Mac half breeds acceptable here?

[miniman82]

I picked up this ratty MC2100 from a guy in cali a couple months ago, and wasn't really sure what direction I wanted to go with it so it sat for a while till I could make up my mind. Since the electronics had been gutted all I had was a chassis, the transformers, terminals, and a whole lot of rust.

In the end I decided this would be a good home for the pair of ECG-1328's a friend gave me, so I set about rehabbing everything.

I had the chassis sandblasted to get off as much rust and old plating as possible, and it came pretty close to original bare metal with some slight pitting where the exposed metal had corroded. I then had the same place do powder coating in satin black all over everything, which is why I call this amp 'BLACKOUT'. The potted transformers couldn't be coated because the baking might have melted the tar in them, so instead I stripped them by hand and shot them with flat black spray paint.

I used a cut in half heat sink from an old JVC receiver for the output modules, and fabbed the rest from stuff I had laying around. The PSU filters are donors from an MC2120.

It sounds pretty good, aside from a slight lack of high end. This is the first amp I've owned where I actually had to use my C32's EQ to brighten the top end a little on program material. Other than that, I'm happy with how it turned out.

Voltage into an 8-ohm dummy load is 28v, which ohms law pegs right at 100 watts so pretty close to what the original should be.

Thoughts? I say it's better than going to the dump, that's for sure. Not completely Mac anymore, but the important stuff is there.

 

[Mike Gibson]

I bet it sounds as good as it looks. Welcome to the asylum.

 

[c_dk]

Have to say I like the stealthy look, don't think I care for the circuit choice but sometimes you got to do what you can do.....are you driving the autoformers with the powerpaks?

 

[miniman82]

Yes, the autoformers remain in circuit. That's another thing that drove me to doing this, since these STK style amplifiers don't really like low impedance loads. Autoformers solve that problem handily, the amp drove my 4-ohm Infinity QB's effortlessly.

If you don't like monolithic IC's, check out the data sheet on them. These particular ones have a diff front end, same type of quasi-complimentary output setup as original, and the outboard components are mainly power supply related. It's not a whole lot different than what it had, truth be told. I haven't measured distortion yet, but even if it's slightly higher than a Mac circuit I doubt I'll notice.



Here's the data sheet on them:

http://www.datasheet4u.com/datasheet-pdf/ECG/ECG1328/pdf.php?id=546941

 

[Hipocrates]

It will be fun to see the internals! Would you share some pics?

 

[damacman]

Very cool! Got an construction pics?

 

[Pio1980]

Congrats on a nice save.
There was a scruffy mod'ed two channel only MC240 listed a while ago without all the front end stuff that interested me, but for some reason I didn't move on it. Kinda regret that now.

 

[c_dk]

The MC2100s output autoformers load the output stage at 2 ohms. Memory said it was 4 ohms but I checked the schematic in the AKdatabase and it clearly shows 1/2 way between 4 and common so 2 ohms it is.

Those ECGs are probably gasping for breath......

The Mc2100's 4 outputs and 2 drivers have a lot of silicon to disapate heat and handle current, the paks not so much.

Neat idea but it could fail in a spectacular fashion and your speakers might be the fuses.

 

[Pio1980]

The MC2100s output autoformers load the output stage at 2 ohms. Memory said it was 4 ohms but I checked the schematic in the AKdatabase and it clearly shows 1/2 way between 4 and common so 2 ohms it is.

Those ECGs are probably gasping for breath......

The Mc2100's 4 outputs and 2 drivers have a lot of silicon to disapate heat and handle current, the paks not so much.

Neat idea but it could fail in a spectacular fashion and your speakers might be the fuses.

It's have to hit the autoformers pretty hard to do that, and I think the fuses would go first. Foiled fuses should cook and open the emitter resistors before carbonizing the autoformers, speakers I don't see going smokeward.
Still, ideally get more margin in the output stage to be reliable under most cdx. Otherwise, go easy and always fuse properly to protect the autoformers.

 

[miniman82]

It will be fun to see the internals! Would you share some pics?

There aren't any to speak of, just some wiring. It's pretty boring inside now, lol.

 

[Pio1980]

There aren't any to speak of, just some wiring. It's pretty boring inside now, lol.

STK style power packs are pretty much modular power amps with very little support circuitry.

