Workflow suggestions for refurbishing MC2125

Moving ahead, here's the parts list for the rebuild of both channels...though I will probably fix the left one first and then evaluate the situation:

Pre-Drivers (Power Output PCB)

#132-148 x 2, Mouser #610-CEN-U57
#132-149 x 2, Mouser #610-CEN-U07

Note: These are Centronics U57 and U07, as per other threads on the topic. Dunno if there's yet a better replacement.


Your amp has 35 to 40 volt rails , you could get by with MPSU-55 and MPSU-05 devices which are way more cheaper and still easy to find. < and they look original in all aspects >
I have them and MPSU-56 and MPSU-06 all NOS gold leads and all I believe these were a much closer part then the CEN parts which are all that most folks can find from Central Semi nowadays.
I have the CEN parts also, they work fine In ZAPCO/ARC car amps which for the most part is there only other most popular usage I know of....

So you might want to look around for the MPSU-55's and -05's they should work with your rail voltages being as low as it is in this amp. The big differences in these MPSU devices in voltage ratings. The lower the number set the lower the breakdown voltage is... Might save you some trouble and bucks...

Both Tek scopes you can borrow are well suited for your needs, either will do well... I think I have a link for a set of test tones you can download and use off your computer as a signal source, let me look around for the link, I will PM it to you so no linking rules are broken...
 
I already have a very nice set of audio test tools that run on my iPhone made by Studio Six. Includes a tone generator so I'm covered there. Thanks.

I'll need to get resistors for my dummy load ordered today or tomorrow.

Oh, and the fellow who can loan me a scope also offered use of a variac. Sweet!
 
Your amp has 35 to 40 volt rails , you could get by with MPSU-55 and MPSU-05 devices which are way more cheaper and still easy to find. < and they look original in all aspects > I have them and MPSU-56 and MPSU-06 all NOS gold leads and all I believe these were a much closer part then the CEN parts which are all that most folks can find from Central Semi nowadays.

Thanks for the recommendation.

The only source I can find for the MPSU parts are on eBay where everyone claims to be selling NOS (and they all seem to be from New Jersey for some reason). The parts run anywhere from $6 to $9 a piece. Since I already have two new pairs of the CEN parts (bought for another project) I will probably just go with them. And they were only $2.87 each.

I am curious if you think the rectifier I selected is OK. The only references I found for replacing the original Mc part (#070-031) were from a few years back and these are now themselves obsolete.
 
Thanks for the recommendation.

The only source I can find for the MPSU parts are on eBay where everyone claims to be selling NOS (and they all seem to be from New Jersey for some reason). The parts run anywhere from $6 to $9 a piece. Since I already have two new pairs of the CEN parts (bought for another project) I will probably just go with them. And they were only $2.87 each.

I am curious if you think the rectifier I selected is OK. The only references I found for replacing the original Mc part (#070-031) were from a few years back and these are now themselves obsolete.



Oh I did not mean to say the MPSU parts were cheap, just original looking in all aspects , you know for the folks that gotta have it looking like 1974 again at all costs LOL..

What other parts numbers did you have for it? I can always check my "way back" parts supplier in San Jose they got a whole wall of antique diodes and I do mean a 8 foot high shelved 30 foot long wall of diodes.... Even mil-spec diodes also...PM me a part number and I will look it up and see if I can get you covered with something closer to original, like original LOL...
 
Does that include the color of new electrolytics? That would leave MUSE bipolars out. :D

I see a business here making little orange and tan sheaths out of Bakelite that slip over a MUSE cap and make it look like an original ROE. We can call 'em FAUX caps!
 
Does that include the color of new electrolytics? That would leave MUSE bipolars out. :D


Oh never meant new caps either, Your stuck with what ever CDE and Nippon Chemi-con want you to have there ,,,, Unlesss you save your old caps clear plastic liners and carefully remove them and slide them over the newer caps.

