• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

DIY McIntosh Amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Moved from McIntosh MC-3500 Schematic Information

Sorry to bring up an old thread... who is this Milo guy? Full name and bio? Is he still alive? Are you guys saying he's the one who designed the best ever McIntosh amp and then left the company? Where can I read more about this?

I've been reading about these MC transformers for decades. Are they really over the head of a qualified professional electrical engineer? It appears the fellow in Germany was able to make his own?

What was the result of this MC3500 clone build?

Can someone build this amp for me?

Thanks if anyone sees this and replies! Sorry if I said something wrong or against diy protocol.
 
Ya gotta love the way McIntosh brings such a crowd of followers and fanatics. Just the name itself makes people's eyes open wide and brings on the shakes. Yeah, they name some decent equipment through the decades, but it's an amplifier/tuner/preamp, essentially a brand of audio equipment..... that's all.

A name.

I've been in 'em, plenty of times, I got two ratty butchered MX tube preamp/tuners sittin on the floor in the shop...... they're just another piece of audio to service. I don't get thrilled over them.
 
Ya gotta love the way McIntosh brings such a crowd of followers and fanatics. <snip>
My dad has a mc240 and MC-20 and all other 60's tube and analog stuff that he only paid hundreds for as a poor college student, I'd like one but I'm not paying thousands of dollars for stereo equipment... he only paid hundreds for his.

Is there a better amp than the MC3500 that can be DIY made with 350 watts and as good or better sound?

I don't really care about the brand, just the sound. The MC240 with L19 speakers sounds as good as anything I've heard... trying to get any better is just OCD and I'm not OCD. (no braided gold and platinum speaker wires for me)

The question/point I'm trying to make is that if the guy in Germany can wind his own transformers with a tool he bought from a metal scrap-yard for a few dollars, then how hard is it really for a professional electrical engineer to reverse engineer one of these MC amps?
 
Unless you want to fill a stadium with 'tube sound' there is no need for that much power. For the most experienced engineers on this forum, it's a challenge not a need. If you want to build a very high power amp, take a look at Tubelab's work on Pete Millett's engineers amp. He took that 20W design to 100+W safely.

250W from a pair of big sweep tubes was shown to be possible as well, but you'll have to accept two things: the transformers get very expensive. Think hundreds a piece, and that's just one component. Second: a power supply of 600V or more, capable of very high current is needed to produce this power. When something goes wrong, parts don't go 'poof' they go 'bang', not a spark but an explosion. The chances on success go down.

You say you like the MC240's. Why don't you aim for that level of power? The MC design is fairly complex and the transformers labour intensive to make, so expensive still. I'm running a sweep tube amp in my living room and really like it. Plenty of power (ca 50W) from cheap EL36's and standard components.
 
Parafeed813 echos what I have been saying for years. As the power levels go up, the probability for success goes down, especially for someone who has not got years of tube amp building experience.

I have had all the hard parts (tubes and transformers) to build a 400+ WPC tube amp, but have not yet built it, primarily because I have no need for one. I am however coming to realize that I must build it soon before I become to old to lift it!

look at Tubelab's work on Pete Millett's engineers amp. He took that 20W design to 100+W safely.

That's a good place to start. I took Pete's conservative 20 WPC design and cranked it up to 125 WPC. The details are scattered throughout the long and winding thread which could be considered required reading for anyone considering a tube amp of over 100 WPC. About 10 to 15 of the 125 WPC versions were built by different builders. I built two, one went to my boss, and I still have one.....it still runs on it's original tubes, Though it doesn't get too much use any more, I did play my guitar through it for about a year, and used it for sound reinforcement art an outdoor rock concert twice without issue. It ran at the edge of clipping for several hours in the hot Florida sun without any issues, though the somewhat undersized power transformers got rather warm.

Posted new P-P power amp design

My exploits prompted Pete to introduce a 50 watt mono block version of the same design.

