On the other hand loss of tension in beryllium is virtually unheard of unless the metal is grossly overheated or over-bent.

Search this phrase:

low current relay contact failure

There's a great deal to read.

73 Tom

KM1H
04-25-2011, 11:47 PM
I've never seen that in any of the amps or antenna relays returned for service or production failures, and loss of tension in beryllium contact carrier bars is certain not ever an issue unless the bar has been seriously overheated. That might be the case when Moon Dog 007 runs 3 kW of 27 MHz AM on a KA11 size relay, but it never happens with properly sized relays in normal service.


I dont see the Ameritron name on the SB-220 and Im sure you mostly saw warranty problems plus they appear to have better quality relays. I have seen 220 arms that were discolored from heat and no 10-4 type ever used them. Maybe just some elderly ham that took a long time to tune up into the wrong antenna. Maybe even a close lightning hit. The SB-220 relays are pushing 42 years old with the first run.



Ameritron sold about 1000-1500 relays a month for the past 25-30 years, and there isn't a single case of spring or contact bar weakening.


Thats an awful lot of replacement relays unless its sales to homebrewers.

Ive also replaced many open coil SB-200 and 220 relays, I wonder if a gradually corroding coil wire connection wasnt responsible for the "lazy" relay action.

Im well aware of the low current problem, its very common with SS circuitry. Some designs utilize a speed up circuit as I do with my GaAsfet preamps on UHF and above. Im also building it into my new Beverage controller using DIP relays.

K1OF
04-26-2011, 05:48 PM
I suggest you confirm that you have correct filament voltage after the unit warms up. The SB-220 filament transformers are known to develop opens or high R connections. My problem was found to be a weak solder joint inside the filament transformer, where the solid enamel wire is soldered to the stranded flying lead. I was able to open the transformer and simply resolder the connection and it fixed everything. Removing all the laminations is not fun, but from the work you describe you probably can do it. I had to immerse my rebuilt transformer in polyurethane and then bake it to keep it quite after the laminations were reinstalled. A simple filament voltage check hot/cold will confirm if you have this problem.

AH6RR
04-27-2011, 12:22 AM
The Filament transformer is a Peter Dalh replacment and the voltage is the same cold and warm that was one of the first checks I did even though the filaments were bright. Since I filled the pit on the Bias contact on the relay the amp has worked fine with no problems and many contacts in SSB, CW and RTTY thanks to the great propagation the last few days. I belive it was the relay bias contact and maybe a cold solder joint in the bias circuit maybe (they all looked good to me but I re did them anyway).
Thanks for the tip though.
73
Roland AH6RR

ad: Nifty