Vintage amplifiers that could challenge or approach current state of the art amplifiers

[Willem]

Archimago just measured a 1978 Pioneer SX880 reciever. It was still original, and in pretty decent shape although it obviously deserves a (first) service: http://archimago.blogspot.com/2020/08/retro-measure-pioneer-sx-880-receiver.html

 

[anmpr1]

There are so many of these units in everyday service, I doubt it would be possible to estimate just how populous they are.

There are a few reports on the 405 floating around. I thought it was a very smart design with good looks.

One thing, it has an input sensitivity of 0.5 V for full output, so match your preamp accordingly. I used the 33, and together they made a nice combo. I wish I still had them, but you know how that goes--the upgrade Jones.

 

[Willem]

I use a refurbished late model 405-2 as my compact desktop amplifier and it is fine. I also use a more powerful 606-2 in my main system. Again, I hear no obvious issues, but I would like to see one tested, or any of the later and largely identical descendents.
And yes, input sensitivity is much higher than is now common, so I need attenuators on the inputs. I still have a 33/303 combo, and that is also fine, but I have the distinct impression that the Q33 would now be a weak link in a revealing system. But it is a very clever and most of all versatile unit. I only recently replaced it with an RME ADI-2.

 

[Coach_Kaarlo]

Benchmark is a respectable and engineering based organization. At the same time they have to sell product to fickle audiophiles, both amateur and professional. So they sometimes straddle the fence.

Steady state measurements are good enough if the amplifier is good enough--that is, if your loudspeakers don't present a weird and unpredictable load. Transient measurements are fine, but as in the case with Bob Cordell, Marshall Leach, Matti Otala and others, you have to ask yourself whether their measurement protocols (which show up very real and measurable differences) have anything at all to do with what can be heard, above and beyond steady state measurements.

In the Tom Holman Advent receiver thread we discussed his (and others') measurements on dynamic distortions as they might concern audible differences in preamplifier circuits. Peter Aczel measured a handful of preamps (two dozen) and found absolutely no correlation among those esoteric tests and how they 'sounded' in subjective auditions. Not long afterwards, Holman himself walked away from his dynamic protocols, deciding it wasn't important. The point is, clever practitioners can always find some or another 'test' to measure differences in circuits. That is not the point. The point is, does any of it make any difference?

FWIW Peter told me that all those 'audible differences' back in 1977 were likely due to his sloppy protocol of not matching levels. I argued that given his front and back end (expensive MC cartridges on expensive turntables and electrostatic loudspeakers) that possibly there were differences. He was not impressed with my argument, and acted as if I hadn't learned anything at all. LOL

Until anyone can prove otherwise I have to subscribe to what he wrote 30 years after his big amp shoot outs:

Any amplifier, regardless of topology, can be treated as a “black box” for the purpose of listening comparisons. If amplifiers A and B both have flat frequency response, low noise floor, reasonably low distortion, high input impedance, low output impedance, and are not clipped, they will be indistinguishable in sound at matched levels no matter what’s inside them.

When they measure the same amp with different speakers, and then repeat the test with other amps and the same varied set of speakers we might have something to discuss. Otherwise it seems anecdotal, and goes against your very valid point earlier; room + speaker + amp = system. Therefore it seems by only testing amps, they were eliminating the key variables in the system which make the greatest difference. This is actually not doing science or engineering IMHO.

 

[Pluto]

I also use a more powerful 606-2 in my main system… And yes, input sensitivity is much higher than is now common, so I need attenuators on the inputs

One thing Quad kept very quiet about – I really don't know why – was that their “Ampbus” link (or whatever they called it) that featured on the 99 series and later units is little more than standard level balanced signals crunched-up on a DB-15 connector. It is possible to convert this to standard balanced XLR with a simple-to-make adaptor. So why they attempted to disguise this as some proprietary bus really beats me…

 

[Willem]

Thanks for that, Pluto. I suspected as much but since my older 606-2 does not have this I never looked into this any further. I suppose it was to suggest you really needed to buy the combination. Probably for the same reason mine does not have a power switch on the front but a power input from the preamp. Using my ADI-2 as a preamp I solved that problem with a master slave power strip.

