Class AB2, not class B, no wonder! Now you have to ask yourself why all this crazy OT configuration and massive feedback stuff.
---------------------
I'm shocked. You mean the OT is not even layer wound? OMG! I would envision a "correct" Mac OT as being layer wound and interleaved on a split bobbin. Even further, it should be on a long-E core or a cut core (or ultimately a toroid with progressive wind). Using the 50% cathode feedback scheme, both bifilar winding sets MUST have equal coupling to the output winding or HF distortion results from imbalanced leakeage L. Only way to do that is with a split bobbin with interleaves.
Although an el-cheeso method would be to multifilar wind the whole thing, and put up with massive distributed capacitance. I shure wouldn't pay $4500 for that. The "Erector Set" amplifier? One remotely possible scheme, but unlikely on a non long-E core, would be to try a progressive wind. This keeps adjacent turns of the winding with minimum AC voltage difference next to each other everywhere by careful proximity winding practice. Probably hand wound and with adhesive on the wire to make it stay in place (a very mechanically awkward winding is required although multiple PI sections are another way).
The original Mac OT was wound using enamel magnet wire insulation and had problems with HV breakdown. The adjacent strands in the bifilar pairs have ground and B+ on them, a recipe for trouble. Modern super magnet wire insulations are available now though that can withstand 1000's of volts.
But if one were going to try to clone the Mac, I would just use a Circlotron output scheme. Functionally equivalent. Better AC coupling, no DC problems, 1/2 the turns, could even multifilar the secondary in using ordinary TFE wire insulation on it (the secondary). And off the shelf OT candidates are readily available. This pre-occupation with duplicating the Mac OT is absurd. Especially in light of the class AB2 operation.
Why clone the Mac when it would be easier to do it better? (differential driver, "magic" Schade feedbacks, Circlotron output on a long-E core).
---------------------
I'm shocked. You mean the OT is not even layer wound? OMG! I would envision a "correct" Mac OT as being layer wound and interleaved on a split bobbin. Even further, it should be on a long-E core or a cut core (or ultimately a toroid with progressive wind). Using the 50% cathode feedback scheme, both bifilar winding sets MUST have equal coupling to the output winding or HF distortion results from imbalanced leakeage L. Only way to do that is with a split bobbin with interleaves.
Although an el-cheeso method would be to multifilar wind the whole thing, and put up with massive distributed capacitance. I shure wouldn't pay $4500 for that. The "Erector Set" amplifier? One remotely possible scheme, but unlikely on a non long-E core, would be to try a progressive wind. This keeps adjacent turns of the winding with minimum AC voltage difference next to each other everywhere by careful proximity winding practice. Probably hand wound and with adhesive on the wire to make it stay in place (a very mechanically awkward winding is required although multiple PI sections are another way).
The original Mac OT was wound using enamel magnet wire insulation and had problems with HV breakdown. The adjacent strands in the bifilar pairs have ground and B+ on them, a recipe for trouble. Modern super magnet wire insulations are available now though that can withstand 1000's of volts.
But if one were going to try to clone the Mac, I would just use a Circlotron output scheme. Functionally equivalent. Better AC coupling, no DC problems, 1/2 the turns, could even multifilar the secondary in using ordinary TFE wire insulation on it (the secondary). And off the shelf OT candidates are readily available. This pre-occupation with duplicating the Mac OT is absurd. Especially in light of the class AB2 operation.
Why clone the Mac when it would be easier to do it better? (differential driver, "magic" Schade feedbacks, Circlotron output on a long-E core).
Last edited:
Smoking-amp,
The transformer was wound in layers with adjacent primary and secondary sections which is good practice, but there was no isolation between layers. With isolation each next layer has a smooth plane so the bobbin can be wound tidy all the way up.
My favorite core for push-pull output transformers is a single c-core (stacked when necessary), and that is what I use for that purpose exclusively. For a single c-core you can wind two bobbins, one for each leg of the core. This way you have accurate center taps.
