Solid state/class d vs tube amps

[Chucky Stiletti]

Interestingly a pure tube gain stage versus pure ss gain stage can't be detected by hearing test within headroom level limits each.

I agree. If both pieces of gear look the same through an audio test set such as an Audio Precision, logic would imply that there should be no difference to the human ear. This said, you would be blown away by the Bay Area's "Golden Ear Fraternity's" claims that there ARE differences. I do not know if rehearsed (they had to be), but I have seen some very convincing blind listening tests some extending to the passive component level.

 

[seamonkey]

I agree. If both pieces of gear look the same through an audio test set such as an Audio Precision, logic would imply that there should be no difference to the human ear. This said, you would be blown away by the Bay Area's "Golden Ear Fraternity's" claims that there ARE differences. I do not know if rehearsed (they had to be), but I have seen some very convincing blind listening tests some extending to the passive component level.

References or Links to those DBTs?
In the modeling world, they do model down to passive component level. And some then actually measure the real passive device. The models are accurate, and there just isn't any audible difference that can't be measured and accounted for elsewhere. Even placement and trace/wire routing is considered. Some differences may arise of physical layout.
Thanks!

 

[Rich Fiscus]

With digital (synthetic)

You keep using that word (synthetic)
I do not think it means what you think it means.

Unless there are record bushes or magnetic tape trees out there that I haven't heard of.

 

[bumble yeti]

Sparkl, my friend, what have you done? Your naïveté has somehow 'sparked' the hi-fi debate of yore.
This fire had been raging for years when I'd gotten into hi-fi/home theater back in the 90's.
Alas, all you can do is bask in the glow, roast marshmallows and wait for it to burn out.
I haven't missed a word.
(Psst… Hey guys, I don't think we've heard from the OP in awhile. Just sayin.')

 

[Chucky Stiletti]

References or Links to those DBTs?
In the modeling world, they do model down to passive component level. And some then actually measure the real passive device. The models are accurate, and there just isn't any audible difference that can't be measured and accounted for elsewhere. Even placement and trace/wire routing is considered. Some differences may arise of physical layout.
Thanks!

Yes, pretty familiar with the modeling world and with Cadence and ADS for circuit simulation. Good thing most component manufacturers provide S parameters that can be imported into the model/sim, these days. Makes things faster. I have used HFSS for physical layout and magnetic sims. Identifying/Quantifying coupling mechanisms and other non intentional three dimensional phenomena is next to impossible without these tools.
I have attached a few links as per your request;

Spotify is testing lossless audio. Can you hear the difference?

[Invalid or Expired Link Removed]


A Brief Guide to Audio for the Skeptical Consumer


I used to see these tests at the Stereophile Convention at the St Francis Hotel in San Francisco back in the 90's where manufacturers would literally rent the rooms and present their products for three days. Manufacturers of electrolytic capacitors such as V-Cap, Ampohm, Supercap and others would slug it out in these tests.
My point is that for two completely different amplifiers (as far as topology, class, tube/ss, etc.) with the EXACT measured audio spec to sound different to an individual something else must be going on. Don't know quite what this is but my guess is predisposition and bias (excuse my pun).

 

[Chucky Stiletti]

You keep using that word (synthetic)
I do not think it means what you think it means.

Unless there are record bushes or magnetic tape trees out there that I haven't heard of.

I used the word synthetic because it was being used earlier in the thread..

 

[seamonkey]

Yes, pretty familiar with the modeling world and with Cadence and ADS for circuit simulation. Good thing most component manufacturers provide S parameters that can be imported into the model/sim, these days. Makes things faster. I have used HFSS for physical layout and magnetic sims. Identifying/Quantifying coupling mechanisms and other non intentional three dimensional phenomena is next to impossible without these tools.
I have attached a few links as per your request;

Spotify is testing lossless audio. Can you hear the difference?

