Suggestions for Tube Amplifier

[Rundmaus]

Hi,

this is my first post in the Diyaudio forum, and I'd like to ask you for a little brainstorming . I'm not totally new in this hobby, already built the Zen Amp an the Bride of Zen preamp, and a pair of transmission line speakers. But my experiences with tube technology are up to now only repairing vintage radios from the fifties or so.

A few weeks ago I got hold of the AudioValve PCL805 headphone amplifier, and now i seem to be infected with the tube virus ...

Well, here are my rough ideas, i would like to hear your suggestions, if possible accompanied by schematics ...

- PP-Design operating in Class A

- Triode output stage

- "common" tubes, please no "exotics" which are difficult to replace

- 10 to 20 watts output power per channel

- "common" output transformer, something i can buy instead of building it myself

- and, i would like to do completely without semiconductors, so i need a rectifier tube vor the anode voltage, and AC heating for all tubes

Well, i'm looking forward to your suggestions, and sorry for my English

Greetings,
Andreas

 

[Bas Horneman]

Something PP EL34 or PP EL84 comes to mind.
Here is a good start for a schematic.

An externally hosted image should be here but it was not working when we last tested it.

sorry...has semiconductors in there...but....


Plenty of choice for standard output transformers.

What preamplifier do you have? (Bride of Zen? what is the gain of that preamp?) What speakers do you have? How big is your room?

Regards,
Bas


ps..your english is perfect.

 

[Bas Horneman]

Another one

An externally hosted image should be here but it was not working when we last tested it.

http://audiotropic.netfirms.com/Projects/mmAudio.html

 

[Rundmaus]

Hi Bas,

no separate Preamp, so a preamp stage needs to be included, in german its called a "Vollverstärker" but I can't remember the english word for this.

Speakers are selfmade transmission lines, should make about 90dB...

As stated before, I have absolutely no experience in designing tube circuits, up to now only repaired or upgraded existing ones... So I would be glad if, after some discussion, this thread might give some reliable, practically tested schematics. I think i know the basics about tubes and operation modes, but I shurely do not have the experience to turn a "standard layout" or principle into a well working amplifier. Just don't know the "dirty tricks" ...

One thing I forgot in my list above:

- I don't want to say that I dislike the "typical" tube sound, absolutely not! But in my opinion an amplifier should also have measurable qualities. A frequency response of about 25Hz-20kHz or better should be possible, and THD should be reasonably low...


Many expectations, I know, hope it isn't tooo much

Greetings
Andreas

 

[Bas Horneman]

Oeps ....both schematics I posted need a seperate preamp

[EDIT]...looks like Poinz's amp need not have a seperate preamp.

Many expectations, I know, hope it isn't tooo much

Not at all too much....very do-able.

 

[Brett]

Hi Bas,

no separate Preamp, so a preamp stage needs to be included, in german its called a "Vollverstärker" but I can't remember the english word for this.

Speakers are selfmade transmission lines, should make about 90dB...

As stated before, I have absolutely no experience in designing tube circuits, up to now only repaired or upgraded existing ones... So I would be glad if, after some discussion, this thread might give some reliable, practically tested schematics. I think i know the basics about tubes and operation modes, but I shurely do not have the experience to turn a "standard layout" or principle into a well working amplifier. Just don't know the "dirty tricks" ...

One thing I forgot in my list above:

- I don't want to say that I dislike the "typical" tube sound, absolutely not! But in my opinion an amplifier should also have measurable qualities. A frequency response of about 25Hz-20kHz or better should be possible, and THD should be reasonably low...


Many expectations, I know, hope it isn't tooo much

Greetings
Andreas

I've built both circuits Bas has linked, and a few variations on each. The Wright PP1C is the better sounding of the two and is in your 10-20W spec. Mine are in storage, but iirc they were driven quite well with a standard CDP output even into less sensitive speakers then my horns. You could always build them and add a pre later, if you think it's neccessary. Use the specified Lundahl iron for them and a good PS as the cascode front end has a poorish PSRR.

Been thinking about a 6550 version (with proper chasis' this time) to drive my KEF 104/2's. I have all the parts here, but little time at the moment....

 

[Jaap]

ecl82 ?

For a first start a ecl82 push pull amp could be a good choice. But I suppose the Allen Wright amp is soundswize better.

Question: Why the Lundahl iron ?
Suggestion: the AW amp with 807 tubes.

 

[Jaap]

el84PP

Your countryman Huber had an interesting EL84 on the web (now disappeared (?)).
Could anybody comment on this design ?
Did anybody build or heard it ?

 

[Brett]

Re: ecl82 ?

Question: Why the Lundahl iron ?

Because it's good, available, sensibly priced and what the designer recommends.

Suggestion: the AW amp with 807 tubes.

Probably be quite nice, but won't get 10+W in class A

 

[Jaap]

Lundahl ?