 

[orangejello]

If you are ever in Sydney (COVID etc) I can connect you with the owner for a listening session in his studio. We tested for 6 hours per session, on 2 occasions (level matched listening and cone excursion but nothing more serious - just focusing on specific musical passages a few seconds long and repeating them). I was interested how it performed against my 2 Benchmark Media AHB2's in mono (comparable output).

The AHB2's had great imaging and clarity - but nothing close on the dynamics. There are a whole string of unanswered questions in my mind regarding measuring the AHB2 dynamically to understand why they sound so different dynamically and yet measure so well statically. But let's just say if the owner had been interested in selling I would have purchased speakers AND an amp that day.

I had the exact same impressions comparing the NAD M22 v2 to the AHB2. The difference in dynamics was stark. The AHB2 sounds dead by comparison. I was as puzzled as you seemed to have been. It took me awhile to get over the stellar static specs of the AHB2 and the prospect of owning the best measuring amp on the planet. But in the end I sent the Benchmark packing.

I would suggest that you try the NAD (or Purifi or NCore equivalents) and see if it doesn’t match the dynamics that you experienced without losing the clarity. Granted you wont have a vintage behemoth that is build like the proverbial brick s*hthouse. But the subjective performance might be similar With the edge going to the class d because of better noise nad distortion specs.

 

[sergeauckland]

One thing Quad kept very quiet about – I really don't know why – was that their “Ampbus” link (or whatever they called it) that featured on the 99 series and later units is little more than standard level balanced signals crunched-up on a DB-15 connector. It is possible to convert this to standard balanced XLR with a simple-to-make adaptor. So why they attempted to disguise this as some proprietary bus really beats me…

I seem to recall that there was either a Service Note, or it could even have been part of the user manual, but they had the pin-out of the DB-15 where it was clearly marked where the L+- and R+- and ground pins were so anyone who understood that could make up their own connectors.

Quite a lot of pro equipment has the I/O on DB 9, 15 or 25 instead of XLRs both for reasons of economy, (XLRs aren't needed for permanent installation) and for space reasons as it's hard to get more than a dozen connectors on a 1U box.

S.

 

[Pluto]

seem to recall that there was either a Service Note, or it could even have been part of the user manual, but they had the pin-out of the DB-15 where it was clearly marked where the L+- and R+- and ground pins were so anyone who understood that could make up their own connectors

I have this detail for anybody who needs…

 

[restorer-john]

Until anyone can prove otherwise I have to subscribe to what he wrote 30 years after his big amp shoot outs:

Any amplifier, regardless of topology, can be treated as a “black box” for the purpose of listening comparisons. If amplifiers A and B both have flat frequency response, low noise floor, reasonably low distortion, high input impedance, low output impedance, and are not clipped, they will be indistinguishable in sound at matched levels no matter what’s inside them.

I would definitely add measureable dynamic behaviour into that quote.

The other issue with the above statement is lack of specifics. It's simply too vague. A 'flat' frequency response is exactly what? An allowable deviation across the audible bandwidth needs to be specified simply because one person's 'flat response' is another person's 'poor response'. In the 1970s, frequency response plots were printed in HiFi magazines with a 50dB range. Everything looked pretty darned flat.

So when is flat, actually flat?

This is a Rotel RA-930AX. Lovely little amplifier, but all like all Rotels, it's rolled off a bit at each end of the spectrum. I don't regard it as remotely flat enough to become a so-called "black box" for amplifier comparisons. It is a perfectly respectable -0.4dB at 20Hz and -0.2dB at 20KHz. Consider the Hypex modules are off by ~0.5dB at 20KHz and people rave about them...