50W monoblock “Engineers Amp”
 
Unless you want to fill a stadium with 'tube sound' there is no need for that much power. For the most experienced engineers on this forum, it's a challenge not a need. If you want to build a very high power amp, take a look at Tubelab's work on Pete Millett's engineers amp. He took that 20W design to 100+W safely.

250W from a pair of big sweep tubes was shown to be possible as well, but you'll have to accept two things: the transformers get very expensive. Think hundreds a piece, and that's just one component. Second: a power supply of 600V or more, capable of very high current is needed to produce this power. When something goes wrong, parts don't go 'poof' they go 'bang', not a spark but an explosion. The chances on success go down.

You say you like the MC240's. Why don't you aim for that level of power? The MC design is fairly complex and the transformers labour intensive to make, so expensive still. I'm running a sweep tube amp in my living room and really like it. Plenty of power (ca 50W) from cheap EL36's and standard components.

I got out of "audio" back in the 90's I come back and check in every now and then, but still, not much has changed... my dad's mcintosh stuff is still the best I've ever heard, I wouldn't want to get so obsessed that I want something better, given that it's almost more rare to win the lottery than to find someone with an amp better than a MC240. They are very rare.

In response to your technical info... isn't hundreds less than thousands? Why is MC240 circuit complex? Are you saying it isn't needed? How many of the people on this site are actual electrical engineers with a BS or greater degree? I've been hearing the same words come out of the mouths of different people since the 1990's... almost 30 years ago. Elon Musk created PayPal, Tesla, Solar City and Space X in a fraction of that time, you'd think someone would have figured out how to clone a a MC240 or any of the good MC amps for cheap by now, but I guess you are all still stuck going in circles. That's how it seems anyway, not trying to be rude, it's just an interesting phenomenon.

I'll probably just splash down some cash for two MC240's some day... this may not be the place but is there a more satisfying amp than a MC240? I think going beyond this is just OCD and I try to not be OCD... it's more about the pleasure and enjoyment... and nostalga at this point... we used to bang the crap out of this amp as a kid and it's still running great. It's paired with some JBL L19's in a medium sized room and it's good enough.

I actually don't know much about how amps work, electronics, and tube amps and I read that the MC30 is better because it's fully tube based where the MC240 has some parts that are solid state? Has anyone cloned the MC30 and bumped up the power on that yet?

Is there some place I can read to get the basics of why a tube amp transformer is so different from any electrical transformer? What do all the different types of wire and bifilar/pentafiler change if it's doing the essential basics of stepping up or down voltage/current? Why is a MC240 transformer better than a cheapo transformer? Why the myth and mystique around how they wound them up by hand? Isn't it basic electronics?

And 350 watts isn't enough for a big venue... they must have used many MC3500 amps for different parts of each speaker. Typical large music venues have thousands of watts in addition to a lot of other equipment and timing the signal so it arrives at the same time in different parts of the venues.

Also... if the patents on the vintage Mcintosh transformers and circuits have run out and expired what's stopping a company from making their own MCXXX and selling it?
 
Re: mc240

A number of MAC clones have built over the years by DIYers. It obviously CAN be done, if you have the OT winding info and the patience, or find someone to wind one for you ($$$). Not an easy wind if you do it right. (no random winding!)

However! The MAC OT cross coupled innovation solved the plate to plate coupling issue for P-P center tapped outputs in class B. This has -never- been a problem for AUDIO amplifiers running in class AB (AB so as to avoid crossover dist., a completely different problem). Only an issue for efficient class B operation of shaker table amps and such.

Now the MAC 50% CFB -IS- a nice touch for low distortion.

