SB-220 Find

I have the stuff to make it, and plan on doing so.....
I have some of that same material, I can cut it with a box cutter and a ruler.
I can't cut straight with scissors. I used a box cutter, file and a pin vice to make the top connector for the plate choke.

I have thought about pushing the fan blade further back, it was a tight fit on the new motor shaft.

The amplifier tests okay after the new choke installation.
I did notice I have an input VSWR issue on ten meters. Power is down on ten as well.
I'll address that the next time I open up Oscar.

 

Since wensday I have been rebuilding the power supply.
Rebuilt rectifier / metering board.
Replaced filter caps.
Replaced high voltage wiring in power supply section to match the PA compartment.
And last but not least...
Added a soft start circuit.
So I have been rather busy.
The largest issue was where to put the soft start, I wanted to mount it in such a way as to make sure it was solidly mounted and wanted to keep it close to the line cord input and the voltage selector terminal strip. This proved to be easily accomplished albeit a bit labor intensive.
First the capacitor bank was removed, it was going anyway...

The aria was cleaned and a pattern was made of the board to be used as a placement and drilling guide.

The pattern was located in order to allow pan head screws to be used in the small dead space between the plastic cap mounting block and the chassis.
Next the holes were drilled and the board was wired up.

In spite of how it looks the board is straight to within less than a thirty second of an inch....
Now the rectifier / metering board was removed, stripped and rebuilt using nearly all new parts.
The plate current shunt had to be reused due to Mouser sending me the wrong part.

caps were installed and wired up.

And here is the finished rectifier / meter board installed..


I did run into one snag, I had swapped two of the black wires going to the soft start board, due to that the amplifier failed it's first power up, (on the variac)
All is good now, I have a little higher average power and the lights don't wink at all when the amplifier is turned on from a cold start.

Stay tuned more to follow....

 

Sue,

Thanks for starting and continually updating this thread. I just read it all and I feel I have learned a lot. Also it is a very good reference for anyone wanting to tackle an SB220 rebuild.

73,Paul
VE1PEW

 

Thank You Paul;
It's not over yet, I still have more to do.
My purpose in posting my progress on the SB-220 is to show others what is involved and what problems can come up.
This forum provides a wonderful place to show what is involved in various tasks in repairing and modernizing older gear.
Pat WA6MHZ and others have posted some good reiteration threads for gear over the years and I have always found them educational. While I didn't go into this project blind, I had worked on an SB-220 in the past, I picked up some new knowledge from others who have more experience, so it has been very profitable for me in terms of expanding my knowledge of the SB-220. In a month I will be gathering the last of the resources together for a QSK mod, after that the old plastic covered wire will be replaced, and that will most likely finish the project for the year.
Glad you enjoyed the thread.

 

In another thread Jim N2EY and I were discussing filter caps and what advantages there are by increasing the size of said caps. Instead of reinventing the wheel and going through all the messy math, I am a bit rusty in math anyway, I'll walk those who may be interested though the process that allowed me to come to the conclusion the 390MF caps I chose were better than sticking with the original value caps.
All the power supply stuff used in rebuilding the power supply....


As many of you know the full wave voltage doubler circuit is somewhat nonlinear in it's behavior the caps being the responsible party, in the right set of conditions a large change in regulation can be had by doubling the cap values. What is large, I say anything over 20% in voltage output.
Some off the napkin figures I started with...
3000V no load plate voltage.
4825 ohms load resistance, that is at .7A, not too unreasonable....
The original total capacitance value 25MF
I increased it to just under 50MF, (48MF) but we will use 50MF, besides the caps may total up to very close to that value given tolerances.
Then I used this chart from my 1988 ARRL Handbook on page 6-12.


The improvement is close to 25%, better than my original eyeballing the chart indicated.
The advantage is lower IMD because of the tighter regulation. It was important to install the soft start at tie time I rebuilt the power supply, the next most important thing will be adding glitch resistors, I am thinking a total of 25-30 ohms should be enough.

 

Thanks for the update. But I see a few issues.....

The first one is that IIRC the original discussion was about the need for filter caps with very low ESR. I don't see any calculations justifying them for this application.

The second is that Rs (the source impedance) is pretty much unknown. In the SB-220, Rs is a combination of transformer, diode, and line impedance. Note how there are different curves for different ratios of Rs to R.

The third is how bigger caps affect the diodes (which is probably a non-issue, since they are being replaced with modern ones with much better ratings.

