Using non-matching caps in McIntosh 2125 repair? Or will it destroy amp/burn down house?

Hi all,

Please take 5 minutes to read and let me know if what I am going to attempt poses any physical danger (fire, broken McIntosh amp, etc.)

today is my last day off before I start an 8 week rotation of working 7 days a week from 4am-8pm (the joys of med school) so I am scrambling to replace the caps on my old McIntosh 2125 amp. I didn't consider it until I would hear an occasional crackle from time to time, and also BOTH meters have never worked (although the light up fine). 
I already cleaned all connections with DeOxit, replaced all 6 IC's on the meter PC board, as well as all of the electrolytic caps. Which offered no change. My plan today is to correct a few signal diodes on that same board which have their polarities reversed, according to the service manual at least, but they are reversed in a symmetric fashion so I don't even know if it matters since the culprits are in pairs, never solo or at a point of termination, so overall the phase output shouldnt be changed by their configuration (I'm guessing).

with that said, I'm working on to the Power Supply PC Board today while I still have the time. 
First major issue I discovered when looking closely between the big parallel axial caps were 2 signal diodes that looked obliterated. I have enough silicone signal diodes and zener diodes to do the whole board, as well as the caps, well almost all the caps. I DONT have exact replacements for the 2 big electrolytic caps (C307, C308) that are rated to "85c, 2200MFD-16v". They don't appear to be In bad shape, but I will have to take them off anyways to get to the busted diodes inbetween them. 

So option #1, remove them from the circuit board and test them with a digital multimeter (Klein Tools model MM2000) if that's possibly? I wouldn't know what readings would equate to "bad" though?

option #2: I have 2 Kimber Kaps .22+/-10% @ 600V. I'm assuming the .22 is micro (u) units, because it's not stated on the cap wrapper.....obviously this is pretty far off from the originals, but just so I'm prepared - when I remove/test the originals if I find they are bad will these Kimber Kaps suffice for a few months until I have time to install a proper replacement? Or would I run the risk of them blowing up/starting a fire/breaking my amp?

any input would be greatly apprefiated! I am damned and determined to get these meters working today...


I see several problems here:

- You cannot replace the 2200uF capacitors with 0.22uF capacitors - the amplifier certainly will NOT appreciate a change of value (four orders of magnitude!) to this degree.
- The diodes near these caps (D302, D303) are NOT small-signal types (i.e. 1N4148) . . . they're rectifier types.  If they're 1A sized (about 2-3mm diameter), you can use i.e. a 1N4004.
- If you're referring to the diodes on the outputs of IC101/102 that are connected in parallel . . . it indeed doesn't matter which way which diode is placed, as long as they are placed opposite to each other.  I'd leave well enough alone, especially if the soldering for them looks original.  My paper copy of the service manual has conflicting reference numbers for these diodes - it's completely conceivable that they're merely documentation errors, not a production or service mistake.

In general, keep in mind that when substituting parts . . . it's far, far more important that they be the correct type and value for the circuit, then they be of "higher quality" in an audiophile sense.
You can find the schematic on line. The 2200uF capacitors C307 and C308 are there to provide transient suppression. In that circuit, there is about a 2 second charging time to bring the voltage up to where it can saturate transistor Q301 to energize the relay to pull in the contacts on the output transformer when the unit is turned on or the speaker switch is selected on.

If you put in that 0.22uF cap you will be defeating the transient suppression. The time constant will be almost instantaneous. 

Chances are cap 307 is shorted, putting D302 in parallel with the 5.6VAC source, causing it to burn. Replace the cap only with an electrolytic of the same value.
wow...I just typed for 20 minutes and got and error. So in an interest of time I'll simply post the photos for now and ask if anyone has any advice as to why there is a signal diode bridging the leads of a single capacitor on the back of the circuit board, and also can I use the same 16V Zener diodes I bought for D304-306 on D119? D119 is currently an empty slot, and it doesn't specify a voltage like the others. Finally youll notice all the red circle of the things I'm going to replace, but just for reference the ones that have a red Z means it should be a Zener diode, but currently has a standard signal diode.

let me know what you think or the parts I selected and if you have any feedback on this project in about to get started on.

thanks everyone!!
D119 is a 6.2V zener. Do not put a zener for D301 -- it’s not a zener, its function is to rectify the 5.6VAC to the base of Q301 (smoothed out by C301). The diode across C118 on the solder side is doing nothing that I can see, unless that cap is shorted and someone thought the diode can prevent the 15V B+ from being pulled to ground.
Good eye Gs, I had that pic mislabeled, I just used the same signal diode as I did for D302, 303 (1N4004).  I did however use the 16V Zener at D119, which was entirely vacant before. I'm not sure why the manual specifies 16V zeners for d304-d306, and then for d119 it just says "Zener"

I changed everything marked in the photo and am about to put it all back together now, keep your fingers crossed
So far not so good, left the front end of amp sticking out of the housing about 5" so I could calibrate the meters if they worked. Unfortunately the power guard lit up right away and a white billow of smoke rolled out of one of those 2200uF....nothing looks fried, aside from that cap, and after turning it on 5 minutes later there was no smoke but no sound either. 

