• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

SAE 2200 Rebuild

sabrefencer said:
Driver Board

This board showed evidence of a failure during the 1980 event that took out the outputs. Several components for ChA had burned and were replaced, including the drivers, pre-drivers (limiters) and some diodes and a 120 ohm resistor. Other issues found included:

- One of the Channel A limiter transistors (Q13) was replaced incorrectly during the 1980 service. This transistor, originally would have been a 2N5210 NPN TO92 with a pin configuration of EBC. The transistor found on the board was a BC237 an NPN with a CBE pin configuration. However, this transistor was installed to match the INCORRECT board markings – it was installed CBE when it should have been installed EBC. It is unclear how the amp functioned with this transistor installed backwards.

- The repair to the divers, only replaced the ChA drivers not ChB.

- There was a service bulletin (2200-3) that identified incorrect board markings, and the need to replace resistors R55,56,57,58 which were 470 ohms when the unit was built (and marked this way on driver board) with 330 ohm resistors – this was not done when the unit was serviced.

Before I did anything, I replaced the differential pairs with discrete 2N5210 and 2N5087 (same as were found on the board) that I matched for hFE on an Atlas DCA55. I then re-checked DC offset, which dropped from 38/21mV to 26/18mV (ChA/B).

Then I replaced everything on the driver board. I also corrected the board markings where needed. In place of the “select” resistors (R21, R22) which were 750k I installed 2M trim pots. However the lowest I could get the DC offset with the trimmers was ~12-15mV, good, but still more room for improvement. I then put in 2.2M ohm fixed resistors resulting in ~8 -11mV DC offset readings, which is better. I will probably replace the fixed resistors with 3 or 5M trimmers.
Click to expand...

Those select resistor(s) are for adjusting the quiescent current. Did you measure this across R37(38) or R35(36). Around 1 / 1.1vdc ? Maybe this is way off and effecting it.
 
Last edited:
What are you using for signal generator? Can you test it for THD using your H/P 334? You need 20KHz at low THD to set the bias properly.
 
I'm getting about 0.6 vDC across the 68 ohm resistors. Bias is holding in the 1.1- 1.2 v range, measured across the 62 ohm resistors.

Regarding the signal generator I am using a Heathkit IG-18 modified with a low distortion kit. THD @20kHz is about 0.35%. I would think that different before and after measurements using the same generator would be measuring differences in the amp, not the generator (which is a constant in the equation).
 
You'll need a generator with no more than .02%THD to use the THD method of setting the bias. You'll have to guess using the voltage method. If you bump up the quiescent current in the VAS the overall distortion will come down. Right now you're at approx. 9ma, try upping it to 15ma or so, increase the "SELECT" resistor.
 
I initially set the bias at 1.1v. Then, with the analyzer set to measure distortion, tweaked it to get it as low as practical, which was about 1.2 v. I would assume I am pretty close. It seems that I still need to work on getting the quiescent current to come up.

The factory value of the select resistor was 750k, I put in a 2.2M. The next larger resistor I have is 10M, which seems like it would be to high. I could try the10M or wait until I get some 5M trimmers.
 
You can put a couple of smaller values in series for testing purposes. I usually use a decade resistor find the value. Did the 11-0080A/83A transistors get replaced during the repair?
 
Yes, the 11-0080A were replaced with KSC2073, and the 11-0083 were replaced with KSA940. I had to cross the legs to get the correct pin configuration, as the board is EBC and the new transistors are BCE. I thought about putting the old transistors back, since they test OK. I assume they were factory installed, but there are absolutely no markings on them, one of each pair is green the other black.

I do have a decade box, that goes as high as 1M, so I could put that in series with the 2.2M resistor to see if I can get the quiescent current to come up.
 
If those transistors have a big tab on them they are probably S-1375 and S-1376, stock from SAE.
 
Yes, they do have the big tabs. I am going to put them back in. I don't see any reason not to run with them, they test fine. Right now the select resistors are 2.0M trimmers, I will keep these in place to see if I can dial in the quiescent current with the S-1375/1376 re-installed.
 
Swapping back to the original S1375/1376 transistors did not make any significant change to the quiescent current reading across the 68 ohm resistors. My readings are about 0.58 to 0.6 vDC. Is the spec 1.0 -1.1 volt?
I have not changed the 2.0M trimmers, nor have I added the decade box in series. I wanted to test only the change from the re-installed transistors, which seems to me nil.
Bias is still around 1.1v. I did not test distortion again. Seems like it makes sense to solve the quiescent current issue first.
 
Were the input transistors replaced? They also will have an effect on the Q.C. Unfortunately SAE didn't put that spec in the 2200 SM. Q.C. will affect the output stage bias. Just like the old points ignition system in cars, set dwell (points) first then set the timing.
 
I’ve had issues with setting the QC on more than one amp. After talking to Craig it was felt that all the new transistors were not giving it the gain as originally. So I increased two resistors and by doing so it all worked once again, with using the pots also. I’ve used CENU07 & 57’s to replace those S1375/S1376’s. I noticed that the schematic for the 2300 showed 1vdc across the resistor which makes me think the 2200 would be the same/similar. R27,28,29 & 30 maybe try 200K instead of the 180K. On the amps I did it didn’t take a giant increase to bring it all back online.
 
I replaced everything on the drive board. The input transistors (differential pairs) 11-0044 and 11-0017 were replaced with 2N5210 and 2N5087, which I matched for hFE on an Atlas DCA 55. The next stage of TO-92s (Q5,6,7,8) 11-0154 and 11-0155 were replaced with MPSA06 and MPSA56. The pre-drivers (limiters) Q13,14,15,16 11-0044 and 11-0017 were replaced with 2N5210 and 2N5087, and the drivers were replaced with MJE15032/33. All resistors were replaced and hand matched to the limit of my Fluke meter. All caps were replaced, ceramics were upgraded to silver mica, or multilayer, all diodes are new. See photos in post #16.
 
R27,28,29 & 30 are 150k in my unit - marked on board, that is what was installed, and what I put in. Perhaps I should try 180k?
 
That will raise the Q.C. also, but the SEL resistor should do the same. If you have them try them.
 
I'll try increasing the 150k to 180k then 200k. In the "select" position I've got 2M trimmers.

Can I get a confirmation that the target is for 1.0-1.1v across the 68 ohm resistor?
 
The smaller 31B has 14-15ma in that stage, .9V across 62 Ohms. It would stand to reason that the 2200 needs at least that much.
 
sabrefencer said:
I'll try increasing the 150k to 180k then 200k. In the "select" position I've got 2M trimmers.

Can I get a confirmation that the target is for 1.0-1.1v across the 68 ohm resistor?
Click to expand...
The schematic I have shows 180K, must have made changes along the production run.
 
R59 and R60 which are the bias measuring points are 62 ohms. I am shooting for 1.1 to 1.2 volts across the 62 ohm resistors for bias, which is adjusted by the 500ohm trimpots on the output board. I have not had any issues getting the bias set.

R35 R36, R37, R38 are 68 ohms. I am of the understanding that the quiescent current is measured across these 68 ohm resistors (R35, R36, R37, R38). For the quiescent current I'm looking for about 1.0 to 1.2 volts across the 68 ohm resistors, this is adjusted by the value of the select resistors (R21, R22), and/or R27, R28, R29, R30 (150k or 180k).

It's not hard to get confused, do I have this right?
 
You must log in or register to reply here.
Back
Top Bottom