[Invalid or Expired Link Removed]


A Brief Guide to Audio for the Skeptical Consumer


I used to see these tests at the Stereophile Convention at the St Francis Hotel in San Francisco back in the 90's where manufacturers would literally rent the rooms and present their products for three days. Manufacturers of electrolytic capacitors such as V-Cap, Ampohm, Supercap and others would slug it out in these tests.
My point is that for two completely different amplifiers (as far as topology, class, tube/ss, etc.) with the EXACT measured audio spec to sound different to an individual something else must be going on. Don't know quite what this is but my guess is predisposition and bias (excuse my pun).

Yep - that's all I've seen so far is that every DBT published even have conclusions that listeners, even trained musicians, don't hear a difference in amplifiers playing in the linear range.
With high powered class-d amps these days that linear range is large.
Which also means listeners don't hear anything superior with any class of amp or devices.
Still some "head" makers actually color their tone ahead of the natural linear range of the amp. Getting a controlled "signature" sound. Power amps, or the amp part of most all heads are just power amplifiers.

Around here there's a lot of focus on the edge case of pushing an amp outside it's natural linear range. Most music doesn't go there, and if does, it is usually at the choice of the musician with the signal processing they use (pre-amp, pedals, etc)

 

[DavesnothereCA]

....Still some "head" makers actually color their tone ahead of the natural linear range of the amp, getting a controlled "signature" sound....

~
Or they put different amounts of (clean) preamp gain and/or odd tapers of volume control pots, leading to the old silly observation of only needing to turn this amp up to 2 to get the volume (to the ear) which that amp delivers when set to 4.
~

 

[AE1]

Well, it looks like proof is not in your favor here. Those amps are CLEARLY MC2300's.

Yes you are correct. I was wrong. Thank your for the correction.

The front panel of the MC2300 SS amp is very similar to, but not identical with, the MI350 tube amp. I based my comments on Bob Weir's reference in the video I posted to "Mac350's" - I assumed he meant "MI350's".

Did he mean Mac 2500's ? Did he mean 600 instead of 350 ? Is he wrong ?

It might be helpful to understand the history of this because only then can you appreciate how absurd your chasing this rat down the hole really is.

I presume "chasing this rat down the hole" means advocating the use of tube amps ?

There are also numerous website references and documentaries to the Grateful Dead's purchase of a stack of MI350 tube amps for Woodstock - long before you started your career in 1978

It seems the band used both types of amps in the references posted by D J Bebop and to which I responded

The concept for the wall of sound was hatched in 1972, the year that the MC2300 was introduced into production. Production volumes were very low, it was brand new technology for McIntosh and it was very expensive. There was ample supply of the tube amps though, which is what they started with, but they turned out to be entirely unacceptable, not enough power for what they were trying to achieve. Once the MC2300's were available at sufficient quantity, the tube amps were immediately swapped out, with a few remaining for various applications where they were more suitable.

In fact the industrial MI350 and its domestic brother MC3500 ceased production in 1971 so would not have been available to the band even if they wanted them. ref http://www.roger-russell.com/amplif2.htm

These typical posts at Grateful Dead's 1966 McIntosh set-up say:

"The Dead also bought some McIntosh amps back in 1973. Their equipment hadn't all made it to the Watkin's Glen site, so (if ,memory serves...) they grabbed a copter to the McIntosh plant in Binghampton, NY. They pulled amps off the line that were not 100% completed with lettering/numbering and threw downs wads of cash, then off to the show. Look behind Jerry in pictures from any year 1973 and after to see his McIntosh amp.

Is that when they went to the MC-2300 amps, bridged into 600 watts mono? I've seen those live and in films. Those things weight about 130 lbs. each! They look great on stage, the blue meters and all.

Separately, McIntosh made a high power vacuum tube amp, the MC-3500 or MI-350 for the "industrial" version. 350 watts of vintage McIntosh tube power. Very, very rare. I've been informed there were 10 of these monoblock amps under the stage at Woodstock!"

"The Grateful Dead used 14 McIntosh MC3500's in the late 60's and early 70's. These mono babies put out 350 tube watts ea. for a total of 4,900 tube watts per show."

they (the MI350) turned out to be entirely unacceptable, not enough power for what they were trying to achieve.