 

[miniman82]

The MC2100s output autoformers load the output stage at 2 ohms, those ECGs are probably gasping for breath......

Maybe. Let's remember that I'm running at less than the recommended rail voltages for these units, so there may be some wiggle room available. Less rail and lower load resistance could make things more or less equal all things considered, what will matter is the total current seen by the output stage. These are rated 7A collector current max, which is currently exactly what they are seeing. Will it make them pop? Only time will tell, I already got 100w out of them and they seem to be cool. I suspect that as long as I don't overheat them or push them hard for long, they'll last a good while. And if not, I'll do something else. There's a 170v winding not being used, perhaps a tube front end will appear in version #2.

The Mc2100's 4 outputs and 2 drivers have a lot of silicon to disapate heat and handle current, the paks not so much.

And yet, MC-2100 outputs fail occasionally as well. There's probably less headroom here before bad things happen, but I don't plan to push this very hard.

Neat idea but it could fail in a spectacular fashion and your speakers might be the fuses.

I read D.C. gets shunted by the autoformers, is that incorrect? (From russels site)

"Should a direct current component appear at the amplifier output, it was shunted by the low DC resistance of the autoformer, instead of passing through the speaker voice coil, which could damage the speaker or even cause a fire."

 

[Hipocrates]

There aren't any to speak of, just some wiring. It's pretty boring inside now, lol.

Oh I see, pretty cool project.

 

[c_dk]

The semiconductors are 45+ years old in most MC2105s and MC2100s, very early designs when high power PNP devices were marginal so not used. Therefore the Szikai (compound) pair for both the original design Was used. The generic EVG is a copy of what I believe was a very early Sanken pak design, therefore it too is a Szikai design.

Sid Coderman and his engineering team could choose any impedence they wanted in designing the autoformer to load the output trasistors, they chose 2 ohms. I do not believe the Sanken engineers expected to see 2 ohms for their design.

When a transitor fails and the magic smoke is released from the amp who knows what complex garbage gets sent to the speaker.

The database has these schematics for the general interest and use of AK posters. Most often they provide needed good information, sometimes not so much.

 

[miniman82]

It might be possible to connect the autoformer differently to present a more acceptable load the the amplifier, but I'll need to run some tests so see if it's possible.

Meantime I think one of the power supply caps is past its use by date, the positive rail is a couple volts lower than the negative and I hear a slight buzz. I'll go replace it and see if that helps matters.

 

[quartersaw]

I picked up this ratty MC2100 from a guy in cali a couple months ago, and wasn't really sure what direction I wanted to go with it so it sat for a while till I could make up my mind. Since the electronics had been gutted all I had was a chassis, the transformers, terminals, and a whole lot of rust.

In the end I decided this would be a good home for the pair of ECG-1328's a friend gave me, so I set about rehabbing everything.

I had the chassis sandblasted to get off as much rust and old plating as possible, and it came pretty close to original bare metal with some slight pitting where the exposed metal had corroded. I then had the same place do powder coating in satin black all over everything, which is why I call this amp 'BLACKOUT'. The potted transformers couldn't be coated because the baking might have melted the tar in them, so instead I stripped them by hand and shot them with flat black spray paint.

I used a cut in half heat sink from an old JVC receiver for the output modules, and fabbed the rest from stuff I had laying around. The PSU filters are donors from an MC2120.

It sounds pretty good, aside from a slight lack of high end. This is the first amp I've owned where I actually had to use my C32's EQ to brighten the top end a little on program material. Other than that, I'm happy with how it turned out.

Voltage into an 8-ohm dummy load is 28v, which ohms law pegs right at 100 watts so pretty close to what the original should be.

Thoughts? I say it's better than going to the dump, that's for sure. Not completely Mac anymore, but the important stuff is there.
View attachment 968754

I have two rusted, "rat" 2100's. It is a great amp. Enjoy!!

 

[c_dk]

DOB amp clinic trick. When you are absolutely deadly postive the caps are fully discharged back off and then retighten the main capacitor screws.....make sure they are discharged because the arc if one isn't will certainly get your attention.

I use a old style Christmas tree bulb with gator clips attached to discharge them.

 

[miniman82]

Turns out the power supply is actually fine, there's no AC ripple coming out of it as observed on the scope. The problem is the cheap coax cable I ran from the input jacks to the boards, noise was getting into them. I'll replace with better quality stuff and retest.