The CDE caps I am installing in a Mc2105 here on my bench right now could do this presto chango plastic sleeve slip on since the newer CDE caps are the exact same size as the older General Instruments caps that were original in this amp. Only noticeable difference is the CDE caps weight practically nothing compared to the shorted out and leaking General Instruments caps I pulled out. Should save about 3 to 4 pounds of lift weight by upgrading to newer caps. CDE are Cornell Dubilier brand name, made in the USA, so hope they hold up as well or better then the General Instruments caps did...

As for internal caps where no one can usually see well the others here seem to have those swaps covered pretty well, I defer to their experience in the matter.. They seem to take no flak on weather or not they match or not ..LOL...LOL...LOL...

Although I am sure with some inventiveness and free time and effort someone will find a way to make everything look like 1974 in time....:lmao:
 
I have recapped ten (I think) 2120, 2125, 2200, 2205 and MC300 amps plus two C34V's using high quality Nichicon and Elna caps. The only electrolytic caps I worried about matching exactly were those used for timing such as the speaker relay or muting circuits but typically staying in the 0% to +50% capacitance range overall. I worry about solder workmanship as much as component selection and treat it almost like an art form.
 
Just wanted to kick this thread back alive and let you know what's up next.

Today I constructed my dummy load using parts from da interwebz. Tomorrow I pick up my loaner oscilloscope. Thursday I'll be opening up the MC2125, hooking all of this up (using the iOS app AudioTools for signal generation) and looking at some actual traces. I'll take snapshots and post for opinions.

It's getting exciting!

Michael

IMG_2249_sm_dummyload.jpg
 
There is nothing wrong with your TV set

So below are some snapshots of what I've been able to look at so far. Please bear with me as I am having to learn how to use the 'scope as I go along. Suggestions on what to measure next and how to make the measurements would be most appreciated.

I'm particularly interested in how to interpret the difference traces.

TEST CONDITIONS

Signal source: AudioTools for iPhone signal generator, 1000 Hz
Load: 200 Watt/8 Ohm resistors x 2 mounted on large heat sink
Gains: Full CW
Probe Coupling: DC

TEST POINTS

Same as for voltage drop study, including the difference between B++/B-- emitter resistors. Ground connected to barrier strip

NOTES

Heat sinks on amp got quite warm. Kill-A-Watt meter indicated power draw was above 100 Watts.

IMG_2250_Right_B-_sm.jpg


IMG_2251_Right_B+_sm.jpg


IMG_2251_Right_B+_sm.jpg


IMG_2253_Left_B+_sm.jpg


IMG_2254_Right_DiffPair2_sm.jpg


IMG_2255_Left_DiffPair2_sm.jpg
 
Last edited:
I'm posting two additional images for the difference trace which provides a bit more detail. Also, I changed the scope to read AC instead of DC, which after I thought about it more, seems like the right way to make this measurement.

IMG_2254_Left_DiffPair2AC_sm.jpg


IMG_2254_Right_DiffPair2AC_sm.jpg
 
Current status of work: I've driven the amp to just below clipping using a 1000 Hz signal into my 8 ohm dummy load. Here's what I'm seeing:

RIGHT CHANNEL clips just below 36V(peak), 25V(rms). Calculated output is 78 Watts
LEFT CHANNEL clips just below 37V(peak), 26V(rms). Calculated output 84 Watts

Something isn't right (or I am measuring something incorrectly) since these numbers aren't even remotely close to rated power nor the rail voltages, which at idle measure +40/-40. However, this does correlate somewhat with an observation of idle power draw using the Kill-A-Watt meter: the draw is roughly half of what the SM indicates:

27 watts measured vs. 50 watts spec
.23 amps measured vs. .4 amps spec

Thoughts on what might be causing this or ideas on what to measure next?

ADDENDUM: measurements made with and without PG enabled to same result.
 
In the interest of continuing to document what I'm doing I want to report that today I remeasured the rail voltages, which as I reported a month ago were a bit north of 40V/-40V at idle. These number haven't changed. However...

