I'd be pretty surprised if the Macs do any real compression, they were known for low distortion in their day. Being high feedback, I would guess they are simply more "clinical" sounding like a SS amp. No "tubeyness" or slack damping. But I haven't heard one, so just a guess.
--------------------------
By the way, that RCA schematic (handbook) uses just a 47 Ohm tail resistor in the driver stage, so it hardly does any of the "magic" crossover fixing. Maybe they didn't know about the fixup scheme back then after all. It needs a CCS tail in the driver stage.
--------------------------
By the way, that RCA schematic (handbook) uses just a 47 Ohm tail resistor in the driver stage, so it hardly does any of the "magic" crossover fixing. Maybe they didn't know about the fixup scheme back then after all. It needs a CCS tail in the driver stage.
Last edited:
Compressed was for lack of a better term. Maybe constipated would have been better... Feedback does not function like an audio compressor.
Smoking-amp, you got it right on that the sound is very tight, much like a SS amp. With feedback the damping goes way up which is part of the reason for the tightness. Here's the deal with MAc amps though... If I play deep notes like organ music through my MACs, the woofers of my speakers will flop around like crazy because they do not do well with sub-sonic frequecies. I can then put an all triode amp that I have that has no feedback and the speakers will track along with the notes; no boominess what-so-ever, but the bass is still full and very tight. This has a lot to do with the internal resistance of the output triodes matching up much better to the impedance of the output transformer. Negative feedback is not needed to better match up the output stage.
Talking about the way an amp sounds is SO subjective; the best way is for a listener to get a tube amp, listen to it, take out the GNF and then listen to it again. A lot of amps will sound worse, some sound better without feedback, and a few will give the listener euphoria from the experience. Now does that mean that those few amps measure better spec-wise than the rest of the amps? No, most likely, they will not, but they sound so good... Back again to my statement that describing an amp's sound is so subjective.
Anyway: this is for smoking amp: I have been thinking about what you had said about crossover distortion and the MAC amp and you are right. I was thinking in terms of the gap in the waveform one sees on an O-scope when the output tubes of a conventional PP output stage is biased too lean. MACs don't really have this but they do suffer from the Gm kick when one of the outputs turns on.
Daniel
Smoking-amp, you got it right on that the sound is very tight, much like a SS amp. With feedback the damping goes way up which is part of the reason for the tightness. Here's the deal with MAc amps though... If I play deep notes like organ music through my MACs, the woofers of my speakers will flop around like crazy because they do not do well with sub-sonic frequecies. I can then put an all triode amp that I have that has no feedback and the speakers will track along with the notes; no boominess what-so-ever, but the bass is still full and very tight. This has a lot to do with the internal resistance of the output triodes matching up much better to the impedance of the output transformer. Negative feedback is not needed to better match up the output stage.
Talking about the way an amp sounds is SO subjective; the best way is for a listener to get a tube amp, listen to it, take out the GNF and then listen to it again. A lot of amps will sound worse, some sound better without feedback, and a few will give the listener euphoria from the experience. Now does that mean that those few amps measure better spec-wise than the rest of the amps? No, most likely, they will not, but they sound so good... Back again to my statement that describing an amp's sound is so subjective.
Anyway: this is for smoking amp: I have been thinking about what you had said about crossover distortion and the MAC amp and you are right. I was thinking in terms of the gap in the waveform one sees on an O-scope when the output tubes of a conventional PP output stage is biased too lean. MACs don't really have this but they do suffer from the Gm kick when one of the outputs turns on.
Daniel
Daniel,
Your statement on feedback, damping and tightness concerning the Mac amps is far from complete.
McIntosh had to apply large amounts of feedback because the primary impedance of the output transformer is a MISMATCH in the first place.
The KT88's are used as TRIODES here, and we know that a good impedance match with loudspeakers point in the direction of 4 to 5k plate-plate impedance, also when cathode feedback is applied!
Well, the MC-275 has only 2k plate-plate impedance (to get that much power out of a pair of KT88's). Therefore the feedback is really used here as a means to cure a mediocre design, and therefore so many gain stages are necessary to compensate for the loss of gain as a result of the feedback.
When the output transformer had a 5k or so primary impedance there had been no need for GNFB, and the amp would work with at least one gain stage less.
