MC2105 Rebuild
- Thread starter Motnick
- Start date
I have been looking over the resistors installed on my output boards and comparing my resistors to the schematics. I have noticed something strange here and I think I should get some input as I am not sure if this is an issue or not. I have been using the later MC2100 schematic as it seems to reflect my actual boards better, and also because I have not located a copy of the late model MC2105 manual so far.
Here is the MC2100 Output board layout. The resistors have different numbers than the MC2105, but I think the layout is the same between the two units.
The schematic says that R28 & R30 should be 1K ohm, and that R20 & R22 should be 470 ohm 5%.
On my boards the 470 ohm resistors are installed into the R28 & R30 locations, while the 1K ohm resistors are installed into the R20 & R22 locations.
I checked the forum and found pictures of other output boards that match the layout of mine, but I also found photos of output boards that match up with the schematic. Is it a problem that these resistors are installed reverse of the schematic?
Here is the MC2100 Output board layout. The resistors have different numbers than the MC2105, but I think the layout is the same between the two units.
The schematic says that R28 & R30 should be 1K ohm, and that R20 & R22 should be 470 ohm 5%.
On my boards the 470 ohm resistors are installed into the R28 & R30 locations, while the 1K ohm resistors are installed into the R20 & R22 locations.
I checked the forum and found pictures of other output boards that match the layout of mine, but I also found photos of output boards that match up with the schematic. Is it a problem that these resistors are installed reverse of the schematic?
Last edited:
My head it's trobbing over this RT909 Reel that is haunting me so I really do not want to pull out any schematics but would suggest if both your boards are the same and you have seen others with the same layout to just accept that they made a number of changes to the circuits over the production run.
That’s what I thought, but I didn’t want to pass it by without at least asking here first. I’ll keep moving forward with my parts order. Thanks!
I have replaced the carbon resistors on both boards with new metal films. I found almost all of the original resistors had drifted out of spec. I have also installed new Nichicon axial capacitors in both boards. I have replaced both bias diodes with two 1N4148 diodes soldered together. Now either output board will work in either channel without making the DBT stay bright.
I spoke with McIntosh labs late last week and they let me order new bias diodes 070-046. So those should be on their way to me soon.
I was thinking of matching up two sets of KSA992FBU transistors to replace the differential pairs in both boards. Should I go ahead and do this next?
I also have some new Central Semi MPSA14 and MPSA64 to use as replacements for the Darlington monitor transistors. Do these need to be matched in any way?
I spoke with McIntosh labs late last week and they let me order new bias diodes 070-046. So those should be on their way to me soon.
I was thinking of matching up two sets of KSA992FBU transistors to replace the differential pairs in both boards. Should I go ahead and do this next?
I also have some new Central Semi MPSA14 and MPSA64 to use as replacements for the Darlington monitor transistors. Do these need to be matched in any way?
Last edited:
I believe xraytony found that a different series pair of diodes offered a better voltage drop that using the 4148s. I have found very few of the originals not performing within spec even at 40+ years of aging.
Without curve tracing equipment your best test for the sentry monitor devices is to observe that the amp clips symmetrically....assuming the main filter caps are still both strong asymmetrical clipping will indicate non equal current limiting.
Without curve tracing equipment your best test for the sentry monitor devices is to observe that the amp clips symmetrically....assuming the main filter caps are still both strong asymmetrical clipping will indicate non equal current limiting.
Gregory
Soundco Kid
There are at least four schematics of the MC2100 here, and the earlier ones where bipolar transistors are used for Q7-Q10 (Sentry) have different resistors than the schematics showing Darlington devices. Have been thinking it over for weeks and concluded that if one updates to the latest Darlingtons, they would also update resistors R19-R22, R23-R26 and R27-R30 (or whatever) accordingly, see schematics.
One guy here @Hunted11 didn't update to the Darlingtons, and I assume kept the old resistors in place? That's what I'm thinking of doing with @jbailey930's 2100, since the PO (driver board) transistors all tested ok. That being cold, out of circuit tests.
-Greg
Last edited:
I received the replacement bias diodes from McIntosh labs today. I’m glad that they had some for me! I’ve also matched up some KSA992FBU transistors to replace the differential pair 132-056. I’m planning on installing them next, and then moving onto the service bulletin updates. Now that I have the bias diodes, I think I have everything on hand for the updates except for the heat sinks. I still need to order those.
Last edited:
MACKIE1975
Super Member
Hope that carpet doesn't kill the devices
It’s actually a soft towel on my bench, but point taken.
I plan to implement the output board service updates next. I need to order the heat sinks still, and I also realized that I still need 4 disc capacitors to replace C115-C118. The service bulletin calls for the a 680pf disc cap replacement for these. Would these capacitors from Newark be ok to use for this?
https://www.newark.com/vishay/d681k20y5ph63j5r/ceramic-disc-plate-capacitor-d/dp/81R2250
https://www.newark.com/vishay/d681k20y5ph63j5r/ceramic-disc-plate-capacitor-d/dp/81R2250
I now have the service updates completed on both output boards. I have also installed the new bias diodes that McIntosh sent me. The amp seems to be working well now.
I think I have come to the conclusion that the oscillation issues that I struggled with earlier in this thread were actually being caused by a poor fitment between the "bad" board and the edge connector of the repaired output channel. The "bad" output board is thinner than the "good" output board, and also has a subtle twist to the PCB. Whenever the thin board was installed into the thick boards edge connector it caused the DBT to stay bright. This just happened to be the output section with the MJ15003 outputs. This situation was very consistent at first, but after I cleaned the edge connectors and terminals on both boards the issue became much better. As long as I make sure the output board is seated properly when I snug down the standoff screw to the heatsink, then either board can work in either channel now.