+/-5dB (note there are 493 data points in this log sweep from 20-20K of 8bit resolution- 256 levels)

Same amplifier, same data at +/-3dB

Now at +/-2dB unsmoothed data

What is a "low noise floor" and a what point can the noise floor be considered good enough to perform a "black box" comparison without clues from hum and noise clouding the result? Any noise not present in one amplifier and present in another renders such comparisons useless otherwise.

Same for distortion and clipping. How much clipping? 0.01% of the time, never, or 1% of the time? We know a 30wpc amplifier will be clipping more often on transients at a decent volume than a 30wpc amplifier with 6dB of dynamic headroom.

 

[anmpr1]

What is a "low noise floor" and a what point can the noise floor be considered good enough to perform a "black box" comparison... Same for distortion and clipping. How much clipping?

All your points show that things are not always 'one size fits all'.

Generally the noise should be less than the ambient room noise. With some loudspeakers and in some living rooms it will have to be much less. With highly sensitive loudspeakers a noisy amp could be most all amps--at least in the hiss department. At the Benchmark site there's a user comment about someone with Avantgarde Acoustic loudspeakers in his living room. He stated that the AHB was the only amp he tried that did not demonstrate audible hiss with his loudspeakers at his listening position. I use relatively efficient horns, but am not bothered with hiss by any of my amps--probably due to my HF hearing loss (age related). So noise floor is definitely a variable thing--that is, it varies among listeners.

I think any clipping ought to be avoided, and mostly can be avoided given the watts/dollar ratio of gear, these days. However, if one's idea of good sound is a 7 watt SET, it might not be possible to avoid, but then some argue that tube clipping is 'softer' and less unpleasant, so that could be a mitigation in context.

I recall the Stereo Review article on the Phase Linear 700, which was one of the first reasonably priced super high powered amps. Compared with the Crown DC-300 (used with AR LST loudspeakers) the SR panel frequently clipped the Phase, and they didn't care or notice it as much as with the Crown, because at the same SPL the Crown pooped out and went harsh, whereas clipping on the much more powerful Phase Linear was shorter in duration and less bothersome at high SPL.

 

[sergeauckland]

I would definitely add measureable dynamic behaviour into that quote.

The other issue with the above statement is lack of specifics. It's simply too vague. A 'flat' frequency response is exactly what? An allowable deviation across the audible bandwidth needs to be specified simply because one person's 'flat response' is another person's 'poor response'. In the 1970s, frequency response plots were printed in HiFi magazines with a 50dB range. Everything looked pretty darned flat.

So when is flat, actually flat?


What is a "low noise floor" and a what point can the noise floor be considered good enough to perform a "black box" comparison without clues from hum and noise clouding the result? Any noise not present in one amplifier and present in another renders such comparisons useless otherwise.

Same for distortion and clipping. How much clipping? 0.01% of the time, never, or 1% of the time? We know a 30wpc amplifier will be clipping more often on transients at a decent volume than a 30wpc amplifier with 6dB of dynamic headroom.

Flat used to be +-3dB, but I accept that these days there's no excuse for 'flat' to be any worse than 20Hz-20kHz +-1dB, and only at the frequency extremes, not anywhere above, say 50Hz and below 15khz. I don't thank anyone other than perhaps a very acute 18 year old could hear -1dB at 20khz.

As to noise, if noise is imperceptible at the listening position it's low enough. That can easily be tested by switching the loudspeakers on and off and in a blind test checking if anyone can tell when on or off with no audio playing.

As to clipping, no, it should never be allowed at all, and in the case of valve amplifiers, the maximum voltage should never exceed their rated output. Clipping is not just a simple squaring-off of the waveform. many amplifiers do all sorts of strange things when clipping, and that should be no part of a test for audibility. In the case of valve amplifiers, some don't clip square, but progressively distort, but one I've measured actually went unstable and the waveform went suddenly from, say a few % distortion, to over 100%, where the fundamental disappeared and it's all distortion. Very odd when it happened, but both channels were consistent, so I took it as part of the design, even if presumably not intentional.