If we set aside the accuracy of some of these recollections, then assuming the same speaker system and ignoring the difference in coupling efficiency, 9,800 watts of SS power v 4900 watts of tube power would produce only a 3 db increase in the total SPL. At say 130 db SPL a 3db difference would be hardly noticeable.

As to tube v ss the MI350 and its sister MC3500 amps currently fetch at least ten grand a piece despite being nearly 50 years old. Must be something about them more than folklore that buyers prefer over SS.

There is a place for both in this world - it is big enough to handle it and has enough buyers to support each genre.

 

[agedhorse]

My experience was as a McIntosh warranty tech starting in 1979.

Back in the very early 1970s, solid state amps were very limited in availability, performance and were very expensive. If you wanted higher powered amps, you had little choice but to deal with the cost and weight of tube amps. The cinema industry was a good example of this, though solid state was just beginning to make inroads with amps like the Ampex 4021063 amplifier, 80 watts RMS into 16, 8 or 4 ohm. Altec and then BGW and Crown offered similar amps around this time, but the larger ones didn't show up in any real quantities until around 1973-74.

Tube amps were a necessity, we as an industry were begging for higher powered solid state amps that were also reliable. There were some notoriously unreliable early solid state offerings, but they were avoided like the plague.

 

[ThisBass]

If we set aside the accuracy of some of these recollections, then assuming the same speaker system and ignoring the difference in coupling efficiency, 9,800 watts of SS power v 4900 watts of tube power would produce only a 3 db increase in the total SPL. At say 130 db SPL a 3db difference would be hardly noticeable.

Another +3dB of headroom is a good bet to run the amplifiers within linear dynamic range.
Soundengineers more often think in headroom rather than total gainable loudness.
We are (alomst always) more interested in improved sound quality rather then total loudness FOH wise when we do double up amp power

 

[agedhorse]

If you don't have enough SPL, the extra 3dB is a goldmine of resources.

 

[AE1]

My experience was as a McIntosh warranty tech starting in 1979.

You were very fortunate to start your career with an industry leader. McIntosh is a hard act to follow.

I wish I could have done the same but was not to be.

Here is a wonderful shot of the MC2300's - I can see 13 on stage

Jerry Garcia's Middle Finger: Maria Muldaur: July 21, 1974, Hollywood Bowl, Hollywood, CA (click to expand the pic to full size)

According to this report the band was still using them on stage up to at least 1992 - Garcia Guitars & Gear Directory - Twin Preamp Mods

 

[agedhorse]

There was a lot of mixing and matching and experimenting going on over the years. Many of the technical folks involved with The Dead, JGB, and Old & In The Way, New Riders of the Purple Sage and even DGQ (both NRPS and DGQ were acts that I worked with regularly) went on to start or become closely associated with companies in this industry and ironically, many ended up getting (or already had) their engineering or physics degrees through all of this. Companies like Furman, Alembic, Meyer Sound, UltraSound, Gamble Associates (the reason for the link to Crest Audio), and others all came out of their association with the Dead.

Sometimes products are used simply because they were available. Nothing more should be read into the use of a product.

 

[Chucky Stiletti]

My experience was as a McIntosh warranty tech starting in 1979.

Back in the very early 1970s, solid state amps were very limited in availability, performance and were very expensive. If you wanted higher powered amps, you had little choice but to deal with the cost and weight of tube amps. The cinema industry was a good example of this, though solid state was just beginning to make inroads with amps like the Ampex 4021063 amplifier, 80 watts RMS into 16, 8 or 4 ohm. Altec and then BGW and Crown offered similar amps around this time, but the larger ones didn't show up in any real quantities until around 1973-74.

Tube amps were a necessity, we as an industry were begging for higher powered solid state amps that were also reliable. There were some notoriously unreliable early solid state offerings, but they were avoided like the plague.

Yeah, remember Phase Linears and Adcom's?..

