Do you think the original McIntosh founders anticipated their original equipment would still be operational and sought after? Are any of them still alive?
The original McIntosh amplifiers came out is the early 50's I believe. What speakers were typically used with these original amps (we they considered "Hi-Fi")? It was a few years before the original AR units came out (I think). Did loudspeaker technology lag behind the McIntosh amp technology then?
The original McIntosh amplifiers came out is the early 50's I believe. What speakers were typically used with these original amps (we they considered "Hi-Fi")? It was a few years before the original AR units came out (I think). Did loudspeaker technology lag behind the McIntosh amp technology then?
Last edited:
lux 3045
I modified several Lux 3045 to KT88's, the 8045's became very difficult to find.
The circuit may be very good, but the amplifiers are very unreliable; had to have new PCB's made. The originals must have been designed by some one from Lux trany radio department, as they used the same materiel. I was lucky to have a friend who copied the PCB's using Fiber Glass. Amazing how a Big firm will destroy a product just to save a few cents.
Phil
I modified several Lux 3045 to KT88's, the 8045's became very difficult to find.
The circuit may be very good, but the amplifiers are very unreliable; had to have new PCB's made. The originals must have been designed by some one from Lux trany radio department, as they used the same materiel. I was lucky to have a friend who copied the PCB's using Fiber Glass. Amazing how a Big firm will destroy a product just to save a few cents.
Phil
Attachments
Dynaco and McIntosh had their chassis made at the same factory. The Mac mainly had prettier front panels.
I am still green in audiophile, but it seems like a well designed amp from the older days can be just as good. All the technologies advancement in electronics do not necessary touching that. my brother in law inhered an old amp that looks very nice and very very heavy. Looks to be of quality amp 50 years ago. I keep telling him don't dump it. Compare to their $400 Nakamichi amp, my bet is that door stopper might sound better.
I think loud speaker is the single source of distortion. I don't know much about speakers, I heard about full range speakers. I just don't know how can they make a full range speaker that have light mass for treble and at the same time have big cone for the bass. I read people complain about the crossover being the single major source of distortion and should avoid at all cost. I better let people that know speakers to tell me. More a question than comment.
Except this is not the truth of the matter.
Other than the fact that both brands used tubes, the similarities end just about there. Compare the schematics. Compare the performance too.
Early McIntosh gear had no front panels on the amps... so...
_-_-
That Lux 3045 circuit looks overly busy with extra primary windings. The use of a separate primary, D, for the differential driver stage feedback means it misses the differential current sensing signal in the CFB winding (winding resistance derived). And why they would want two windings, A & B, that are cap coupled with just a DC voltage difference between them, when some resistors would do instead. Taking "global" feedbacks from both sides of the OT seems like it would be corrupting the final output. I could see if the feedback networks did a transition at HF to local Fdbk, but they don't. SY, are you sure about this amp as an example?
On the Mac Unity Coupled style amplifier absorbing leakage inductance spikes in class B cutoff, this is only really necessary in class B. Class AB should ameliorate this problem with both sides conducting during transition, insuring a low Rp absorber. Another approach would be to just put some HV Zeners or shunt regulators on the primary winding ends, set just above the normal max voltage excursion, so as to absorb the inductance cutoff spikes. Same problem occurs in switching supplies.
On the Mac Unity Coupled style amplifier absorbing leakage inductance spikes in class B cutoff, this is only really necessary in class B. Class AB should ameliorate this problem with both sides conducting during transition, insuring a low Rp absorber. Another approach would be to just put some HV Zeners or shunt regulators on the primary winding ends, set just above the normal max voltage excursion, so as to absorb the inductance cutoff spikes. Same problem occurs in switching supplies.
I have seen some speaker designs where two drivers were put together front to back in the same speaker cabinet. By driving the back speaker appropriately, the driven front speaker sees no back pressure from the box. So a small box can be used that emulates a large box. So if one put a SS amp on the back speaker and a SET amp on the front speaker, one maybe gets an efficient speaker in a small box. Seems like any old speaker cabinet could have two speakers mounted (drivers front to front and phased correctly) in the same existing speaker mount. Then wire them up separately to two amps.
