• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

MC2105 Rebuild

Motnick

Motnick

AK Subscriber
Subscriber
I picked up a nice condition MC2105 from the original owner over the past weekend that I would like to update. The amplifier is working but it is in original condition. I have opened up the amplifier and right away I’ve noticed that one capacitor looks like it has started to vent.

I can see that one channel has been repaired in the past. MJ15003 transistors are installed in that channel, and it also looks like the 5w resistors for that channel have been replaced. Some work has been done on the driver board for that channel as well. Fortunately the work looks like it was done properly. The solder side of both driver boards looks nice. I did notice that a mica was installed on all 6 MJ15003 transistors but I don’t think this is correct for the MC2105? Hopefully someone can clarify that.

My plan is to replace the 2 filter caps, replace the multi cap, and recap the input and driver boards. I’d also like to implement the driver board service bulletin. Should I plan on updating all of the outputs to MJ15003? It always bugs me when the two channels don’t match up.

Any guidance or advice would be great. I have been reading all of the MC2100/2105 threads that I can.


index.php

index.php


index.php

index.php

index.php
 
Last edited:
Register to hide this ad
I would be first concerned with the main filter caps because of the discharge as well as the replaced wirewound resistor that can be seen.

They must be inductive ones, typically supplied by Cober or Rockwood. Oscilation is very possible if not ones supplied by Mac parts.

I have to think thru the improper use of all the micas....there should only be one on the negitive heat sink to bring into circuit the odd WW resistor. The heat sinks are supposed to be "hot" with supply voltage or audio.......
 
Are they mica or metal?

Yes, replace PS capacitors, the multi-cap (with a new multi-cap or separate caps), new electrolytic caps throughout, test all of the wirewounds, update the driver boards, and replace all of the output transistors with matching new transistors.
 
Originally only one device was isolated via a coated thus insulated aluminum "mica".

The two pictures posted show one sink with no isolators, which is correct, and one with three clear micas? Incorrect.

I do suspect the WWs due to the white color. Originals we're blue or beige.
 
They definitely look like clear mica insulators. I will post better pictures of all 4 heat sinks tonight.

What I actually have is one channel, two heat sinks, that look untouched. All of the transistors look original, and only one out of six transistors in this channel has a mica. All of the wire wounds look original in this channel.

In the other channel, all six transistors have been replaced with MJ15003 and all six have a mica. It also appears that all of the wire wounds in this channel have been replaced. All replacements are bright white.
 
Last edited:
Channel that looks to be untouched:
index.php
index.php

index.php




Previously repaired channel:
index.php
index.php

index.php
 
Last edited:
I’m thinking step one should be to replace the two main filter caps. I don’t want to power up this amp any more with the leaking filter cap that’s installed.

Step 2 should probably be to figure out what to do with the mica situation. Should I try to remove the micas that don’t belong? Is it possible that they were installed by mistake, and I can simply remove the ones that don’t belong?
 
Last edited:
I'm seeing tiny cylindrical ferrites on yellow wires under the output sockets in OP's photos. Those would be to suppress parasitic oscillation? Normally a choke like that is to suppress RF surface currents from propagating along wires or cables.

What's the requirement for matched outputs or pairs, if any? how tolerant is the circuit for transistor variation?


-Greg
 
Last edited:
I went ahead and order two new capacitors. It looks like others have had success installing these. Hopefully they’ll work out for me as well.

https://www.mouser.com/ProductDetail/Cornell-Dubilier-CDE/DCM413U050DC2B?qs=AcmJtBVhzkTn9Hepf9PP%2bQ==

I am thinking of installing them once they arrive and then running the amp some to evaluate the amplifier more. With the new caps installed I should be able to measure some voltages and also check the temperature of the heat sinks. Let me know what you guys think of this plan?
 
Last edited:
The wire-wound power resistors on the `updated´ side appear to have a date code 9th week of 1998?

2105_WW_date_code.jpg

Does that sound right?


No word on the ferrites??

2105_output_ferrites_725x.jpg

Aren't those put there to suppress higher frequency signals? Chris?


