Hello all and Happy Thanksgiving to you and yours. Whereas I've been reading the DIY Forum for years, this is my first posting. I have gained so much knowledge and have had endless fun learning from you gays/gals. So Thank you! OK, on to my reason for posting. About a decade ago I purchased two (2) Model 16/A's amplifiers from an estate sale. Neither one of them worked, but with the assistance herein and pure tenacity I was very successful in getting one of them perfect, the other one not so much. The problem stated in the simplest terms I can come up with is the Bias and null (offset) mirror each other for both channel A and channel B. I can adjust VR1 (null) and VR2 (bias) to get bias at specified 50mv, but null tracks that in lock step. I can set bias to zero and null is zero. Note that when bias is sitting at 50mv, the heat sinks never get that warm (as does my functioning amp). So, bias=null and visa versa no matter what I do. I'm here to seek any and all guidance available.
Now that the problem has been stated, it's appropriate to provide some details as to my work. My work includes replacing the PS main filter caps with previously unused mid-90's dated coded units, Hexfred diode installed, driver board completely gone through and rebuilt (Wima, Panasonic, blah, blah) Note the 16A driver boards have the discrete can, not the opamp at the input stage. Gold RCA's installed, all output trannies removed, cleaned, new mica and compound, all emitter R's replaced (and a couple random resistors that look like they've suffered enough heat cycles, VR1 (1meg) and VR2 (500) replaced with multi turn units, faceplate removed and clear anodized just to clean up the physical look as they were pretty beat up, A few bad LED's replaced in the V, Bias and null set to factory specs.
So as stated previously, one amp is 100% and I swap it in and out of my system when the mood strikes me. It sounds wonderful driving my Martin Logans. I'm guessing someone is thinking why don't I trouble shoot swapping parts between the two. Well, I sorta did but only to satisfy one aspect. I did take one of the driver boards from the good amp (channel A to be specific) and put it into Channel A of the bad amp. The symptom remained, the reading were different as I suspected, but bias and null were again mirroring each other. That tells me my driver boards are probably not the problem and the issue is global. Channel A and B react the same. Last point, and it's interesting because I can't identify it. The Relay PCB in the good amp is assy#00132979 and this assembly is correct according to the manual. However. the Relay PCB in the bad amp is assy#00134489A and it looks totally different. I can't figure what this is, where it came form or otherwise. I was clearly soldered on by someone way before it landed on my workbench. So is that my global issue? I wish I knew.
I'm a novice, trying to save a neat old beast. That beast has a relay board I can't identify and its bias/null mirror each other. Lastly it runs! I hooked up some garage sale bookshelf speakers and it sounds fine. But obvious lots of reason to fear a flame on. I prefer to fix it, not burn it to the ground. Perhaps someone could point me in the right direction. Thank you in advance. Stay safe.
Now that the problem has been stated, it's appropriate to provide some details as to my work. My work includes replacing the PS main filter caps with previously unused mid-90's dated coded units, Hexfred diode installed, driver board completely gone through and rebuilt (Wima, Panasonic, blah, blah) Note the 16A driver boards have the discrete can, not the opamp at the input stage. Gold RCA's installed, all output trannies removed, cleaned, new mica and compound, all emitter R's replaced (and a couple random resistors that look like they've suffered enough heat cycles, VR1 (1meg) and VR2 (500) replaced with multi turn units, faceplate removed and clear anodized just to clean up the physical look as they were pretty beat up, A few bad LED's replaced in the V, Bias and null set to factory specs.
So as stated previously, one amp is 100% and I swap it in and out of my system when the mood strikes me. It sounds wonderful driving my Martin Logans. I'm guessing someone is thinking why don't I trouble shoot swapping parts between the two. Well, I sorta did but only to satisfy one aspect. I did take one of the driver boards from the good amp (channel A to be specific) and put it into Channel A of the bad amp. The symptom remained, the reading were different as I suspected, but bias and null were again mirroring each other. That tells me my driver boards are probably not the problem and the issue is global. Channel A and B react the same. Last point, and it's interesting because I can't identify it. The Relay PCB in the good amp is assy#00132979 and this assembly is correct according to the manual. However. the Relay PCB in the bad amp is assy#00134489A and it looks totally different. I can't figure what this is, where it came form or otherwise. I was clearly soldered on by someone way before it landed on my workbench. So is that my global issue? I wish I knew.
I'm a novice, trying to save a neat old beast. That beast has a relay board I can't identify and its bias/null mirror each other. Lastly it runs! I hooked up some garage sale bookshelf speakers and it sounds fine. But obvious lots of reason to fear a flame on. I prefer to fix it, not burn it to the ground. Perhaps someone could point me in the right direction. Thank you in advance. Stay safe.
Thank you for the response. Well, not really. I had replaced the cap back there and sorta wondered if a PP was the correct type. I will though, next week after the holiday. Admittedly, I will have to figure out how to measure/verify the choke piece. I will post my findings. Thanks again!
Yes, it idles quietly and passes a signal per my post. I've listened to it play rock through cheap speakers. Although I hesitate to grade what I was listening to based on the quality of speakers...
Drive boards are plug-in, correct? Check all solder joints on the plug-in connectors, male and female.
Craig
Craig
Hi Craig. I think perhaps you may have inquired about the 10VDC supplies as well? Anyway, they are 12V supplies on this puppy. The +12vdc measured 12.4 with .38v ripple (so 3.06% if I did the math right). The negative rail measured 13.4 with .010v ripple (so 0.07%) I replaced the 1000uf filter caps for grins, but it really didn't have much effect on that supply. Regarding the driver PCB/connector and connection points, they were all cleaned. I have not reflowed solder, but agree that's prudent. Frankly I was never in love with the overall shape of all wires associated with that and even back into the guts of the heat sink. I've got some heavy digging to get to the Zoble so probably a good time to follow your suggestion.