The older generation Mac amps used c-cores as well but I think in a double core set up with a single bobbin.
The older Mac transformers used what was called Formvar wire, which is magnet wire with thicker enamel. It will work OK as long as the bifilar winding is tidy and there is no room for resonance. I am curious how one of the earlier transformers were wound as I would not be surprised if it would look much better. Apparently the modern EI output transformer is made on a budget basis, which is a pity in an amplifier where the output transformer is its heart and soul.
The transformer was wound in layers with adjacent primary and secondary sections which is good practice, but there was no isolation between layers. With isolation each next layer has a smooth plane so the bobbin can be wound tidy all the way up.
My favorite core for push-pull output transformers is a single c-core (stacked when necessary), and that is what I use for that purpose exclusively. For a single c-core you can wind two bobbins, one for each leg of the core. This way you have accurate center taps.
The older generation Mac amps used c-cores as well but I think in a double core set up with a single bobbin.
The older Mac transformers used what was called Formvar wire, which is magnet wire with thicker enamel. It will work OK as long as the bifilar winding is tidy and there is no room for resonance. I am curious how one of the earlier transformers were wound as I would not be surprised if it would look much better. Apparently the modern EI output transformer is made on a budget basis, which is a pity in an amplifier where the output transformer is its heart and soul.
Modern magnet wire has very good insulation. At work I was playing with the new HIPOT tester and twisted two strands of 32 awg wire together. It passed the 5KV AC/DC test with no arcing or leakage. However the coating is delicate. All it takes is a minor scrape to short against a metal core.
it's my understanding, that the original Mc amps were C-core ...
McIntosh MC275 Amplifier Reviews | ProductWiki
McIntosh MC275 Amplifier Reviews | ProductWiki
Them thar thangs is called bail-n coils around these parts. 
I recall seeing a lot of old telephone loading coils made with steel wires. Must have been enameled. I don't think you can get grain oriented stuff without rolling it though, but permalloy wire would be interesting. Why cut it then. Use a cheap random wind toroid winder to put it over the (large diameter) wire coil bobbin.
I recall seeing a lot of old telephone loading coils made with steel wires. Must have been enameled. I don't think you can get grain oriented stuff without rolling it though, but permalloy wire would be interesting. Why cut it then. Use a cheap random wind toroid winder to put it over the (large diameter) wire coil bobbin.
Last edited:
Smoking-amp (and others), I appreciate this discussion though I understand precious little of it. I do not plan on cloning the legendary MC275, but do hope you guys figure out the (apparently somewhat mysterious) design decisions that were made. I accept the notion there are better designs now, and less complex designs, but I'm just not sure those old units were terribly more complex than other tube amps. It's not as though they are FILLED with miles of wire and thousands of components. They were different, and better, than all the others of that time. Again, I'm wondering if some of their wisdoms are still applicable. Bet they are. These were not a bunch of idiots.
Again, I am too ignorant to answer your questions. I can only express my goal in terms of sound, and of course that is very subjective. The best tube amps (according to my ear) have certain characteristics. Heck.....I'm wondering if all that current passing throught the 12AX7's is part of the secret. I hope to build an amp that has this "fat" -ness in the lower tones. Warmness throughout the range, but as this warmness covers those lower tones, you can get this "fat" -ness that is absent in most solid state gear. Again, a warmness that covers the entire range, and an "air" -iness in the mids and highs. An "organic" sound. And now for the hard part: These warm, fat, airy, organic features need to be achieved without flabbiness. Notes played through this amp need to retain all their amplitude, without that resonance stuff. If a rock drummer hits his kick drum, it needs to NOT sound like a kettle drum. I wonder if this helps.....hope it does.
I am keeping an eye out for a variac or some device with which I can regulate voltage to this Yaqin MC100B. Because the consensus wisdom seems to be that this improves sound, and perhaps tube longevity (which I don't care about). Replacing some caps might help too. This amp does not sound right yet. It's got some decent fatness in the bass, and some airiness in the mids and highs, but it's flabby. Not tight-sounding. And there is a tad too much brightness at the very top.