-Greg
 
From your pics it would appear that a repair was made with generic WWs and Motorola 15003s.

The tech did not follow the original build in regards to isolation of the transistors via micas.

The use of ferrite "beads" to control high frequency oscilation can be standard procedure in any electronics.

Amps going into oscilation is hit or miss which is why proper bench testing is always recommended.
 
Gregory said:
The wire-wound power resistors on the `updated´ side appear to have a date code 9th week of 1998?

View attachment 1699419

Does that sound right?
Click to expand...

The original owner told me that he had a repair done to the amp around twenty years ago, but he couldn’t remember what exactly had been done or why. He remembered that the repair was around $60 at the time.

The MJ15003’s have 9752. 52nd week of 1997?
 
Motnick said:
The original owner told me that he had a repair done to the amp around twenty years ago, but he couldn’t remember what exactly had been done or why. He remembered that the repair was around $60 at the time.

The MJ15003’s have 9752. 52nd week of 1997?
Click to expand...

Yes, that would all add up to a repair 20 years ago.


Ferrites attenuate EM fields that propagate along wires and cables.

https://en.wikipedia.org/wiki/Ferrite_(magnet)

"Ferrite cores are... commonly seen as a lump in a computer cable, called a ferrite bead, which helps to prevent high frequency electrical noise (radio frequency interference) from exiting or entering the equipment."

Inductive WWs would act as a choke, and would need to have XL high enough to block ultrasonic and RF energy, but low enough to NOT roll-off the amp's freq response. From what I've heard, parasitic oscillation in these amps is far above 20 kHz. Any comment or clarification on that?


-Greg
 
Last edited:
My discussion with Mac engineering was centered on the MC2125 and others in that family specifcally because my question was in reference to if the circuit mods were because of a change in production's output devices from Motorola sourced to RCA. I was informed why the mods were put in place due to Rockwoods production changes to its WWs.

The inductive WW possible issue was inferred out of that discussion but had not been specifcally addressed because at that time Mac parts still had orginal stock of .56 and .33 WWs and because of the use of 15003 devices we have always had to carefully watch out for oscilation anyway.

If oscilation does happen the correct remedy is to replace the driver TO3 devices for the darlington and Szikali (compound) paired devices with lower gain devices.
 
Last edited:
The new capacitors arrived today and I was able to get them installed with no trouble. I powered up on the DBT and all was well. I moved to full power and measured the rail voltages at 38.0v and -38.0v. The 90v supply measures at 105.4v, and the 80v supply is measuring 90.6v. So everything is measuring higher than the schematic.

I also noticed that the positive 35v rail drops to 0v almost immediately when the unit is powered off, while the -35v rail takes a minute or more to reach 0v. Is this the expected behavior, or should they both fall slowly to 0v together?

index.php
 
If my memory is correct without pulling out a schematic it appears the sentry monitor transistor were changed most likely when the output stage failed years ago.

I would certainly replace the axial caps on the driver boards and input board.

The time difference between the caps discharge could be because of the value drifting higher for the one discharge resistor.

Thinking of drifting resistors. Even if you choose to hold off rebuilding the driver boards do check the bias resistor value. Even a two ohm drift high can greatly effect the notch distortion.
 
I ran the amp and listened to an entire CD at a normal listening volume. Meanwhile I kept an eye on the output heatsinks with a temp gun. All of the heat sinks got up to around 85f equally and seemed to stay at that temperature.

I will definitely be rebuilding the driver boards before putting this unit into any real use though. I measured the bias resistors in circuit and both measured around 26ohm. I’m debating about rebuilding the output boards next, or completing the power supply by replacing the multi cap.

Is the multi cap that hayseed offers any good, or am I better off using separate capacitors? Seems like a nice clean stock looking solution, but are the caps any good? Anybody used the hayseed multi cap?

https://hayseedhamfest.com/collections/mcintosh/products/mcintosh-amp
 
Last edited:
At 26 ohms I would guess you would measure close to 1% distortion if you have access to a bench...Maybe more.

I believe spec is 22 ohms without the update parts installed.
 
You must log in or register to reply here.
Back
Top Bottom