As I drive the amp toward clipping (with a 1000 Hz input into 8 ohms), the rail voltages start to drop and at clipping, I'm only supplying about 37/-37V (or 26V(RMS)) And of course, that correlates to what I'm seeing at the output across the dummy load resistors, roughly 36V(peak) or 25V(RMS).

So: is this kind of drop normal? I don't think the amp would be able to deliver close rated power (120 watts) with the rail voltages this low.

Michael
 
Remember - the 2125 has power guard while the 2105 does not. Have you compared the two with the power guard "out" on the 2125?
I read somewhere (possibly the McIntosh "For the Love of Music" book) that the reason that this 2120/2125/2200/2205 had a power guard switch to disable it was that the 2105 generally sounded "louder" when compared to the 2125 with power guard engaged. I have a 2100 and a 2120 powering very similar speakers (AR-3a's on the 2120 and AR 10pi's on the 2100). The 2100 always sounds "louder" although I have not used a db meter to measure. On the bench, the 2100 puts out more power than my 2120 when power guard is active. If I turn the power guard off, they measure pretty close in terms of power.


I am a little bit curious about this issue. I think if you have the same power with or without power guard, the amp without power guard will sound louder, because you are listening to high distortion.
It is the same like some friends of me, are always telling me, that there tube amps with 20 Watt Power Output are so loud like a 100 Watt SS amp. I am always telling them, that they are listening to high distortion and that this is a subjective effect. Is this right? :scratch2:

Best regards

Martin
 
I'm just looking at voltages and current draw at this point. The only difference PG makes is what the waveform looks like when the amp is driven to clipping.
 
Replacement for thermistor 144-074 or 144-075

Would anyone happen to know what to use as a replacement for the original thermistors, Mc part #144-074 or #144-075? These are the ones inside the heat sinks. The originals have a shrink covering with the lettering "500" on them. They are thermally "connected" to the sink with some heat sink compound.

Thanks!

Michael
 
Not considered a semiconductor by McIntosh, so no parts cross. I always considered them to be a mechanical device, heatsink overheats and it opens the relay, cools and restores AC.
 
Still scratching my head...

Thought it would be time for a few substantive updates on this project:

First, I had a something of an epiphany regarding voltages I was measuring across my dummy load. If you recall, it seemed my amp was woefully under performing. I spent time thinking about how my digital multi-meter took its readings. I realized whether or not it was a "True RMS" meter, it needed to do *some* kind of sample averaging on the input voltage. I took a peek at the manual and found this spec for AC voltage:

Response: average, calibrated in rms of a sine wave

So while my DMM is not a "True RMS" meter, it treats sampled input voltage as if it were a sine wave and then applies a simple RMS conversion to it. And in this case, the input voltage is a sine wave, the 1 KHz test tone, the power of which is unvarying over time. In other words, the readings my DMM was giving me were accurate w/o any further conversion. I was incorrectly converting the numbers a second time, leading me to the low power output figures I reported earlier.

Thus taking my observed voltages as is, I get the following output in Watts for each channel:

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 resistance of my dummy load

That's a huge relief!

Second, I'm still scratching my head regarding emitter resistor voltage drops and bias setting. I have at this point completely rebuilt my left channel output section replacing the transistors, resistors, and diodes. I have also replaced both of the pre-driver BJTs on the Power Output PCB as well as replacing the single turn trimmer with a 25-turn part. I am still however observing a very significant difference between voltages on the final two emitter resistors: the one on the positive rail hovers around 5 mV while the one on the negative rail reads close to 0 mV. I would expect to see this one read close to -5 mV. I still don't understand why this is the case or whether I should even be worrying about it. Certainly open to suggestions.

A few other changes I've made:

  • Replaced all the opamps on the Meter PCB and calibrated meters under load.
  • Replaced the coupling caps on the RCA inputs, changing value from 1uF to 2.2uF as per later revision of the amp. Used a MUSE bipolar for this.
  • Replaced the .47uF coupling cap on the Power Output PCB with a 1uF as per some discussions I read about improving bass. Used a MUSE bipolar for this.
 
Back
Top Bottom