BTW, we're getting some cold weather in my location, and possible storms. If things get really frigid, I will put a fire in the fireplace for ambiance, and we'll huddle around this amplifier for warmth. This thing makes HEAT.
Since you are looking for a power amp you'll of course have a preamp to feed it. I'm reading all the characteristics you are looking for described as "warm" and "fat" and "airy" and "flabby"....
Why not set your amp goals to find one that doesn't do anything to the input signal to color it at all, just amplify it. You then use the controls on your pre- and use speaker placement to get the sound you want. Do you really want the amp to do anything other than amplify? Anything else "added or lacking" is a failure of the design. After all, the amp designer didn't intend for it to do anything else except reproduce the source material as closely to perfectly as possible.
That's a good suggestion, one I can understand, and is a big part of my current thinking. A power amp that is just a very obedient amplifier. It amplifies. Then, "cook" the sound using a preamp.
Yup, in theory, amplifiers are all obedient. In the real world though, they color the sound. And if they don't.....they sound sort of harsh. Clinical. There is a sound I am looking for, and it is warmer and more organic than most of the solid state stuff I have heard. All of it, actually.
Yup, in theory, amplifiers are all obedient. In the real world though, they color the sound. And if they don't.....they sound sort of harsh. Clinical. There is a sound I am looking for, and it is warmer and more organic than most of the solid state stuff I have heard. All of it, actually.
Is that what the recording engineer/producer would say about the quality of the recording?
If you feed the amp a great recording, what would the "warm" output signal look like compared to the input? How different? Or from the other perspective, what does the "harsh, clinical" signal lack that the input signal has? In either case, say how you think the signal is different other than being bigger.
I've watched guys mixing music using digital equipment where they're looking at waves on a computer screen, but I've never gotten into that stuff. They use those graphic representations of the sound waves to make decisions about processing. Again, it's an art and a science, and I don't have any experience with those graphic representations, other than watching over someone's shoulder.
Nevertheless, aside from these newfangled tools, recording engineers make all sort of decisions about sound processing. The board contains sound-processing circuitry. The tape machine contains some. There is no such thing as "correct."
That said, I could imagine there are machines capable of looking at the line level signal and comparing it to the amplified signal, to check for accurateness. Looking for distortion. Looking even for distortion that cannot be heard. Then that amplified signal still needs to go through some speakers, where more color is attached to the sound. I do not know how, or whether, testing equipment can isolate the "warmness" feature. Ears can, though. It makes sense to use sophisticated equipment to identify and learn and repeat and control, but the bulk of our sound preferences are subjective, not objective. In fact, all of our sound preferences are subjective.
Nevertheless, aside from these newfangled tools, recording engineers make all sort of decisions about sound processing. The board contains sound-processing circuitry. The tape machine contains some. There is no such thing as "correct."
That said, I could imagine there are machines capable of looking at the line level signal and comparing it to the amplified signal, to check for accurateness. Looking for distortion. Looking even for distortion that cannot be heard. Then that amplified signal still needs to go through some speakers, where more color is attached to the sound. I do not know how, or whether, testing equipment can isolate the "warmness" feature. Ears can, though. It makes sense to use sophisticated equipment to identify and learn and repeat and control, but the bulk of our sound preferences are subjective, not objective. In fact, all of our sound preferences are subjective.
The list of questions I posed earlier was mostly related to aesthetic questions. What type of tubes would relate to buying replacement tubes later that were either (hopefully) still in production or of fading availablity (but cheap now) NOS Sweeps. Some of the Sweeps can be made to work better than the Audio tubes, Mac even used the 6LQ6 for an amp for example. Plate caps would mainly be a safety versus classical looks issue (kids around?). Power output needed would depend on speaker sensitivity, type of music, and room size.
