A little research: old tube amps from the past

[Giaime]

Hello everybody!

Problem is: I'm so tired about poor performance from triode amplifiers. Poor damping factors and high distortion even at low volumes are an issue. I know, a 300B PP amp is one of the best sound I've ever heard, and it also measure quite well, but 300B are too expensive.

So my interest recently moved toward the great amplifiers from the past.

I would like to collect information about them, how did they sound, focusing on what are the technical "tricks" that they used (back in the days were audio meant research).

I'm talking about Quad II, Leak, McIntosh, H/K Citation, etc etc...

What's good in them, and what can be still succesfully used those days to build good performing amps.

- triode / tetrode-pentode / ultralinear?

- what kind of negative feedback?

etc etc... and please post schematics of the amp you're talking about.

Thanks in advance to everyone.

[ErikdeBest]

I don't quite understand...I always thought (read) that triode amplifiers have higher damping factors than pentodes - that's why a PP amp with EL34 in triode can play without feedback, while EL34 in pentode will need some. That's already one point: making good use of negative feedback will increase damping factor!

Distortion is an issue with SE (which were your interest in some recent threads) but I see that the Red Light and others show good distortion figures and the employed tubes don't cost as much as 300B's. Better distortion figures may be reserved to SS terrain.

Here I am again recomending Morgan Jones Valve amplifiers third edition, where he explains the Williamson, the Quad II and a mullard design (5-20, if I am not mistaken). And after reading the problems with those circuits you will probably like his crystal palace!

[SY]

Damping factor is only one performance critereon. Besides, the damping factor without feedback is only worth looking at if you're running without feedback; one of the beauties of pentode operation is that one can achieve quite low distortion and good damping because of the high open loop gain. (It's also possible to get a very bad result if you don't know what you're doing!)

The analyses by Morgan Jones in Valve Amplifiers are excellent! One thing becomes clear, though: many engineering decision in classic amps were made for cost reasons. Others were made because of the economics and size of components in that era, which may or may not be applicable now (e.g., good active regulators were bulky and expensive in 1960, but not so today). There's a weird current in the tube community which takes the unexceptional idea that not all newer technologies are better and transmutes it into the idea that NO newer technology is better. The Crystal Palace is a perfect refutation of that silly notion, making appropriate use of LEDs, chips, bipolars, and the like to achieve performance that would have been prohibitive 40 years ago.

[jlsem]

I'm so tired about poor performance from triode amplifiers. Poor damping factors and high distortion....

This is a heretofore unknown discovery.

Seriously, the key to a good performing amp isn't whether you use a triode or pentode, but how much development time you put into the design involving both measuring and listening. I've heard all kinds of tube amps and the most highly developed (read engineered) ones almost always sound best. I personally like in this order:

Single ended DHT
Push-pull DHT
Push-pull Ultralinear

The differences are minimal in well-designed and well-developed systems and each type has its strong points. In other words, you can't expect good performance fiddling around with SPICE and throwing something together in one shot.

John

[phn]

There's no trick. Guys like SY can improve on most of those amps you mentioned. At least as far as measurements go. There probably won't be any sonic difference.

Japanese audio gurus tend to rate pre war stuff as the best ever made. And it mainly comes down to the output transformers. I might have posted this earlier. The Western Electric 111C repeat coil has a ruler flat resonse from about 28-28,000 Hz. It was designed to balance lo-fi phone lines in the 1930s. Some WE and UTC audio transformers have much better specs than that. I recently saw a single WE OT priced at $1,800 on eBay. I don't agree with the price. But I can at least understand why.

I sold a pair Telefunken pro audio step-up/balancing transformers from 1964 to a Reimyo PAT777 owner in the US as an upgrade to his Shindo step-up transformers. I was a bit worried about sending them across the pond. I was afraid they wouldn't be an improvement in his system, which is much better than mine. He convinced me, saying he was willing to take a chance. He was very happy with the transformers, saying the highs are much more open and detailed. Now, I have only owned vintage step-up transformers. But from his and other reports, I'm positively sure that you won't find a modern step-up transformer that can hold a candle to those beefy (close to 500 grams) Telefunkens.

Sorry for the rant.

[poobah]

I think there are 3 factors to the success of the old amps:

1) Iron

2) Iron

3) Iron

[jlsem]

Sorry for the rant.

Not exactly a rant. This is very good information and demonstrates that some of the better UTC and Triad signal transformers for sale on eBay are actually a good deal. Unfortunately, eBay also killed the $1.00 HA100X at the local hamfest.

John

[rdf]

Hi Giame, what caused you to arrive at that conclusion? Of the output tubes I've tested, from EL84 to the 8XX series of transmitter tubes, triode has always yielded lower distortion, much reduced high order distortion harmonics, and much better damping factor. On one occasion though, an old Mullard circuit, conversion from pentode to triode operation raised the distortion and killed the top end. The driver was marginal and couldn't handle the extra grid capacitance of a triode-connected output tube.

That's not to say pentode can't make for an excellent amp. Like SY says, the increased ca be put to good use in the feedback loop.

[rdf]

Re: iron, someone somewhere posted a comparative measurement analysis of old vs. new iron. I recall only one antique unobtanium sample which ran with the new kids with surnames Jensen and Cinemag, the rest got spanked mercilessly.

[ErikdeBest]

I have to, again, mention Morgan Jones. He measured a deteriorated performance of older iron, an example are chokes that lost inductance over years and rattling trafo's.

Erik