Workflow suggestions for refurbishing MC2125

Have meant to do this but have not......I am curious how close the HFE is matched with these dual devices.
 
Have meant to do this but have not......I am curious how close the HFE is matched with these dual devices.

I'll measure tomorrow. It will be only as accurate as what my relatively inexpensive DMM is capable of but hopefully...good enough.
 
Well, all things being relative, here's what I measured for hFE:

#8020: 457/455

#8002: 233/230

So the two sides of each part seem pretty well matched with each other but the parts themselves are rather different. Hmm....
 
According to Mac's semiconductor data sheet a LM114, Mac# 132-155 sourced from Intersil had a minimum HFE of 200.
 
Went in today with intention to replace the Current Sense transistors on the left channel. Hooked up DMM to establish baseline readings for voltage drops across emitter resistors in the output section and, color me surprised, but things looked really even, consistently within a couple of tenths of an mV between the + and - sides! The right channel is a different story---not as even---but I as I still haven't rebuilt the output section (transistors, diodes, resistors) that might be why.

It does occur to me that I might have been making a mistake in how I took my readings after rebuilding the left output section, putting ground probe on the output terminal "com" wire instead of the lead from the output section that attaches to the relay. One's on the far side of the autoformer and the other is before it. That might explain the unsatisfactory numbers I reported.

At any rate, the left side looks pretty good right now with bias set a little north of 5/-5 mV. Don't think I'm going to get it any better than that.

Time for a brew, methinks.
 
Double checked, as of the 1990 data sheets LM114 is part #132-155

Thanks for checking. I would assume that that's what they sent me too. There's a very tiny logo on the side of the can and it might possibly be an "i" of some sort. Hard to make out even with magnifier.
 
c_dk said:
The MC2125 and its kin need a structured restoration plan IMHO. If a OMsemi MJE15028 or something similar can reliably replace the MPSu07 for example, the parts cost with replacement caps would be less than $30.00 to remanufacture a driver board. If you are recaping anyway why wouldn't you replace the $12.00 worth of transistors. I think it would be negligent not to.

@c_dk: I found this in a thread from August of 2011 [*]. It seemed at the time you committed to make the replacements ("Times awasting so trial parts and known subs are being ordered.") but you never posted a follow up. (You also had bigger fish to fry in the form of the damaged transformer.) Obviously I have a vested interest in the topic and now I've obtained replacements for those very same transistors (from McIntosh, for the record). I'm wondering if you'd care to comment on that experience and what you might have learned since.


[*] - http://www.audiokarma.org/forums/showthread.php?t=317291
 
Both digikey and mouser have been stocking the Central Semi u07 and u57. I pulled out my trial devices.
 
Both digikey and mouser have been stocking the Central Semi u07 and u57. I pulled out my trial devices.

I too replaced the pre-drivers with new u07/u57 parts. I was specifically curious to know if you chose to replace any of the other transistors, an approach which it seemed you might have been considering at the time.

Since I already obtained replacements for these from McIntosh---as you say, the parts are cheap---I was thinking of installing them in one channel to see what difference, if any, they might make. I can always put the original parts back in if the result is that I convert my amp into an oscillator. ;->
 
So today I decided to rebuild the right channel output section replacing transistors, resistors, diodes, and diodes using the same selection of parts I used for the left channel a few weeks back [see below]. This channel now appears to be sharing load across the rails much more evenly, albeit based for the moment just on idle voltage drop on the final set of emitters. I've set the bias on both channels around 5/-5mV. There's still a little bit of a slow drifting evident: around .5mV up and back over a period of about a minute. I don't know if this is the way the circuit always worked or something that could be corrected with more work.

Also put a fresh dab of thermal paste between the thermistors (#144-074) and the heat sinks.

Oh: and here's a photo of that nasty looking diode from the right channel along with the sandblock resistor which was next to it. I assume those greenish splotches are from the diode. Some water must have gotten into this amp at some point.

IMG_2309_sm.jpg


Parts used for output section rebuild:

PNP, #132-151: ON Semi MJ15025G x 2 (drivers): Mouser #863-MJ15025G
NPN, #132-152: ON Semi MJ15024G x 2 (drivers): Mouser #863-MJ15024G

NPN, #132-164: ON Semi MJ15003G x 4 (outputs): Mouser #863-MJ15003G
PNP, #132-165: ON Semi MJ15004G x 4 (outputs): Mouser #863-MJ15004G

Emitter Resistors, #139-105 (inductive): Ohmite .15 Ohm/5W, 1% x 8: Mouser #588-15FR150E (non-inductive)

Resistor, #144-079: 12 Ohm/.5W x 2: Mouser #603-MFR50SFTE52-12R (metal film)

Diode, #070-031: 1N5394-E3/54, 1.50 amps, 300 Volts x 6: Mouser: #625-1N5394-E3/54
 
Last edited:
@jlovda Thanks for the response. If you read a few posts back you'll see that I have this meter and made a bias adjustment using it. The observed current draw was only half of the spec numbers given in the service manual. I made the adjustment anyway and the emitter numbers reported above are the result of that. My questions remain the same:

Why was the current draw half of what's listed in the SM and is that a problem?