Last edited:
Some people agree with that assessment. I happen to be one of them.
Alan0354 said:
That is a very good analogy to explain likely what's happening with these old Mc's. Too many bits and baubels in the circuit path. And for the amps with follower output stages (Essentially all the tube poweramps), the increase in complexity of the driver stage in order to drive that output stage swamps all the benefits of that output topology. There is likely a good answer to the CF tube output stage, but I don't think the McIntosh solution was all that good.
Their chassis were from the same vendor, not the circuitry. This I know for a fact. And their current tube amps also lack a front panel.
To me, the older McIntosh equipment was sonically mediocre at best. At least the Dyna stuff was able to be upgraded to reasonably good sound at a low cost.
You could place the input and output jacks on opposite side of the chassis. The power supply should be nearest the power section and as far from both inputs as you can be. I built an amp like that. It was designed to be someplace out of sight and had no volume controls. I've started using Speakon connectors for power outputs, Those are so easy to use that you can connect/disconnect speaker cables on the back of a chassis by "feel without looking
My next build (parts ordered) is going to be a microphone preamplifier. These have a "ton" of voltage gain (about 60db) so I'm placing input and output XLR jacks on opposite ends again.
I definitely think my modified Dynaco sonically bests the McIntosh units. Whether their performance does or not is a moot point. Tone is everything. I am concentrating on small signal work right now because tone really is an issue (I am finding) there.
I will say this though. The distortion in my power amps below 10k varies between 0.15 and 0.25 percent. McIntosh relies on gobs of feedback to achieve those numbers, where I am only using 12dB. If I want any better at the upper frequencies, I will need better iron. That is silly because the harmonic power is inaudible up there anyways. In addition, those numbers require a matched set of output tubes, 50 to 70 percent maximum plate dissipation bias, and the long tail B+ resistor pot adjusted when the tube is changed.
McIntosh does not provide provisions to adjust bias. I assume notch distortion due mismatches are scrubbed away with the feedback. Please correct me if I am wrong about this.
I will say this though. The distortion in my power amps below 10k varies between 0.15 and 0.25 percent. McIntosh relies on gobs of feedback to achieve those numbers, where I am only using 12dB. If I want any better at the upper frequencies, I will need better iron. That is silly because the harmonic power is inaudible up there anyways. In addition, those numbers require a matched set of output tubes, 50 to 70 percent maximum plate dissipation bias, and the long tail B+ resistor pot adjusted when the tube is changed.
McIntosh does not provide provisions to adjust bias. I assume notch distortion due mismatches are scrubbed away with the feedback. Please correct me if I am wrong about this.
OK so now you have two drivers, each making sound exactly as loud as the other except they are 180 out of phase. The sound from each driver will mix either constructively or destructively based on the frequency and you get a "comb filter" effect. Not good. The frequency response is full of spikes
Why not face both drivers the same way and run them in phase. This doubles the efficiency and does not have that comb filter effect.
There really is no free lunch.
BTW you would not need a second amp to drive the out of phase driver. Just run them electrically in parallel and swap the + and - terminal on one of them.
On other problem with the out of phase rear driver. For bass the wave length is longer then the path around the cabinet. This will KILL the bass as all you do is move air around the sides of the cabinet. A rear facing tweeter does work.
OK one more comment: Open baffle designes are exactly what you describe but with just one driver. The back of the cone is out of phase by 180 degrees. But note to work the baffle needs to be very large
If you have an 8 WPC SE tube amp, you really should look at some tall horn loaded speakers using an 8" full range speaker.
That's not exactly true- this loading effectively halves the Vas and the efficiency. Done correctly, it also provides superb bass without the "comb filtering" effect someone else mentioned.
smoking-amp said:
Not overly busy at all. The use of multiply nested feedback resulted in a tube amp that had remarkably low distortion that was flat across the frequency range and a similarly low source impedance that was likewise flat across the frequency range. Overload recovery was excellent. The amp was let down a bit in thermal management (board failures also mentioned above), but one cannot criticize the performance. Like the Macs, this is a design lesson for anyone interested in tube amps that are neutral and not effects boxes, but perhaps even more so.