The 12VDC supplies are for all of the "housekeeping" circuits, relay, LEDs, etc. The amplifier circuit contained on the drive boards generate their own 10V supplies. Swapping the boards should have proved those supplies are OK. When you attempt the adjustments do you have + or - offset?
Correct, the 10 is derived from the main rail on the driver PCBs (in my case 69vdc mains and very clean) and you are right when I think about it. I swapped channel A PCB from the good into the bad and mirroring of bias/null remained consistent although the figures were different based on vr1/vr2 pot adjustments required for the good amp. Yep. Easy to loose the forest for the trees. Your Q relative to positive or negative offset I'm not sure I can answer. I want to say I took positive to zero and stopped, notes aren't handy at the moment. One oversight in my original post was I did swap A and B channel driver PCB within the bad amp. Nothing changed. Then I swapped channel A from good to bad amp only to note the symptom still had not changed.
Per suggestion from Huggygood, I've got the 1W R's in route for the Zobel and going to dig in back there as see what is see. I don't recall swapping the Rs, only the C's. Any clue on the weird relay board assembly? I have both sets of prints (Model 16 opamp input and Model 16 A discrete input) and nothing matches. Thanks again for your assistance.
Happy Thanksgiving to all assisting me.
Hello, back again after some level of effort. I'm making positive progress I think but I'm coming back to the well here seeking advice. First my efforts:
- r&r relay pcb in effort to get to zobel. I check every component, looks good. Pulled the cover on the relay, pulled there 12v battery back up for my home alarm and tested the coil/relay function, all is well. BTW, I found a picture on the internet of the relay PCB, seems legit, just not documented.
- Dug out the zobel and built another. Say hello to Mr. Googlie Eyes🙂
-
- I looked at the 10 volts supply suggested previously. Measured at r10 and r26, also looked at the Zeners cr5/cr6. Everything checked out. Then I put it all back together and tested. NO Changes, Bias and Null still mirror each other.
- Back to basics once again, I removed Channel A (Left) Heatsink (H/S) and noted a broken wire, Pin 11 on connector P1. This had to be at least somewhat connected or the amp would not have made music, so I must have done it. It gave me enough pause to cause me to completely rewire the H/S with some teflon coated silver wire I had lying around. net result looked like this..
- OK, put all that back together and fired the amp up again. Within a few minutes I note the H/S was coming up to temp normally. I set bias (more on that in a moment) and plugged it into my audio system with a 15" DVC subwoofer as it's load. Sounded fine and let it run for an hour at decent volume. I noted after the hour channel B (R/S) H/S was noticeably cooler than the left. Got my infrared thermometer and looked at all 20 outputs. R/S was 15 degrees on average cooler than the L/S. Ok so,
- I pulled R/S H/S and performed the same JuJu on it., no pic however🙂
- With both H/S rewired, I fired it up again. Progress it seems. Both H/S now come up to temp equally.
- Upon looking at the Bias, Left side now has a delta between Bias and Null (50.2 and 43.7 respectively) but Right side still mirrors each other. NOTE: The Bias pots (new multi turn) has little to no effect. You can see it vary, only a few MV's, but quick settles back to previous reading. The Null pot has dramatic effect and it's what I've been using heretofore for adjustment. With that in mind a chunk of the schematic follows showing circuitry in and around the Bias piece. I will follow up with some measurements and ask for input.
- R31 measured 1.12v on one side and 97mv on the other. The schematic shows 0.90v on one side and if Im interpreting correctly a -0.25V on the other. Base on Q18 is obviously the same 97mv (positive).
- R32 measured -1.70V on one side and -0.69 on the other side.
- Q15 collector measured 1.69V, Q17 collector measured -1.70V
- So with that in mind, R31 off? New pots screwy? If I were to experiment with R31 (970ohm) I could only do so with 2 470's in series which would put me over spec by 30 or so ohms. Enough to worry about and therefore stand down?
- Im going to close out here. Thanks for all previously responded and to those forthcoming. Remember, Im just a novice attempting to fix this beast and it would seem some of my steps have has a positive outcome. I do appreciated your assistance.
With your DMM connected to the collector of Q15 and the collector of Q17 what happens when you try to adjust VR2?
Craig
Craig
Thank you Craig. It will adjust. Sitting @ 3.3xx vdc and I can go either way with it. Sorta rules out the pot I reckon doesn't it? I'm learning!
Since you are learning the lower this voltage the colder the bias and vice versa. So now we know the voltage on the VAS, Voltage Amplifier Stage, is working let's go to the next stage. The predriver stage is next in line, Q21 and Q22, check the voltage on the emitters or Pins 14 and 5, measure just like last time, does it change with VR2?
Thank you. Yes, smooth linear increase/decrease starting out at .85x ( turn on) and slowly increased to about 1.080v over several minutes and more or less settled in. At that point I adjusted on either side of that reading.
P.S. I did pull both input boards and checked everything one more time. Every R is to spec just to close out a previous comment I had made regarding R31.
P.S. I did pull both input boards and checked everything one more time. Every R is to spec just to close out a previous comment I had made regarding R31.
Thanks Craig, not the case..
Across R3 = 0vdv (either end 2.5mv to ground)
Across R4 = 7.39vdc and note their is a diode in series with it (Looks like 1n914 general purpose to me). This mod is on both boards and I remember see them on my other 16/A. I measured across the R only..
Across R3 = 0vdv (either end 2.5mv to ground)
Across R4 = 7.39vdc and note their is a diode in series with it (Looks like 1n914 general purpose to me). This mod is on both boards and I remember see them on my other 16/A. I measured across the R only..