The thinner board still seems like it could fit tighter in the edge connector. I have cleaned the boards with a pink eraser and the connectors with D5 and that helped things out a bunch, but I still think the fit should be tighter. Is it possible to tighten up the edge connectors so that they grip the output boards tighter?
I think I have come to the conclusion that the oscillation issues that I struggled with earlier in this thread were actually being caused by a poor fitment between the "bad" board and the edge connector of the repaired output channel. The "bad" output board is thinner than the "good" output board, and also has a subtle twist to the PCB. Whenever the thin board was installed into the thick boards edge connector it caused the DBT to stay bright. This just happened to be the output section with the MJ15003 outputs. This situation was very consistent at first, but after I cleaned the edge connectors and terminals on both boards the issue became much better. As long as I make sure the output board is seated properly when I snug down the standoff screw to the heatsink, then either board can work in either channel now.
The thinner board still seems like it could fit tighter in the edge connector. I have cleaned the boards with a pink eraser and the connectors with D5 and that helped things out a bunch, but I still think the fit should be tighter. Is it possible to tighten up the edge connectors so that they grip the output boards tighter?
You can check with micrometer caliper how much thinner is one board. If indeed board is thinner, you can put one or two layers of isolation band tape on board at the opposite side of connector. It will give you some thickness, at least for test purpose until you find something else like laminating foil or something like that.
This is what I mean:
This is what I mean:
Is there any possibility that there is a bad solder joint on the "thin" board that connects in one edge-connector and not in the other? I ask because I had the exact failure mode, worked in one connector and not in the other, thought it was the edge connector until I went over all of the original solder joints and it worked fine.
This also was a 2105, when I put it in the "good" connector it worked fine, in the "bad" connector I could put some pressure on it and it worked, .... I don't remember what was different or how I suspected solder joints but that was the cause.
This also was a 2105, when I put it in the "good" connector it worked fine, in the "bad" connector I could put some pressure on it and it worked, .... I don't remember what was different or how I suspected solder joints but that was the cause.
I found my metric micrometer and measured the boards. The thick board measures 1.52mm, and the thin board 1.32mm. So they are quite a bit different.
With all of the work done to both output boards I think that I have been on almost every solder pad already, but I’m going to go over the remaining few that are still original just to be sure.
I found an old thread where someone with an MA6100 mentioned that he used some small needle nose pliers to pinch the edge connectors one at a time to get them to bite the boards better. Anyone else done this? Does the MA6100 use these same style connectors as the MC2105?
With all of the work done to both output boards I think that I have been on almost every solder pad already, but I’m going to go over the remaining few that are still original just to be sure.
I found an old thread where someone with an MA6100 mentioned that he used some small needle nose pliers to pinch the edge connectors one at a time to get them to bite the boards better. Anyone else done this? Does the MA6100 use these same style connectors as the MC2105?
Last edited:
jbailey930
AK Member
I have only listened to it using test speakers so no real critical listening so far. I still need to replace the multi cap and also go through the input board. I’ve been moving very slowly with this unit. Taking my time to think each step through before moving ahead. I have plenty of other gear so no real need to rush. I’m sure it will be well worth the wait when it’s all done.
I have an MX113 waiting under the bench to go through next.
jbailey930
AK Member
Sounds like your are on same steady pace we have been. FWIW, @Gregory helped with recap and re-stuff of my MX-113. See this post of our work https://audiokarma.org/forums/index.php?threads/mx-113-to-dos.844273/page-2
The unit became dead quiet after this update.
The unit became dead quiet after this update.
I’ve read your mx113 thread a couple of times already. My unit seems to work but it has a hum. Hopefully new caps will clear that up for me like it did for you guys. I want to finish up the 2105 before I open the mx113.
I’ve been moving along the same path with the 2105 as you have with the 2100 output boards. My unit must be a little newer as I have the later monitor transistors in mine. At C_dk’s suggestion I matched up some KSA992FBU transistors and used them to replace the differential pairs. That brought the offset down around 5-6mv in both channels.
I went back over both driver boards and reflowed the last few solder pads that hadn’t been touched during all of my parts updates. Now every solder pad on both boards has been reflowed and inspected.
I used a small screwdriver to lightly pry behind each clip of the edge connectors of both channels. This has given the connectors more bite on the driver boards. Before doing this the driver boards would sag to one side or the other without the standoffs installed. The driver boards were very easy to pull in and out of the edge connectors before. Now the boards can stand up straight when installed even without the standoffs. I’m 99% certain that poor fitment between the edge connectors and the driver boards was causing all of the issues with the DBT staying bright before.
I also went ahead and installed all new MJ15003G transistors in both heat sinks. So now both channels match. At this point I think the output section and driver boards are done.
I used a small screwdriver to lightly pry behind each clip of the edge connectors of both channels. This has given the connectors more bite on the driver boards. Before doing this the driver boards would sag to one side or the other without the standoffs installed. The driver boards were very easy to pull in and out of the edge connectors before. Now the boards can stand up straight when installed even without the standoffs. I’m 99% certain that poor fitment between the edge connectors and the driver boards was causing all of the issues with the DBT staying bright before.
I also went ahead and installed all new MJ15003G transistors in both heat sinks. So now both channels match. At this point I think the output section and driver boards are done.
Last edited: