Workflow suggestions for refurbishing MC2125

Today I went through and measured all of the resistors forward of the pre-driver transistors on both Power Output PCBs---just to eliminate them as a cause for biasing problems---and discovered something rather odd. Diodes D215, D216, D217 and D218 (070-047) have each had a 100K resistor (not shown in the schematic) added in parallel to it---the resistor has actually taken the place of the diode in the PCB and diode rides piggyback on it. In later revisions of the PCB (045-144 etc) this change would have been one of several made to this section of the circuit. However, on my PCB this is the ONLY change that's been made.

I'm thinking that either McIntosh issued a service bulletin that called only for this one change or someone went in and tried to upgrade the board (from 044-626) without making any of the other changes. I'd appreciate any insights or opinions that folks might have about the matter. Thanks!



P.S. For comparison, below is a link to photo of a later revision of the board where you can see the 100K resistor formally added to the circuit:

http://home.comcast.net/~mc_audio/MC2205_Love_RIGHT_installed.jpg

 

I can not and have not been able to keep track of your meanderings.....but have you checked the differential pair? I have replaced them with two matched high HFE transistors with exellent results.

 

I have not yet spent any time looking into the differential amp, Q205/206, part #132-155. So far no one has suggested it as a potential cause of my problems with bias adjustment.

If my postings seem meandering I am sorry. I try to report what I am observing and doing in the interest of documenting as I go along. Since I only have limited time every week during which I can work on the amp, my progress is rather slow. It's also taken some time to obtain all of the required tools (e.g. an oscilloscope), which I now have. And there's also the matter of learning what to do, how to do it, and how to interpret it. For much of that that I am rather dependent on the good graces of forum members like yourself.

I am still very interested to find out what you might know about the change to those diodes connected to the Current Sense transistors on the Power Output PCBs. Since I have found no schematic which corresponds to this one isolated change it has caught my attention.

Thanks!

 

On my 2125 I replaced the dual transistor differential pair on each driver board with NS LM394 "super match" devices. They are electrical and physical drop in replacements. I can't say they made any difference but they are as good as it gets for that purpose. I got them on ebay since they are no longer made.

 

Just checked: all the sellers right now are from China. Reliable?

 

Is there notch distortion? Can you see it on your scope or can you measure it with your distortion analyser? What does ithe distortion do as you test it at various frequencies? How linear is the amps output at clipping for various frequencies? Lots of things to test to determine why you semm to believe you have low bass output. Under full power at the point of clipping how much do the supply rails sag at 20 hz verses 2000 , 20,

I had a tough time reading though your supply voltage readings. I seem to rember that the supply sagged a lot at high power, did you test it at 20 hz?

And yes I do believe you might be fussing over a negative feedback/ dc offset issue. Echowars posted an interesting test using very high HFE zetex devices years ago on some project he was working on.

Remember also this design is not based on a darlington pair but a Szikai or compound design. Wiki Sziklai pair for an explanation of the differences. Rod Elliot's analisis at Soundwest is always educational.

I actually talked to Sid Coderman about setting the bias on these amps and the Mc2255 family years ago.......and yes the setting process seems a bit sloppy for the MC2205 family. That conversation went downhill quickly.

 

I don't know. I bought two from someone in China 6-7 years ago but have no idea who. Mine work fine according to my below average hearing. All I can say is that no Chinese counterfeiter is going to get rich on this item based on my wild guess of the installed base of everything that might use it from 30-40 years ago. Ironically, I was never able to find out what the "original" part was that Mac probably used or the parts National Semi claimed to be improving on. Mac only seems to have used them in the 2120-2200 family with common driver boards. The 2150's use discrete transistor pairs.

 

Thanks for the considered reply. Much appreciated!

So far I've only done one study @ 1000 Hz and did not observe a notch with the amp set to output about 2W into my dummy load. I can make a plan to repeat this at a variety of frequencies next week. What would be a reasonable set of frequency steps to test at?

Yes, there's a sag of about 3V when 1000 Hz clips, however the amp is still outputting plenty of power. Turns out I was unnecessarily converting observed voltages into RMS (for a sine wave, multiply by .707) but the meter was already reporting RMS values. Here they are again:

RIGHT CHANNEL: 36V(RMS) @ clipping into an 8.9 ohm load [*], 145 Watts

LEFT CHANNEL: 37V(RMS) @ clipping into an 8.9 ohm load [*], 153 Watts


[*] - actual measured value of the resistors in the dummy load.

 

I'm going to hold off messing with this part of the circuit for now as I have no evidence that anything is wrong. Meantime, I can keep an eye on eBay offerings and see if someone in the US offers the part for sale. And I suppose I could ask make an inquiry with the folks in Binghampton.

 

20hz will show you if the power supply will keep up, might sag your wall supply so a variac is usually used as part of the testing.

 

Perfect. Thanks. (And I have a variac as well now too.)

 

You have to check at 20k also.

 

Roger!

 

Sorry but I keep thinking about that diode/resistor mod on my Power Output PCBs. On the original schematic the symbol indicates this was supposed to be a Zener diode but in the parts list it uses the same part number other normal diodes (070-047). [*] A Zener would definitely change the way current flows into the bases of those Current Sense transistors---and strapping a resistor across a diode of any sort would also change things again. So I am still left wondering whether this was a formal Mc mod or the result of some techs attempt to remedy a problem with this section of the circuit.


[*] Later versions of the schematic show it as a regular diode (070-047, again) but with other modifications including the resistor in parallel.

EDIT: Actually, I just checked the SM for the MC2205 and it seems that the Power Output PCB has that exact mod, including the change from a Zener to a regular diode though the part # for the PCB is identical. It even leaves the resistor that sits between the diode/resistor pair and ground with the same value: 47K. In later revs the value of this resistor was changed to 56K.

 

If it's a regular silicon diode then it is forcing the voltage across the resistor to be 0.7 volts so it is forcing a certain current through the resistor depending on its value. V=I/R so I=V*R Pretty sure that's correct.

 

Thanks, @jlovda.

Finding that there was an in-line revision made to the PCB in the 2205 that matches what I see on the version installed in my 2125 satisfies me that I'm not looking at the inventive handiwork of a former owner or tech.

 

A little set back...

So...

Measured clipping voltages on the left channel across a range of frequencies (100, 1000, 10,000, and 20,000 Hz) and it seemed pretty linear, 39-40V.

Started working on the right channel and then my scope lost its magic smoke. POOF! I opened it up and looked inside. Nothing was scorched on the logic boards. I think it was the PS but as I have no idea how to service such a thing, I'll leave it up to a tech to determine what needs to be replaced. Also: scary HV warnings on the cover. I've got that in motion but I'll be w/o the scope for a while. :cry:

Also still unable to see a notch at low power, 2V. I fiddled with scope settings quite a but I can't see anything but a clean sine wave. Any suggestions?

Meantime, I've got the parts in stock to rebuild the right channel output stage so will do that tomorrow.

 

132-155

Was ordering a few parts from Mc and decided to proactively purchase replacements for the differential pair (Q205/206) on the Power Output PCB, part #132-155, rather than roll the dice purchasing parts on eBay from China. The part number for this hasn't changed interestingly enough. Below are photos of what they sent. No obvious manufacturer or external part number (one is screened 8020 and the other 8003 -- more like batch #s) though they are screened "155" on the side and there's a dot of red paint on the top making me think they've possibly been graded.

 

I'll bet they were made in 1980.

 

That seems reasonable. Amazing they'd have back stock for so long but also easy imagine they ordered them by the boat load.