So we come to the Circlotron OT, 50% CFB and it -never- even had the P to P coupling issue, like center tapped P-P OTs have. The Circlotron OT is FAR EASIER to wind than the MAC OT, with HALF the primary turns, and only standard primary to secondary interleaving. No need for bifilar winding, unless you want to do the small center primary section (1:1 AC voltage portion bifilar coupled to the secondary). The Circlotron OT is also known for being the highest bandwidth OT. OT bandwidth (well beyond audio band) is required for extensive global N Fdbk, like used in the MAC. (although local N Fdbk around just the output stage and driver, excluding the OT, can provide much the same distortion clean up for lower BW OTs)

The 3 Neg. Fdbk windings in the MAC OT only increase leakage L in the OT, plus winding complexity, and should be dispensed with. One could put taps into the Circlotron primary, cap coupled, for the same purpose if preserving the MAC front end design.

Although the main emphasis should simply be on providing for some form of local N Fdbk around the output tube stage and driver. Then limited global N Fdbk around the OT and full amplifier. Various approaches are available.

Using TV Sweep tubes, and especially parallel ones besides, with their high current capability, will bring the number of turns down even further on the OT primary. The lower the primary turns, the higher the OT bandwidth (Allowing more global N Fdbk, like the MAC). While just two TV Sweeps in P-P can provide all the power and fidelity needed for most home applications, and will avoid the complications, risk and expense of high power amplifier construction.

There is also the Norman Crowhurst Twin Coupled design which can achieve much the same as the MAC OT design, using just two off the shelf standard OTs.
 
Last edited:
#1: Yes, the MC240 schematics is rather complex. This is due to the relatively high drive levels that the power pentodes require in Unity Coupling mode and due to it's high open loop gain that allows some considerable amount of NFB. But it isn't as complex by far as the MC3500.
#2: The MC240 is an all-tube amplifier. It's only SS components are the rectifiers in the PSU.
#3: Have a look at the '69 Woodstock PA system for instance. They indeed used many MC3500 or MI350 power amplifiers, along with some SS by Accuphase.
#4: Nowadays we can get the same output power with much less effort, much less power consumption and very much cheaper from class D amplifiers. And both their tech data and sound are impressive as well. The only drawback is their lack of nostalgic reputation. Reissuing a MC3500 amplifier would most probably lead into economical desaster, due to a very limited market it could be aimed at.

And yes, I'd also like to own a pair of MC3500's. But I'm quite sure that this will remain a dream to me.

Best regards!
 
How many of the people on this site are actual electrical engineers with a BS or greater degree?

I started out making amps out of old TV's, radios, and HiFi sets rescued from the county trash dump. This was in the early 60's, so of course they were all tube designs. My amps were usually make do with whatever parts you could find, and by the time I was in high school they were good enough to sell.

I learned electronics on my own, and I walked into a Motorola plant at age 20 looking for a job. I landed on the assembly line, and worked my way up to being a product development engineer. Then Motorola sent me to college on their dime for a degree, and then another degree, so yes, I have a masters degree in electrical engineering. I have the knowledge to clone a McIntosh amp, but no desire to do so.

My amp designs both tube and solid state have usually been about getting the best performance from the least amount of money, but some builders of my designs often throw $$$$$ at them to make them even better. Where do you spend extra money in a tube amp to make the best improvement? Just about any tube amp builder here knows it's the transformers. Why?

why a tube amp transformer is so different from any electrical transformer? What do all the different types of wire and bifilar/pentafiler change if it's doing the essential basics of stepping up or down voltage/current? Why is a MC240 transformer better than a cheapo transformer?

An ordinary "electrical transformer" works at one frequency, 60 Hz in the US, or maybe two frequencies, 50 Hz and 60 Hz. If it distorts the "electricity" a little bit, who cares as long as the voltage and current is close. The power transformer in a tube amp falls into this category, and could be made cheaply, but cheap ones get hot, and sometimes make annoying buzzing sounds.....not a problem in your microwave oven, but you don't want that buzzing sound to annoy you when you are trying to enjoy the music.......