How do I interpret the numbers for the individual emitter resistors and the resistor pairs?

Should both channels be more or less the same? If so, which channel is "right"?

Are the small voltages I measured across the pairs on the left channel indicative of a problem or acceptable?

Michael

my 2125 has half the current draw as yours on the KWHM also ( actually 23W @ 120 V), I only used this meter with a variac for power up after R channel rebuild to check things before setting bias across the emitter Resistors.
I think the lower current draw is because I removed all the 6 meter lamps and used LED replacments because of heat so they draw less current
 
Power Guard discussion again. Well to me it was always a big plus back in the 70's and 80's. When you are using Mac amps for discos and live bands, anything that will protect expensive diaphragms and woofer cones is a big financial plus. Big Altec and JBL drivers could be pushed by a 2120 and 2200 all day and all night with twice the power without failing than you could from a 2105 or 2300. Now if you put a Urei peak limiter in front of the 2105/2300 and set up the 2105/2300 with an oscilloscope to make sure no clipping was taking place you could get almost the same result. The only problem was DJ's and club techs like to twiddle knobs,Then when the owner had to shell out a bunch of money for my time to replace the expensive diaphragms the Power Guard was soon learned to be loved. Because of the digital revolution owners of stereos have learned they need ample head room to produce short peaks without clipping and to buy speakers that can produce the required levels in their homes without stress. Its not like the old days when folks bought speakers that were rated at 35 watts and used Mc 240 amps or 2505 amps that when just pushed would clip and take out their T-35 and T 350 tweeters or blow their Altec 802 diaphragms that could only handle 12 watts rms in the flash of an eye. Most folks under stand they need a safety margin of 10 to 16 db in their systems today. If they do then Power Guard isn't as noticed as it was back in the day when there was either no head room or normally less than 6 db in the systems. All Power Guard has to do now is handle the occasional peak. But when it is asked to work over 30 % of the time then just like any limiter its action can be heard. 207 puts out 250 watts per channel and more depending on how many channels are being driven simultaneously to those levels. In the worse case only 4 of my 14 channels ever seen a power guard lamp flash for more than fraction of a second. The rest of the time I normally operate at 1/2 to 2 watts for loud levels and an occasional 20 to 50 watt peak for HT sound effects. So I really don't need power Guard, but as I value my speakers and replacments aren't easy to find off a shelf, I'm glad its there in the background to protect my speakers. It also services another purpose a limiter can't. If a component fails that will cause more than a fraction of a percent in distorted waveform shape then the Power Guard will react at any level protecting your speakers. Blow an out put transistor or a driver and the power guard will prevent levels much beyond a 1/2 a watt. What happens when your 2105 or 2505 has a minor heart attack? Well if you are not careful you will have lost valuable irreparable speaker components of your classic non repairable loudspeakers. I have had 2105's and 2300's wipe out more Altec, EV and JBL diaphragms and McintoshML series tweeters than I wish to remember. One night a guitar player forgot to cover his strings and their was acoustic feedback with his marshall amps and the sound got picked up by the mics and I lost 8 288C diaphragms in less than a second using a 2300 power amp. So first buy a speaker that can produce the level you wish in your room with ample head room. Then buy an amp that will produce those peaks with out any sign of strain, but so matched to the speakers that the power Guard will turn on above those peak levels. Lets say you buy a pair of JBLs that are rated at 375 watts and will handle 2000 watt peaks for 10 milliseconds. Should I buy a 1.2 K amp, or a 2k amp or a MC 450. Well a 1.2 k amp will put out 2000 watts for 40 milli seconds so it meets the peak requirement. But the spekers RMS rating is only 375 watts, so I'll compromise and by a pair of 601's. Wrong> The speaker can't handle the 900 watt capability of the 601. So How about a MC 450. Well if you reed the fine print at 100 watts give or take a few the woofer is already compressing 1 db and at 375 watts it compressing over 3.5 db. So if the speaker isn't loud engough at 100 watts then its not loud enough. 100 watts equals 116 db and 200 watts 119 maybe for the JBL 4435 with over 300 watts. So I would buy a pair of 301. Their peak rating is close to 450 watts and the Power Guard would probably come on some where between 360 and 400 which perfectly matches the JBL speaker iRMS handling n this case. I know some will say I'm not taking full advantage of the speakers capability. But I am. But he'll say they sound great 1000 watt Crown amps. Fine as I have said in the past we each have distortions we love and have to live with in our systems. If you have the money to back up your listening style fine, more power to you. But I would rather spend the money somewhere else, buying a new lNikon or pre owned Hasselblad lens or staying gone another month in the RV traveling the States and Canada, and maybe buy a new turntable and Dynavector XX2 cartridge. Look at the money I would save between 301's and 1.32 K's. Even though the 1.2 k is my favorite amp in this example give me the 301's. And let the Power Guard flash once or twice a year. If you can hear it its protecting your speakers.. .
 