My guess is that you'd listen to a much better amplifier.
Hugh Alvin Lockhart did it right in 1956.
Pieter
Your statement on feedback, damping and tightness concerning the Mac amps is far from complete.
McIntosh had to apply large amounts of feedback because the primary impedance of the output transformer is a MISMATCH in the first place.
The KT88's are used as TRIODES here, and we know that a good impedance match with loudspeakers point in the direction of 4 to 5k plate-plate impedance, also when cathode feedback is applied!
Well, the MC-275 has only 2k plate-plate impedance (to get that much power out of a pair of KT88's). Therefore the feedback is really used here as a means to cure a mediocre design, and therefore so many gain stages are necessary to compensate for the loss of gain as a result of the feedback.
When the output transformer had a 5k or so primary impedance there had been no need for GNFB, and the amp would work with at least one gain stage less.
My guess is that you'd listen to a much better amplifier.
Hugh Alvin Lockhart did it right in 1956.
Pieter
Pieter,
Not quite sure what you mean by this as the voltage between the cathode and screen stays constant. That spells pentode operation to me...
Looked at a stock diagram for a 275. Doesn't look like a triode strapped circuit. They wouldn't be able to call it a 275 if it was factory triode.
A quick search shows KT88 triode power would only be about 30W/channel 485v B+ with 4K output Z. I can't speak from experience.
20to20,
Not only I dismantled the shorted output transformer to check the winding ratios, I also measured the amp.
So the plate-plate impedance is 2k048 referred to 8 ohms to be exact, not 4k.
A triode connected KT88 with the same load in the plate and cathode circuit is able to swing 200 VRMS (for comparison: a normally operated 300B swings about 150 VRMS max).
At full power the amp works in class B, so we have 400 VRMS swing plate-plate.
The winding ratio for 8 ohm is 16:1, so we have 25 VRMS at the loudspeaker terminal at full power; that is around 75 watts.
Not only I dismantled the shorted output transformer to check the winding ratios, I also measured the amp.
So the plate-plate impedance is 2k048 referred to 8 ohms to be exact, not 4k.
A triode connected KT88 with the same load in the plate and cathode circuit is able to swing 200 VRMS (for comparison: a normally operated 300B swings about 150 VRMS max).
At full power the amp works in class B, so we have 400 VRMS swing plate-plate.
The winding ratio for 8 ohm is 16:1, so we have 25 VRMS at the loudspeaker terminal at full power; that is around 75 watts.
If you listen to a well made LM3886 chip amp, like the Jeff Rowland clone, you will hear extremely bright accurate music. Then you will be able to understand how compressed or not so blatant music sounds.
These terms are very hard to express in the written language it is just the best way I have found to compare one type of sound to another.
I prefer the Mac sound it seems more musical and natural like my Leach amp.
Of course each persons opinion is theirs to own.
On another note I have a hard time trying to understand why output transformers for tube amps cost so much. There is not a terribly large amount of material involved in there manufacture. Besides the Output trannys and some quality tubes the rest of the amp is off the shelf items and rather cheap.
You can fabricate a nice aluminum or stainless chassis in the garage with minimal tools and a bit of patience.
If I could find a really nice 100 watt tube layout I would give it a go. I have always liked the way they sound. Very easy to listen to.
Tad
These terms are very hard to express in the written language it is just the best way I have found to compare one type of sound to another.
I prefer the Mac sound it seems more musical and natural like my Leach amp.
Of course each persons opinion is theirs to own.
On another note I have a hard time trying to understand why output transformers for tube amps cost so much. There is not a terribly large amount of material involved in there manufacture. Besides the Output trannys and some quality tubes the rest of the amp is off the shelf items and rather cheap.
You can fabricate a nice aluminum or stainless chassis in the garage with minimal tools and a bit of patience.
If I could find a really nice 100 watt tube layout I would give it a go. I have always liked the way they sound. Very easy to listen to.
Tad
On the diagram I see the upper tube screen connected to the lower tube plate. Is that configuration still considered triode mode, if the screen isn't connected to its own tube?
20to20,
I am not so sure anymore; you might be right that it is not pure triode configuration.
Someone must throw some light on this matter.