I thought that your suggestion had to do with something Allen Wright said on the tube asylum:

"In my experience, P-P amps need the standing current in the OP tubes to be VERY closely matched - or you will have a DC component in the OPT that really limits the fine detail µdynamics of the whole amp and causes it to sound closed down and lost.

Some OPTs are more sensitive to this than others: toroidals are the worst and cannot accept more than a mA before losing quality bigtime - 'normal' (non airgapped) P-P OPTs should be balanced to a mA if possible to be at their best, but Lundahls (which have a airgap even in their P-P OPTs) can be off a couple of mA without any noticiable loss."

It is not a secret that Allen has his PP-1c amp updated with a ccs under the cathodes of the outputtubes. In this way the outputstage is perfectly balanced and I think there is not a specific need for a Lundahl. You could easly use a famous vintage opt I suppose, or a cheaper one from Asia, or whatever quality-opt from own neighbourhood.

 

[Rundmaus]

Hm,

first let me thank you for that big bag of answers ... Great response!

So, in return here's my new bag of thoughts and ideas:

- the Wright PP1C sounds interesting to me, Bas do you have more info on it? Couldn't find anything on the forum... Especially performance (frequency response, THD, output power) and the second part of the schematics, containing the constant current source...

- somehow i feel that this ccs will be using semiconductors ... is it possible to design a tube based ccs instead? (at ANY cost )

- Brett are you sure sensitivity will be high enough to drive the circuit directly via line level source and volume pot? How difficult would it be to add an appropriate preamp stage?

So much for now, thanks in advance for your ideas

Greetings
Andreas

EDIT: as soon as the idea becomes more detailed, I'll start a new thread on the planned amplifier, i think I'll have enough questions for more than one thread... css, preamp, ps, soft start for tube heating, delayed switching on of anode voltages etc...

 

[Bas Horneman]

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

 

[Johan Potgieter]

Oh dear. . . .

I would like to respectfully say here that we must be careful before we take observations such as this by Allen as dogma. I do not know the gentleman and will have no hesitation in affording him the greatest of respect, but this observation is contradicted by the experience of others, as well as basic magnetic principles. Why would this total absence of dc magnetisation be so mandatory? At the other end of the scale we need to note that the S.E.T. fans swear as vehemently to the quality of their products, which carry the full output tube dc! Also, in the famous Quad II amplifier, the effect of an unbalance of output tubes of up to 25% was not audible. (Admittedly this was a special circuit, but still.)

If Lundahl has a gap in their p.p. output transformer core it would linearise the transformer - but also decrease the primary inductance needed for good low frequency response, or vice versa, necessitate that much bigger a component. According to basic magnetic theory any dc magnetisation makes just that little less flux available for signal use, to put it simply. This would affect the low frequency response to that degree, but have no effect on higher frequencies. (In the basic transformer equation flux and frequency are simply inversely related.)

This may not be the place to expound on Allen's explanation of "quality lost big-time" (if there is a reference available I would like to read it; Jaap, can you help?) but just to caution that observations can and have been subjective. This loss of quality due to a few mA of unbalanced dc was certainly not my experience over decades of tube amplifier design of all sorts, neither have I come across this before in the wide experience of others. Output transformers have been designed with distortion significantly lower than that of the output tubes, so where is the total picture?

 

[Brett]

I thought that your suggestion had to do with something Allen Wright said on the tube asylum:

"In my experience, P-P amps need the standing current in the OP tubes to be VERY closely matched - or you will have a DC component in the OPT that really limits the fine detail µdynamics of the whole amp and causes it to sound closed down and lost.

I agree with the general sentiment.

Some OPTs are more sensitive to this than others: toroidals are the worst and cannot accept more than a mA before losing quality bigtime - 'normal' (non airgapped) P-P OPTs should be balanced to a mA if possible to be at their best, but Lundahls (which have a airgap even in their P-P OPTs) can be off a couple of mA without any noticiable loss."

Generally true re the toroidals, but not always so. I use a 30W rated mains toroaid in a guitar amp as an OPT and it will take as much as 8mA before distortion goes up too much (in hifi terms not MI amp). The LL1620's will also take about 6-8mA, the smaller ones a bit less. You can also order them with no gap if you choose.
Plitron/Amplimo PP OPT's have gaps too iirc.

It is not a secret that Allen has his PP-1c amp updated with a ccs under the cathodes of the outputtubes. In this way the outputstage is perfectly balanced and I think there is not a specific need for a Lundahl. You could easly use a famous vintage opt I suppose, or a cheaper one from Asia, or whatever quality-opt from own neighbourhood.

I've also built the PP2C, and had the schematic very early on (and posted about it here a number of times). Allen asked that I never publish it* so I never have, even though it was given to quite a few other people since. No idea how what I have varies from what he now sells in his 300B amps.