Last edited:
Another thing to consider is the power rating of our speakers. Some speaker are rated full bandwidth peak levels. But what happens if all the power is focused to the bottom octave of the woofers capability. Can it handle it. Well obviously the JBL couldn't in the above case. And I doubt todays current Mac speakers could either. So unless your into horn speakers and don't ask them to produce below the capability of the LF horn, you really need to do your home work before buying any speaker system. There still is no substitute for the volume displacement of a woofer. So if it takes two 15 inch woofers moving 4/10 of an inch to produce 110 db at 20HZ. How far are they going to have to move to produce 120 db . Its like Magico and B&W saying their S7 and 802 can produce 120 db at 50 HZ with 2 % distortion. What happens at 30 hz let alone 20 hz. And of course you have to ask do I need that capability. Now you loose close to 8 to 10 db if you listen at 10 ft from the speaker. So your system is now limited to 110 db. Then you decide 25 hz instead of 50 so your system loses another 6+ db in capability at those distortion levels if your lucky. Now we are down to 104 db. and may be the speakers share a little bit, but we'll say 104 db as a safety margin. Most of use listen around 85 db and with 6 to 10 db for peaks we are very safe. But then we decide we want to listen to some life like classical Orchestral or Ja zz concert and boost the level to a 90 db average. We we are starting to run out of safety margin fast and if its a Reference recording, Sheffield, or Telarc we are sunk and hope and pray our Power Guard is calibrated properly. We are lucky today we have choices and can reach life like performances in our homes. Money is the limiting factor, Yester- year there wasn't the capability. Life like levels were possible, but with extremely high forms of distortion. I decided to limit my dynamic range in favor of LF extension. You'll have to make choices too. Good luck. We have the Power , the source material and vast selections of performances, all we need to do is select the right loudspeaker system first. Good luck.
 
So today I decided to rebuild the right channel output section replacing transistors, resistors, diodes, and diodes using the same selection of parts I used for the left channel a few weeks back [see below]. This channel now appears to be sharing load across the rails much more evenly, albeit based for the moment just on idle voltage drop on the final set of emitters. I've set the bias on both channels around 5/-5mV. There's still a little bit of a slow drifting evident: around .5mV up and back over a period of about a minute. I don't know if this is the way the circuit always worked or something that could be corrected with more work.

Also put a fresh dab of thermal paste between the thermistors (#144-074) and the heat sinks.

Oh: and here's a photo of that nasty looking diode from the right channel along with the sandblock resistor which was next to it. I assume those greenish splotches are from the diode. Some water must have gotten into this amp at some point.

IMG_2309_sm.jpg


Parts used for output section rebuild:

PNP, #132-151: ON Semi MJ15025G x 2 (drivers): Mouser #863-MJ15025G
NPN, #132-152: ON Semi MJ15024G x 2 (drivers): Mouser #863-MJ15024G

NPN, #132-164: ON Semi MJ15003G x 4 (outputs): Mouser #863-MJ15003G
PNP, #132-165: ON Semi MJ15004G x 4 (outputs): Mouser #863-MJ15004G

Emitter Resistors, #139-105 (inductive): Ohmite .15 Ohm/5W, 1% x 8: Mouser #588-15FR150E (non-inductive)

Resistor, #144-079: 12 Ohm/.5W x 2: Mouser #603-MFR50SFTE52-12R (metal film)

Diode, #070-031: 1N5394-E3/54, 1.50 amps, 300 Volts x 6: Mouser: #625-1N5394-E3/54
 
Hello Manhattenup;

I'm currently trying to rebuild an MC2125 that has problems with the 044-626 board. I have a (supposedly ) working 045-144 board that I would like to replace it with. Obviously, there are differences in these
boards and I've been ohming them out to see those differences. It would be wonderful to get a schematic from McIntosh (I've tried and you need a model number that contains the 045-626 board to get one).
Do you have a schematic for this board? I am new to this group and don't know my way around yet or I would have posted a more general request. It looks like you are very familiar with the 2125. Can you help?
 
I have two versions of the service manual:

For units beginning with serial #AZ1001, Channel Card PCB #044-626

For units beginning with serial #AZ2805, Channel Card PCB #045-144 (units AZ2805 - 5345), and Channel Card PCB #045-557 (units starting AZ5346)

My guess is that the newer cards are electrically compatible with older units. (This was the case when I replaced cards in a 2105 a couple of year ago.) The real problem with going from the earlier to the later cards is mechanical: while the earlier card used an 8-pin connector plus a single post to connect to the inputs, the newer cards used a single 9-pin connector.

What's wrong with the current card that you are choosing to replace it?

Michael
 
Back
Top Bottom