As to dynamic headroom, that's an odd one, as I recall one magazine (can't now remember which!) publishing some tests that established that two amplifiers with identical continuous power outputs, but one with a stabilised supply, so no dynamic headroom, the one with the conventional sagging supply sounded better due to higher dynamic power, but when the stabilised amp had it's power supply raised so that its continuous power was the same as the other's dynamic power, then the stabilised amp sounded better.

Those tests although carried out blind, did allow the volume level to be set by the listeners, so clipping was a distinct possibility, and indeed was thought to be the reason behind the audible differences, as the amps were otherwise identical other than power supply stabilisation.

S

 

[Shanghai Wally]

All these pages and no mention of Yamaha B2 VFET? Yes, it definitely challenges modern amps in sound quality, it is hypnotic to put a word on it.

My Sample, all original (except for VU meter LED), sounds amazing:
View attachment 71443

I have a Yamaha MX 600u, too, that is also all original, powers some surrounds speakers. It sounds OK, but nothing approaching the B2.

Aloha all
agree 100percent with your words
how about the B-3 they doin a gorgeous job in BTL modus all I can say The JBLs and my Ears like it alot just right no need more for the moment
Cheers and stay safe

 

[cjfrbw]

Aloha all
agree 100percent with your words
how about the B-3 they doin a gorgeous job in BTL modus all I can say The JBLs and my Ears like it alot just right no need more for the moment
Cheers and stay safe

View attachment 88469

Yup, B3 are Yamaha VFET amps, like the B2. Still one of the best sounding production amps around. My sample is hypnotic to listen to.

 

[Sal1950]

The JBLs and my Ears like it alot just right no need more for the moment

Of course not, 140 WPC in bridged mode is nuclear power level for those JBL's.
Sweet pair of amps.
I had to google for a bit more info on them and noticed strangely they have input level controls on both the front and rear panels. very unusual.
Looks like they have a ton of heat sink available and I see copper transistor mounting chassis inside, top shelf construction it appears. Enjoy!
https://audiokarma.org/forums/index.php?threads/yamaha-b-3-v-fet-power-amp.42090/

 

[wemist01]

I run an Arcam Alpha 10 power amp. Someone tell me what I've got. Vintage measurements said it was worthy.

 

[Shanghai Wally]

Victor M-7070 in BTL this pair of Amp's is bulletproof covered in timeless design
Sounds very good too with adequate Ls
Never serviced and would be interesting test them the way Amir test the
Yamaha A-1 which brought me actually to this cool ASC website
Cheers from Hamburg stay safe

Inside

 

[precisionav]

Hafler amps have been mentioned earlier in this thread. They get my vote for cost to performance ratio. I purchased my DH 500 twenty years ago used for 250.00. I performed some Musical Concepts upgrades for about the same amount. At my previous job I was able to AB this amp with some from Mac, Krell, Classe, Bryston and a few more. The Hafler held its own with every one of them at a fraction of the price. I believe an amp should do nothing more than amplify the incoming signal faithfully. Nothing added, nothing taken away and my Hafler does this very well. I want what I hear coming from my speakers to be as close to the original recording as possible, what the producer and artist intended. When I close my eyes I want to believe as though the music is being performed live right there in my listening room. Well designed, engineered and executed audio equipment can and should do this and it doesn't have to cost a fortune.

 

[Doodski]

I was able to AB this amp with some from Mac, Krell, Classe, Bryston and a few more. The Hafler held its own with every one of them at a fraction of the price. I believe an amp should do nothing more than amplify the incoming signal faithfully. Nothing added, nothing taken away and my Hafler does this very well.

That's exactly what a good amp is like up to it's clipping point or point where distortion starts.

 

[yodog]

Epic specs. What year was the hafler released?