The "sound" of the amplifier is tough to nail down. There are the rather obvious harmonic distortions seen in SE amps, 2nd harmonic generally still considered benign.
Then there are things like damping factor (generally lower for tube amps, ie Zout greater than 0), some old amplifiers even had damping factor adjustments (there is a concept called critical damping factor, related to speaker Z, that was popular at one time).
There is the issue of "constant" output impedance versus signal or reflected EMF from the speaker. SS amps drive the output Z to zero with their high feedback, but if you want a specific damping factor it should be constant versus signal. Most class AB low feedback amplifiers are not. (Class A does not, but is a room heater) The "magic" Schade feedback scheme I mentioned earlier is a means to ensure constant Zout over the crossover region, where typically amps have the problem.
Then there is the issue of variable capacitance in the amplifying devices, possibly causing minute dynamic phasing effects. SS amps have a problem here, and demand massive feedback to overcome it. The tube devices are much more constant, and so will work well in this regard still using low feedback.
And other issues like hum, regulated power supplies, clipping, crossover distortion artifacts, bandwidth (usually HF rolled off for most tube amps) and any uneveness of gain verus frequency.
(brightness.....)
You can find some general guidelines to what is better, but getting a specific "sound" of some amplifier is tough to order up. Lots of DIYers seem to be able to tweek in effects using exotic components subbed in later, I haven't tried any of that, so couldn't say how effective that is. A higher feedback design will be more immune to component changes though. But even there, there are some tricks that can be played with inverted triode feedback or feedback to the input tube screen grid (ie, slightly non-linear feedback networks).
The "sound" of the amplifier is tough to nail down. There are the rather obvious harmonic distortions seen in SE amps, 2nd harmonic generally still considered benign.
Then there are things like damping factor (generally lower for tube amps, ie Zout greater than 0), some old amplifiers even had damping factor adjustments (there is a concept called critical damping factor, related to speaker Z, that was popular at one time).
There is the issue of "constant" output impedance versus signal or reflected EMF from the speaker. SS amps drive the output Z to zero with their high feedback, but if you want a specific damping factor it should be constant versus signal. Most class AB low feedback amplifiers are not. (Class A does not, but is a room heater) The "magic" Schade feedback scheme I mentioned earlier is a means to ensure constant Zout over the crossover region, where typically amps have the problem.
Then there is the issue of variable capacitance in the amplifying devices, possibly causing minute dynamic phasing effects. SS amps have a problem here, and demand massive feedback to overcome it. The tube devices are much more constant, and so will work well in this regard still using low feedback.
And other issues like hum, regulated power supplies, clipping, crossover distortion artifacts, bandwidth (usually HF rolled off for most tube amps) and any uneveness of gain verus frequency.
(brightness.....)
You can find some general guidelines to what is better, but getting a specific "sound" of some amplifier is tough to order up. Lots of DIYers seem to be able to tweek in effects using exotic components subbed in later, I haven't tried any of that, so couldn't say how effective that is. A higher feedback design will be more immune to component changes though. But even there, there are some tricks that can be played with inverted triode feedback or feedback to the input tube screen grid (ie, slightly non-linear feedback networks).
Last edited:
The manufacturers don't list those subjective qualities you wanted in the specs though, for real "objective" reasons. How will you know what to buy, or try to build? If an amp truly sounds "warm" compared to another, what is the difference in the output. That "subjective" quality will be measureable because it is a real difference in the signal. What spec does it represent?
"How will you know what to buy, or try to build?"
Well, its easy to get annoying artifacts if some things are cheapened in the amp. So these kind of issues need to be dealt with. Ie, regulated power supplies instead of flimsy ones, adjustable bias and gain balance adjustment to match up tubes better, DC filament power, eliminate crossover distortion ..... These typically make the amplifier expensive when fully dealt with. But this will only get you part way to some particular "sound".