 

[agedhorse]

I do. Phase Linear warranty service paid for quite a bit of my engineering education. They were an example of the struggles early high powered solid state amps faced with regard to reliability.

I wonder if all my vintage service manuals will ever be worth something?

 

[Chucky Stiletti]

I do. Phase Linear warranty service paid for quite a bit of my engineering education. They were an example of the struggles early high powered solid state amps faced with regard to reliability.

I wonder if all my vintage service manuals will ever be worth something?

Did you ever meet Bob Carver? I like his tube design. Early SS on the other hand...
There was a story going around back in the early 80's about a supertech (cannot recall the name) that flew to a ZZ Top gig with a bunch of output transistors to repair a whole bank of PL's that had cooked and managed to get the show started on time. Do you recall anything like that?

 

[agedhorse]

I did meet Bob Carver. He is a very, very talented engineer who was responsible for some truly novel designs that were way ahead of the time. This also was responsible for the tweaky, temperamental nature of many of his products and the associated failure rates that (often) go along with expanding the envelope of conceptual design.

There are so many stories, I don't know how accurate many of them are. Rarely were the failures just limited to the output transistors in any of the PL-400 and 700 models, so that causes me to doubt it.

 

[AE1]

My point is that for two completely different amplifiers (as far as topology, class, tube/ss, etc.) with the EXACT measured audio spec to sound different to an individual something else must be going on. Don't know quite what this is but my guess is predisposition and bias (excuse my pun).

Depends upon how you determine what "EXACT" is

Nowadays most quality amplifiers claim distortion and frequency response specs which are well beyond human hearing or awareness

There are several measures which could be used but not usually quoted, such as frequency response, distortion, stability and dynamic response at rated power into the various rated loads - eg 2, 4, 8 and 16 Ohms - but particularly in the normal frequency dependent speaker impedance range of 40 to 80 Ohms

Linear gain from zero to max is another issue. Obviously power supply sag will reduce gain in the upper power region, affecting not only max continuous power but also dynamic peak power.

All of the above plus more can be measured to give the designer much information about how well the amplifier will perform

As to how will it sound - to my mind that is a consequence of numerous cause and effect relationships within the design, as well as interactive elements such as negative feedback and speaker impedance characterictics v frequency. Negative feedback from the output terminals ensures speaker characteristics are fed back into the amplifier, sometimes creating instability under certain but maybe random conditions.

The underchassis layout has interactive effect via inductive and capacitive coupling. Components such as resistors and capacitors also display inductive properties. Internal shielding and proximity effects play a big role.

The number of discrete stages in the system is also relevant, as is the conceptual topology of each stage

The use of integrated circuits also confuses comparisons.

So what might appear to be the same will most likely be very different

Tube amplifiers appear to be more tolerant of varying load whereas typical SS amplifiers are very sensitive to load - often not in a linear relationship as load changes with frequency

In my experience a solid state amplifier supplying a nominal 2 Ohm load will deliver more power and more solid LF bass than a nominal 8 ohm load simply because the maximum load presented by the speaker to the SS amplifier is proportionately less.

With reference to the above posts re the Grateful Dead, both the McIntosh MI350 tube amplifer and the MC2300 SS amplifier have inbuilt output transformers, allowing the load to be accurately matched to the amplifier

The MC2300 SS amplifier has capability to match loads of between 0.25 Ohms and 16 Ohms nominal - quite a feat - ensuring maximum power is available at any load within the rated range. The actual load on this amplifier when correctly matched is less than 2 Ohms, thereby achieving maximum performance for bass

If amplifiers DID NOT sound different to an individual then we could all have the same make and model for the same result. That would eliminate "difference" and put a big dent in the number of manufacturers.

Is worth thinking about.

 

[agedhorse]

Solid state amps in general are less sensitive to load variations because the source impedance is usually order of magnitude lower.

Delivering greater power into a lower impedance load doesn't mean it's more sensitive, it's purely a scaling thing that is a side benefit to very low output impedance.