The output transformers connect the tubes to the speakers, they must pass every frequency from 20 Hz (low bass) to 20 KHz (extreme treble) EQUALLY without distorting any signal, or mixing them together.....this is NOT a trivial task. There are a few, very few, custom transformer winders out there that do make very good tube amp transformers, they are not cheap.

You can find cheap transformers for $20 each, and they will SUCK!

You can spend $100 for Edcor or Hammond, and they won't suck, and to most people will sound good enough.

You can spend $500 to $1000 or more for custom transformers wound with silver wire (silver conducts electricity better than copper), and they SHOULD be better than the $100 transformers...How much better? That depends a lot on how critical you are, how good the rest of your system is, and what you listen to. For me, I use the Edcors and design the circuit to make the best of them.

you'd think someone would have figured out how to clone a a MC240 or any of the good MC amps for cheap by now, but I guess you are all still stuck going in circles.

In the 50's and 60's most consumer stereo amps used pretty much the same designs. Yes, there were some outliers like the Citation 2, but many were cookie cutter designs often copied from the tube manufacturers books.

McIntosh took the high road and designed a rather unique circuit which required some unique and expensive to manufacture transformers. It performed better to most users than the usual cookie cutter amps, but cost more, and were built like tanks. They didn't break.

The details of the McIntosh design are well known now. Anyone with the resources available to them could make one, as we have seen, people have done it. Getting the transformers made usually involves someone who knows how doing it themselves. Why?

Well let's say that you make transformers for a living. If it takes you 5 times longer to make that special McIntosh design than it takes you to make a common tube amp output transformer that you sell for $100 each, and you must get special parts to do it, and tear down and setup your equipment every time you make one, how much are you going to charge for it?

The tube amp market is small and getting smaller. McIntosh amps are unique, and never had a large percentage of the market, so what are you going to make if you want to stay in business? Uh, stuff that most people want to buy!

Want that MC240 sound? Find a used one with good transformers and restore it. You could replace every part in the amp for less than you would spend to have one of the two output transformers made.

Want something that sounds just as good? Your choices are unlimited. Everyone has different preferences in sound. Figuring out what you like about the MC240 is the first step. Then find something with those qualities.

I once had a pair of quite rusty, but perfectly working MC30's. I didn't really like the way they sounded, so I sold them....but then all the amps I have now are ones i made.
 
As far as "quality" levels and such, I stick to sensible, reliable, and reasonably-priced areas for parts like transformers. Tubelab's statements above are good advice, and realistic. The Dynaco "clone" transformers offered are excellent for new builds, and I've used them in my own work. The Triode website posted a "shootout" of various transformer brands, comparing the Dynaco clones with several others, and rated them, with the Dynaco's at the top of the heap. Specifically, the Z565 transformer used for 15/15w amps. That was enough for me to decide on, and I'm beyond happy with the results.

That amp that I built from them rocks the house if I choose to.
 
McIntosh output circuit with Plitron-made transformers.

The transformer sets that I have for building my own 400+ WPC amp were made by Plitron....over 12 years ago. They are very good transformers, but as with all things built for sale, Plitron left the tube amp market a few years ago. There are still a few links on their web site, but no products listed for sale.

These were Vanderveen designs, which were also available from Amplimo in Europe. I do not know if Amplimo still makes them.

The OPT's that I have were sold for over $400 USD when they were in their catalog, but dumped for $134 each 12 years ago. when they quit making them. They can be seen in several of my big amp tests over the years since then, including the test where I extracted 525 watts from one of Pete's 20 WPC amp boards with 4 BIG TV sweep tubes wired as a mono block.

See posts #6 and #7 in this thread about building big amps.
 
I have wound many of the popular McIntosh Unity Coupled Outputs...
The demand for these OT's is low and does not justify the cost of building them... In todays world of purchasing wire and iron...you cant order low quantity... I will need to order a pallet of wire and and a pallet of C-cores... The C-cores are available but are custom built to order, not an off the shelf item. The wire is not commonly stocked. it is a special coating that can withstand the corona that occurs in this type of design...