The 3045 is a great sounding amp, I've had a pair here, compared them directly with my IT coupled GM70 SE amplifiers, surprising to say the least. I could have easily lived with them given the opportunity.
I can't speak to all McIntosh amps and am not very impressed with the current lot, however I did own a very nice pair of MC-30s for a long time in the 1980s and they are what made me finally and irrevocably jump to tube audio from solid state. I've had a lot of ST-70s as well, and with modern speakers at least (difficult loads) the MC-30 easily outclassed both stock and heavily modified ST-70.
Everyone talks about massive feedback employed in the McIntosh amps while possibly not recognizing that the unity coupled output stage had massive voltage driving requirements. I no longer remember the feedback margins, but suspect it was under 30dB if I recall correctly.. Hopefully someone can provide some illumination on this subject.
I can't speak to all McIntosh amps and am not very impressed with the current lot, however I did own a very nice pair of MC-30s for a long time in the 1980s and they are what made me finally and irrevocably jump to tube audio from solid state. I've had a lot of ST-70s as well, and with modern speakers at least (difficult loads) the MC-30 easily outclassed both stock and heavily modified ST-70.
Everyone talks about massive feedback employed in the McIntosh amps while possibly not recognizing that the unity coupled output stage had massive voltage driving requirements. I no longer remember the feedback margins, but suspect it was under 30dB if I recall correctly.. Hopefully someone can provide some illumination on this subject.
One of the interesting features of the Mac and Lux amps was the use of positive feedback (bootstrapping) in the drivers. Crowhurst's paper discusses this in detail (as well as feedback margins and how the nested loops really worked).
High negative feedback (with good stability- these were smart engineers) is responsible for the low distortion and source impedance, which can be offputting if you want the amp to have a "sound" rather than simply make the input signal bigger.
High negative feedback (with good stability- these were smart engineers) is responsible for the low distortion and source impedance, which can be offputting if you want the amp to have a "sound" rather than simply make the input signal bigger.
The CFB feedback to the driver stage in the LUX and in some MACS, I would definitely agree with. No need for the highish 50% CFB level then with the extra loop gain (and leakage L spikes can be fixed in other ways). But the Lux seems to have gotten carried away with extra windings that have the same signal on them. You don't want to put extra stuff in the OT that just makes leakage L bigger. But keeping the %CFB down to where it can be bifilar with the output secondary is a smart move.
A speaker with doubled up drivers would just share the loading and double the mass if they are driven with the same Amp signal. The inner driver needs to have its drive level increased slightly until the outer driver no longer sees back pressure.
A speaker with doubled up drivers would just share the loading and double the mass if they are driven with the same Amp signal. The inner driver needs to have its drive level increased slightly until the outer driver no longer sees back pressure.
Last edited:
Dynaudio did an interesting trick here- they used different drivers on the inside and the outside, had the inside driver fire through a Variovent, and had a choke in series with the inside driver.
"now you have two drivers, each making sound exactly as loud as the other except they are 180 out of phase. The sound from each driver will mix either constructively or destructively based on the frequency and you get a "comb filter" effect. Not good. The frequency response is full of spikes"
No, the drivers are IN phase and close together to avoid comb filter effects.
----------------------------
"Dynaudio did an interesting trick here- they used different drivers on the inside and the outside, had the inside driver fire through a Variovent, and had a choke in series with the inside driver."
The choke in series I assume was the usual electrical one. (not sure what an acoustic choke would look like) This is quite interesting. Maybe what one really wants is a current output amplifier for the inner driver. Then the outer driver can be the "voltage source" setting the sound. (or stiff sound pressure setter)
No, the drivers are IN phase and close together to avoid comb filter effects.
----------------------------
"Dynaudio did an interesting trick here- they used different drivers on the inside and the outside, had the inside driver fire through a Variovent, and had a choke in series with the inside driver."
The choke in series I assume was the usual electrical one. (not sure what an acoustic choke would look like) This is quite interesting. Maybe what one really wants is a current output amplifier for the inner driver. Then the outer driver can be the "voltage source" setting the sound. (or stiff sound pressure setter)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.