Well, its easy to get annoying artifacts if some things are cheapened in the amp. So these kind of issues need to be dealt with. Ie, regulated power supplies instead of flimsy ones, adjustable bias and gain balance adjustment to match up tubes better, DC filament power, eliminate crossover distortion ..... These typically make the amplifier expensive when fully dealt with. But this will only get you part way to some particular "sound".
Last edited:
Right you are. The goal is subjective, and there are no specs to describe it. Like the meaning of the term "obscene," I'll know it when I hear it.
In the meantime, yes. I will likely massage this Yaquin POS, and perhaps improve the sound. But then I might fall down the slippery slope of building, tweaking, researching, tweaking, discussing, tweaking......... Sound familiar?
In the meantime, yes. I will likely massage this Yaquin POS, and perhaps improve the sound. But then I might fall down the slippery slope of building, tweaking, researching, tweaking, discussing, tweaking......... Sound familiar?
I just looked at schematic for Yaquin MC100B. It lacks 16 ohm output tap.
So, not an easy candidate for cathode feedback mod per ARC-ST-70-C3.
Anyone look at this and see another way to clean it up?
http://www.diyaudio.com/forums/tubes-valves/109958-yaqin-mc-100b-schematic-diagram.html
I suspect its flabby cause there is no accurate feedback of any kind from
the load side of the transformer. Any feedback has to couple back through
the iron to the plate and/or screen. Not bad, but damping could be better.
It would be easier to add GNF to tighten damping, but risks becoming an
unstable and less transient tolerant amp.
-------
Nevermind, there is already GNF from the 8 ohm tap, I just didn't see it.
A lot of direct coupling going on, should be almost no evil phase shift to
mess up the GNF. No idea now why it might be flabby in the bass dept?
So, not an easy candidate for cathode feedback mod per ARC-ST-70-C3.
Anyone look at this and see another way to clean it up?
http://www.diyaudio.com/forums/tubes-valves/109958-yaqin-mc-100b-schematic-diagram.html
I suspect its flabby cause there is no accurate feedback of any kind from
the load side of the transformer. Any feedback has to couple back through
the iron to the plate and/or screen. Not bad, but damping could be better.
It would be easier to add GNF to tighten damping, but risks becoming an
unstable and less transient tolerant amp.
-------
Nevermind, there is already GNF from the 8 ohm tap, I just didn't see it.
A lot of direct coupling going on, should be almost no evil phase shift to
mess up the GNF. No idea now why it might be flabby in the bass dept?
Last edited:
How big is the Output Xfmr? Maybe its not passing LFs well.
-----------------------
By the way, since the MC-275 re-issue is only using the straight-forward two winding OT (primary side) now, one could actually do a trivial clone of it using a Circlotron output stage and xfmr. Hammond 1650T with 8 Ohm load on its 16 Ohm tap in Circlotron gives the same 4000 Ohm total primary (equivalent P-P, actually abt 1000 Ohm winding). I'll bet you can't tell any difference from the real thing. Mc-275 for $200, if you've got some bench supplies around and a bake pan chassis.
Well, the tertiary global feedback winding would have to be adapted to come off the actual secondary (possibly just revert to the earlier MC-275 front end setup there.)
http://farm5.static.flickr.com/4037/4204605299_bdfa12827f_b.jpg
-----------------------
By the way, since the MC-275 re-issue is only using the straight-forward two winding OT (primary side) now, one could actually do a trivial clone of it using a Circlotron output stage and xfmr. Hammond 1650T with 8 Ohm load on its 16 Ohm tap in Circlotron gives the same 4000 Ohm total primary (equivalent P-P, actually abt 1000 Ohm winding). I'll bet you can't tell any difference from the real thing. Mc-275 for $200, if you've got some bench supplies around and a bake pan chassis.
Well, the tertiary global feedback winding would have to be adapted to come off the actual secondary (possibly just revert to the earlier MC-275 front end setup there.)
http://farm5.static.flickr.com/4037/4204605299_bdfa12827f_b.jpg
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.