A few slick business guys tried to have me make them a few, only so they can strip them apart and have the winding info for themselves... good thing It was too complicated for them and the made a mess of it... The MI-350 OT is made of of 2 C-cores, same C-core the MC60 and MC75 use.... The MC75 has had more failures than any other OT in McIntosh line... This is because they took the MC60 OT, kept the same C-core and added more turns of wire to it, only problem is that they had to reduce the wire gauge to do it, which led to more breakdown of the wire... McIntosh experimented with several wire coatings for the MC75...eventually they found a wire that worked fairly well....So the earlies of MC75's OT's are more prone to problems..
 
These were Vanderveen designs, which were also available from Amplimo in Europe. I do not know if Amplimo still makes them.

I communicated with someone and learned that there was another European company making Vanderveen transformers as well and they were less expensive than Plitron, even shipped to the US from Europe. I can't remember the name of the other company but I'm sure it wouldn't be too hard to figure it out.

The problem with the Vanderveen Unity-Coupled transformer is the ratio: 6.4k:8. This puts you considerably under the knee of the Vg=0 curve and is hard on the screen. I solved this by putting a simple floating mosfet circuit for each output tube that eats some of the screen voltage put presents a low zout to the screen. I have a lower screen voltage than B+ without the need of a third primary winding like some other Unity-Coupled designs have had to lower screen voltage.
 
That’s a sweet amp you made there
Do you make and sell them? If so how much would one cost?
Great job and work. Looks very well made and looks like it will be wiring for many years to come

Thanks!

I just made one for myself. It was so much work, I don't think I'd even want to think about making one for someone else. The output transformers alone were almost a grand...
 
You can just order the OTs from Edcor or your favorite vendor. Why wind anything at all? And they will automatically be high performance. (see below)

Just get Edcor to make a 600 Ohm to 800 Ohm range P-P OT (depending on B+, power, and tube type) and you will have an excellent OT for Circlotron mode.

Alternately, a 1200 Ohm to 1600 Ohm range P-P OT pair (depending on B+, power, and tube type) would work for the Crowhurst Twin Coupled Amp. (the effective P-P primary Z is twice the nominal xfmr Zpri when used in Twin Coupled mode with paralleled outputs, full winding 4x in parallel with full winding 4x giving 2x effective result, so two 1200 Ohm P to P OTs used to get a 2400 Ohm P to P effective OT) Only a single grounded B+ supply is used for this design. You need a 16 Ohm secondary on them to get an 8 Ohm output when paralleled.

Edcor have multiple output secondaries of 0,4,8,16 Ohm as a standard option, but I would request an additional tap at 2 Ohm so as to have a balanced tap to the 8 Ohm tap for local N Fdbks or alternant speaker impedances. (the 4 Ohm tap is the center tap between 0 and 16 Ohm. 2 Ohm and 8 Ohm would then provide another balanced pair for Fdbks or alternate output impedances, can even shift the paralleled windings by 2 Ohm tap points to get higher speaker Z) Primary UL taps would be helpful too for N Fdbks. Pretty standard to get those.

--------------------------

The Circlotron at 600 to 800 Ohm primary Z will have amazing bandwidth. You can't do any better than Circlotron.

The Twin Coupled design also gets a BW boost. Each tube now uses all 4 primary winding halves of the doubled-up OTs. (there are power supply caps between the two OT primaries at all tap points used so they act as one) This reduces leakage L by 1/4 versus a standard OT. You effectively get twice the interleaves in the OT plus using the full primary instead of half a primary as usual for P-P. Twin Coupled turns a low cost OT into a premium OT, but of course the price is doubled for two xfmrs.

----------------------------

The Mac unity coupled design puts bifilar wires adjacent to each other, with full B+ between them. Either use expensive polyimide insulation or expect failure before long. Transformer wires do vibrate magnetically unless encapsulated, wax, tar, epoxy etc.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.