Is that cap the culprit or could there be more serve damage? I'm experienced at soldering, so I am sure it had nothing to do with a short.

Functionally, two 2200uF caps (C307/C308) and D302/303 is that of a half-wave voltage-doubler negative supply.  It's fed by the ground-referenced 5.6VAC that runs the bulbs, and produces (about) a -12v output.  To do this, C307 is charged on the positive half of the cycle, when D302 is forward-biased, and D303 is reverse-biased.  As the input voltage descends, D302 becomes reverse-biased, and C307 passes its charge to C308 through D303, which in practice becomes forward-biased near the negative peak of the input AC.  This setup is really inefficient, requires oversized capacitors, and makes very poor use of a transformer's current capability . . . but it's a cute way to get a separate -12v from the lamp winding.  The only real function it serves is to power the "normal/limit" indicator lamps - the relay startup delay is separate, rectified by D301 with R301/303 and C301 forming the time-constant.  There's also a switch formed by Q304, whereby the +15v that runs the meter circuit is turned on after the -12v indicator-lamp supply comes up . . . this is presumably to keep the meters from jumping around on startup.

As far as your issue goes . . . if it's C307 that blew up, this could also be caused by something drawing too way too much current on the little -12v supply . . . but I'm not seeing any likely "standard" failure modes that would cause this.  Generally, electrolytics blowing up usually means that either it or its associated rectifier diode(s) were installed backwards.

From your photos, it looks like you have the caps in correctly, but it's not clear which way the (blown-up) diodes were installed.  Also, there's a misprint in the manual for your version, pertaining to the switch for the meter +15v . . . only D306 is fed from the little -12v supply, while R309 and the emitter of Q304 are actually fed from the output of the 15v regulator (emitter Q302).

The first thing I would do would be to leave the blown-up parts in place, and run the amp without the meter card installed . . . and check to see if the +15v and -15v regulators come up and work correctly.  Then check/replace the diodes/caps for the little -12v supply, and make sure it comes up properly . . . then check that Q304 is working correctly by verifying the same voltage exists on the collector and emitter, and that there's about 0.6v between its emitter and base.

After the power-supply is running, then remove the ICs from their sockets in the meter card, and put the meter card in place.  Verify that the power-supply is still working correctly, and that there's indeed +15v and -15v getting to the right places on the meter card.  Then . . . install the ICs, and again, re-verify the supply voltages.

It's important to go through the diagnostic process step by step . . . and NOT to replace parts in various different parts of the amplifier until you solve the basic issue.  Otherwise, you can't separate whether you're looking for something that's the root cause of the original problem, or a possible mistake when changing parts.
Thank you so much Kirkus for taking the time to explain that. Very helpful, and I'm sure people will continue to use that info for years to come when googling for help on old McIntosh meters!

if it makes any difference here is some other info:

-All lights work EXCEPT the meters. They were brightly illuminated prior to my "repair"

-The powergaurd red light comes on for the right channel if I turn the meter knob to "watts" or "Hold" (and its brighter when turned to the "hold" position. The light is not illuminated on "0db, -10db, -20db"

-No sound from either channel except an occasional crackle or whisper like it's trying to play but the signal isn't making it..., at all levels of gain, one speak off the other on, etc. etc, for all combinations of volume and channels.

-one positive note, the METERS MOVE! Although they never hit above -20watts, they were atleast flicking a tiny bit and were synchronized. It didn't seem to effect them when i adjusted the db/wattage dials on the meter board though.

does this new information help narrow down the source further? I pulled both boards and there is no physical evidence of a short, even on the capacitor I thought was smoking. Which it only did for 5 seconds the first time I started it up, never since.

thanks again for your help! If pics would help let me know and I'll snap a few of whatever you need.

Just discovered it was D303 that was smoking....not the capacitor. D302 looks fine...what's the deal??

There is a tiny drop of solder on C307 that happens to be right above a pin hole on the circuit board, which on the other side (back) has solder over the hole. Does this mean anything? Should that C307 be tilted a little to make a connection??
I also just found a few of these transistors (2N3904-A06) seen here:

are these usuable able anywhere on the Power Supply PC board? 
Post removed 
Okay, I re-did everything. All new caps and diodes, even swapped out that D119 that was missing originally for a 6.2V Zener and removed that mystery diode on the back of the board as suggested.

After replacing everything except transistors and the C307/C308 caps I triple checked polarities plugged only the power supply board back in (leaving out the meter board)...turned on the power (preamp disconnected), and this time all the lights were brightly illuminated, no power guard warning, then 2 seconds later a plume of smoke....yep same thing. BOTH D302/303 cracked in half

either way I know I need to replace C307 and308 (recommendations??) but in addition to that, could the batch of rectifier diodes I’m using be bad? This is what I am using for d302/303:

pleaseeeee help I just want my music back
Hello, sorry I've been away . . .
I was under the impression that you'd replaced the two 2200uF caps per gs5556's advice, but in skimming through again I'm not certain now.  BTW your choice of diodes is fine, provided they're not counterfeit or anything weird.

If you haven't replaced the two 2200uF caps yet, that's the first step.
I'm dying to know how this ends?  Kirkus, hockey4496, gs5556 you guys are awesome!  Next time I blow something up I'll make sure you're all on speed dial.