The fourth is how the bigger caps affect the power transformer and crest factor.

Fifth and last is how much difference the change makes in real life SSB service. That is best determined by measurement.

None of the above may turn out to be a big deal, because the increase is not dramatic (it's not as if you replaced a 25 uF cap string with a 250 uF cap string) and you added soft-start.

In the SR-2000, there was a specific reason not to increase the filter cap size too much - a flashover in an 8122 could destroy a tube if the filter caps are too big. That doesn't seem to be a concern with 3-500Zs.

More pictures!!

 

I did forget about the ESR issue.
I have no idea what the original ESR rating was for the old caps I could measure the ones I have, but that would be "stacking the deck" since they are thirty eight years old.
The ESR issue is perhaps not as important in terms of any one issue with the power supply except for heating of the capacitors, I think we agree that ESR is a multi-edged sward.
On the one hand it ensures better regulation, on the other hand it increases stress on the transformer and diodes. I wouldn't have gone to larger caps without a step start circuit. I also have plans on replacing the plate transformer with one from Hammond Engineering when they have the Peter Dahl replacement line available.

Sorry I forgot to include that number which is 12.5 ohms, that is the secondary winding resistance of the plate transformer. That number is actually available on the net as well as my own confirmation subtracting the test lead resistance on my DMM.

It is hard to find the original specs for the diodes used in the power supply, no 1N number is given. If I had to make a wild guess I would suspect they are 2-2.5A diodes, they were replaced with 1N5408 diodes.

The inrush current will be higher, so will the peak current during overload conditions and to a lesser extent during normal operation. As I understand crest factor there would be no change since the waveform into the voltage doubler circuit remains essentially unchanged. If I am wrong on this and you have the knowledge, please enlighten me.

I agree that measurement is needed to see what changes have taken in terms of real life operation.
Not having a spectrum analyzer up to the task, I could only guess. The spectrum analyzer in my IFR 100S is a joke, but then again so is the spectrum analyzer in most every other service monitor.

I know about the situation with the SR-2000, those tubes are fragile and like the tube used in a Signal One CX-7A they are very difficult to find.

More pictures to follow as Oscar's rehabilitation proceeds.

 

I started to watch this thread at the beginning , but drop out early because I do not have one .
Sue , was there a problem with the transformer ?
I was trying to figure what the cost of reconditioning a 220 , but this sounds more link a major up grade ?

But now a local ham wants to get out from under one he bought at a SK estate , we both bought stuff and I took down all the antenna stuff .
With out seeing he is offering it up for $250 , a blown resistor , should be some room there for a number directions to go , recondition , upgrade or 6m amp .
Dang now I'm going to have to read starting back at page one
After getting into all this stuff , and what the other guy bought , it would seem that almost everything was in pore to DOA condition , after taking down the antennas , it would seem that they were in bad enough condition to cause all kinds of problems .
Well used .

 

There is really nothing wrong with the plate transformer in the SB-220.
They work well just don't work them too hard.
I want a replacement simply because by the time I get done I will have an almost new amplifier, so why not take the extra step and get a transformer of a more modern design. The amp will be good for another 38 years then.

 

I am quite sure that even when brand new the ESR of the original caps was higher than modern ones.

Ok, although the original iron may be OK for the application. You can probably sell the original iron for a good price too.

The actual Rs is higher than the DC resistance of the secondary. There are all kinds of factors involved, such as leakage reactance, the resistance of the primary, core loss, etc. What Rs really represents is all the factors which limit the current available to the rectifier.

As I suspected, the replacements are higher rated. Back when the 220 was designed, the cost of silicon diodes was a serious issue; today they are perhaps the least expensive major component in the supply.

Crest factor is the ratio of peak to average current. Using bigger C and lower-resistance diodes will make it worse because the conduction angle will be less. Whether that's an issue is unknown.

With a capacitor filter, the diodes only conduct when the transformer output voltage is greater than the capacitor voltage. As the caps get bigger and the output voltage approaches the theoretical 2.828 times RMS, the shorter the conduction angle. So all the current has to get into the cap during less time.

An easy test is to see how much the HV changes under load.

Which is one big reason the SR-2000 and amps like the NCL-2000 didn't achieve much in popularity.

We had an NCL-2000 at the U when I was there. Nice amp - it had the whole time-delay setup, too. But it was complicated to tune up, and cost about double what an SB-220 cost.

Great!