I recommended the Lundahl iron for a number of reasons, as stated above. I have a large number of OPTs, both new and vintage from a number of suppliers including a number of vintage types (I moved house recently, so I know I have >250kg of them) and ultimately could care less what I use. I did try a number of different OPTs in the PP1C and PP2C I've built, and kept coming back to, guess which one?

In the original post, this was stated "- "common" output transformer, something i can buy instead of building it myself" and as I am recommending something, I thought it wise to recommend both what I and the original designer suggested as it meets both our criteria, as well as Andreas' and increases my confidence that another circuit built will give similar excellent sonic and measured performance to what I acheived.

* Though it's not too hard to work out the general idea of what he's done, if you think about it a bit.

 

[Brett]

This may not be the place to expound on Allen's explanation of "quality lost big-time" (if there is a reference available I would like to read it; Jaap, can you help?) but just to caution that observations can and have been subjective. This loss of quality due to a few mA of unbalanced dc was certainly not my experience over decades of tube amplifier design of all sorts, neither have I come across this before in the wide experience of others. Output transformers have been designed with distortion significantly lower than that of the output tubes, so where is the total picture?

I've also been designing and building for a long time, and I know the electronics and magnetic theory well. However, in your actual experience have you built a differential output stage and measured / listened to it?
Nelson Pass also makes reference to this, indirectly, in one of the later Zen amp articles (# 7) also telling how to optimise the circuit.

At the hime I was actually using these amps in my system, my system had a sensitivity of circa 104dB/W/m which gives a greater insignt into the low level performance of amps at even very modest listening levels.

Philosophical comment: everything is subjective (a human interprets the results from the test gear, no?), so if simple measured perfomence was the key to musical enjoyment, then we should all be quite happy listening to SS recivers from the 70's.

 

[Bas Horneman]

I would like to respectfully say here that we must be careful before we take observations such as this by Allen as dogma.

This is always a good attitude to have.

I do not know the gentleman and will have no hesitation in affording him the greatest of respect, but this observation is contradicted by the experience of others, as well as basic magnetic principles.

It seems to me it has not been contradicted by the webzine Enjoy the Music as his latest amp where he takes balance one step further has gotten the best product of the year for the SECOND time. So it must have some merit.

http://www.vacuumstate.com

 

[argo]

Hi,

.....

- PP-Design operating in Class A

- Triode output stage

- "common" tubes, please no "exotics" which are difficult to replace

- 10 to 20 watts output power per channel

- "common" output transformer, something i can buy instead of building it myself

- and, i would like to do completely without semiconductors, so i need a rectifier tube vor the anode voltage, and AC heating for all tubes

Well, i'm looking forward to your suggestions, and sorry for my English

Greetings,
Andreas

- PP in Class A

- Triode output Stage

- Common and cheap tubes

- 10 Watts per channel output

- Common 5000ohm PP output transformer

- Just replace the diodes in the power supply showed in schematic with rectifier tube like 5U4G, 5AS4 or GZ37


Cheers,

Argo

 

[Tubes4e4]

Re: Re: Suggestions for Tube Amplifier

Hi Argo,

- PP in Class A
- Triode output Stage
- Common and cheap tubes
- 10 Watts per channel output
- Common 5000ohm PP output transformer

And don´t forget:

- burning only 135V in the cathode resistors

Tom

 

[argo]

Well, there is hardly ever free lunch

 

[Brett]

- the Wright PP1C sounds interesting to me, Bas do you have more info on it? Couldn't find anything on the forum... Especially performance (frequency response, THD, output power) and the second part of the schematics, containing the constant current source...

Bas has given the original CCS and Allen's distortion spectrum; mine was similar iirc. I also used a cascoded 2540 or a Pimm BBMCCS instead of the 820.
I also used a somewhat different PSU in the end, using my preferred 6CJ3 damper diodes in a FW bridge, LCLC using motor run caps.

- somehow i feel that this ccs will be using semiconductors ... is it possible to design a tube based ccs instead? (at ANY cost )

Why? A tube CCS is bigger, adds heat and cost and would give only one performance advantage; you can tell all your 'phile mates "it's all tubes!"

- Brett are you sure sensitivity will be high enough to drive the circuit directly via line level source and volume pot? How difficult would it be to add an appropriate preamp stage?

You can tweak the gain by adjusting the cathode resistors in the front end. It had enough for my system almost all the time. Only when I played one of those early issued CDs that were recorded at stupid low levels did I have a problem. I f you decide on this amp, build it this way and add a preamp if you really need it. I'm of the idea that a system should have the minimum gain (and number of stages) it needs and max headroom. I'm also not scared of winding the volume control past 9 o'clock as too many people are.

EDIT: as soon as the idea becomes more detailed, I'll start a new thread on the planned amplifier, i think I'll have enough questions for more than one thread... css, preamp, ps, soft start for tube heating, delayed switching on of anode voltages etc...

The CCS, I've mentioned above. The soft start and delayed B+ can be gained by the damper diodes.