Anyhow the same amp would yield 40 watts of power with a 4k output transformer, much more in the direction of the 30 watt, but then the Mac applies cathode feedback.
I am not so sure anymore; you might be right that it is not pure triode configuration.
Someone must throw some light on this matter.
Anyhow the same amp would yield 40 watts of power with a 4k output transformer, much more in the direction of the 30 watt, but then the Mac applies cathode feedback.
Core material is dirty cheap (even very high quality GOSS or amorphous steel) and sells in tons or huge bobbins of several hundreds of Kg. However, design and test involves a LOT of tedious labor, yet the selling quantity is miserable. For example, one of the largest (and best) transformer factories in the world - Tamura have (if I'm not mistaken) a year turnover over 600 millions of $$. Are they really interested to sell let's say 100 units per $100 each??
You can have output transformers quite affordable if:
1) You design them by yourself + you find who can make them in small quantity at reasonable price. For example, when I inquired transformer factory in Germany (I wanted just a few chokes, not transformers), they wish to charge me 100 EURO/pc + 3 months lead time. Factory is a factory, they are not one-man food shop and have no interest to acquire $100 from someone, it will not pay off even material/warehouse/production management, accounting, and paper work.
2) You design & wind them by yourself using simple home-made machinery.
PS. Right now I'm started testing transformer of my own design, wound on custom-made core, 60W PP @ 12500G, 20Hz - 100,000 KHz frequency response (upper limit higher for sure but I have no suitable equipment to test it), leakage inductance < 20 mH, unit seem to capable handle significant imbalance of output tubes idle current without LF drop. Comparable brand-name transformer (Tamura, Tango, Hashimoto) costs over 500 EURO/pc in Europe ($1368 for a pair). Lundahl offered them for 300 Euro/pc, but as former software programmer I could not resist temptation to design my own trafos.
Last edited:
Pieter,
Plate and screen are not tied together. Plate is tied to the other output tube screen.
AC-wise, tube 1's cathode tracks tube 2's plate with a DC offset equal to B+. Pentode operation without a separate screen supply, but the screens have to be at B+.
I promise, they are acting as pentodes. That's why they used a low plate load. The knee of the curves is high due to the high screen voltage.
Plate and screen are not tied together. Plate is tied to the other output tube screen.
AC-wise, tube 1's cathode tracks tube 2's plate with a DC offset equal to B+. Pentode operation without a separate screen supply, but the screens have to be at B+.
I promise, they are acting as pentodes. That's why they used a low plate load. The knee of the curves is high due to the high screen voltage.
Its easier to see whats really going on here with Crowhearst's twin coupled.
You can clearly see the cross coupled caps that hold each screen in a pure
pentode mode.
McIntosh, might or might not have discrete caps in parallel with the bifilar
pair. Capactitance of the bifilar pair alone can serve the same purpose.
Each plate is capacitively cross-coupled to the other cathode. Therefore
V1's plate tracks the cathode of V2, and provides the perfect voltage to
feed V2's screen in pure pentode.
Triode woulda been V1's plate into its own screen, but we are reaching
across to grab ahold of the exact opposite...
Weirdly, screen currents cancel out in the cross coupling. Magnetic field
across the OPT is only Amperturn difference of the plates.
Last edited:
I agree on the pentode operation. The screen voltages are bootstrapped to keep constant cathode to screen V. It's restricted to a screen V at B+ with that trick. Some Mac versions may have used an extra screen winding on the OT to allow lower screen V applied to the center tap.
Almost forgot, one of our members has done a semi clone of the Mac. Jan Vieset, and he has a web site describing it:
BigMac, Unity coupled output amplifier
Some .pdf papers there too, relevant to the Mac.
Link for the Twin Coupled amplifier by Crowhurst:
http://www.audioxpress.com/magsdirx/ax/addenda/media/1960crowhurst.pdf
Almost forgot, one of our members has done a semi clone of the Mac. Jan Vieset, and he has a web site describing it:
BigMac, Unity coupled output amplifier
Some .pdf papers there too, relevant to the Mac.
Link for the Twin Coupled amplifier by Crowhurst:
http://www.audioxpress.com/magsdirx/ax/addenda/media/1960crowhurst.pdf
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.