coinmaster said:

just the components.

Looks complicated :

Click to expand...

The source follower grid drive mod would be simple and easy to implement unless MrCurwen has some fancy pants version of it.
I would be extremely interested in someone trying this out because I am very interested to see if the claims that a source follower does not effect sound are true. Source follower have often been said to sound better than cathode followers as well.
I won't be able to test this my self until next month since I've realized that I'm missing about $100 in components I need for my power supply and I can't pay for it atm.

but a reason we like tube amps.

Click to expand...

Technically it is current that drives headphones and transistors are as good at current duties as tubes as at voltage duties.

It will probably be one of last things attempted,

Only because we are backed up with focusing on optimal tubes and bias now, and finishing overdue mods.

Click to expand...

Force balancing tubes is an even worse idea to try first. If we want to use 4.5x resistor value of plate resistance then that doesn't leave us an easy option with our 6SN7 either.

Click to expand...

Well...
To answer first two,
I wil post a reason why I do not want any hybrid or transistors in this design:
Look here:




http://www.theaudioarchive.com/TAA_Resources_Tubes_versus_Solid_State.htm

Read it and weep.

Of course out of curiosity, nothing will be overlooked including this.

Also, you are mistaken I am not forcing anything with the tube circuits.
We are simply reviewing limits and ideal parameters for all the tubes,
Since this thread is turning into a complaint thread about bias without giving suggestions.

Edot*:
More info:

http://kenrockwell.com/audio/why-tubes-sound-better.htm


This show triode as best device over SS or pentode:
http://www.vac-amps.com/sciam.html

Also, you are mistaken I am not forcing anything with the tube circuits.

Click to expand...

I'm not sure if you understood what I meant. Go back to MrCurwens explanation of force balancing.
 

Since this thread is turning into a complaint thread about bias without giving suggestions

Click to expand...

Are we reading the same thread? It's the opposite of that.
 
 

Read it and weep

Click to expand...

There's an eternal argument on both sides. You shouldn't make opinions based on a few sources. There's always more to it then that. As MrCurwen said it's about the right component for the job, among other factors.
I really don't care about the theory at this point, for every person claiming one thing there's another person claiming another thing.
I'll believe what I hear.
There's a whole lot of people that claim source followers sound really really good, even compared to tubes, coming from people that build tube amps, therefore it is worth looking into.
Also you should let the people on the forum decide what mods they want to do instead of dictating it. 
 
My bias pot idea is a bust due to the high RA requirement for the 6sn7. The only way to optimally bias the 6sn7 is to create bulky complex circuits or increase the power supply voltage by a few hundred volts which is unlikely to happen.
 
This leaves out input stage optimization unless or until we have another idea.
We've already dealt with the output stage other than CCS tails like Mrcurwen suggested but that is delving into circuit creation again. 
Current source tails could be doable without too much issue in theory given the layout of the board but it would take some effort to create an official guide for it and I question what happens when the signal is moving horizontally on the bottom triode and diagonally on the top triode, you might need to use a gyrator or inductor on the plate of the top triode which turns it into a whole nother mess of things.
 
The only mod suggested so far that is easily doable and explainable is the source follower grid drive mod.
Let people decide individually what they want to do.

baronbeehive said:

Then with the bigger fans the cooling should e sufficiet, hopefully! It will be interesting to see whose amps are still working in a years time!

Click to expand...

Looking at this picture really chills my soul! Good luck!

Every day I'm nerveous for 5 seconds when I think about how the components are cooking in my amp builds, especially the flagship one I delivered last week. The power buffer could be used as a grill...

Active cooling does wonders.

maxx134 said:

The design on driver stage is autobias

Click to expand...

Yes.

and common cathode.

Click to expand...

No.

Common cathode gain stage is a gain stage that has it's tubes cathode at AC ground. Either it is outright grounded, or it has a direct AC path to ground via a large enough capacitor.

A long tailed pair has a tail that cannot be bypassed with a capacitor. If it is bypassed, you do not have a LTP anymore, you have something else.

What this circuit has, is a LTP. Not a common cathode.

The fact that the cathodes are connected together at the cathode does not make it a topology named common cathode gain stage. This nomenclature was in use before WWII, and still is.

So it's not the design as flaw,
As reading MrCurwen posts may imply.
But the cost cutting measures used to carry out the design objectives in our amp.

Click to expand...

Potato, potato.

yes there is interest but let's keep the modest quality out and at least say "good quality" !

Click to expand...

Modest quality not good enough, eh?

In soviet russia you don't choose the cap, the cap chooses you!

In future, I propose only use cap of very cheap price and horrible aesthetics. No pretty boy caps anymore!

So..
A cheapo cap > to transistor > to the grid...
Blasphemy!.. Lol

Click to expand...

All just components. The question is, which one performs best in which role? You wouldn't go hunting with a housecat, and you wouldn't have a Lapland husky at your bedside inside house while sleeping.

No one component type is good at everything. Even everything it could possible be used for.

I do not want any SS device in my signal path! (as a choice.)

Click to expand...

Well you do at the moment. Several.

Looks complicated :

http://www.bartola.co.uk/valves/tag/schade-feedback/

Click to expand...

Schade is not applicable in any way here.

One of the many many reasons I have read tubes are a superior devices to SS.

Click to expand...

They most surely are not. Depends a lot on what you are doing.

Try filtering a 11V rectified supply to get superclean 4.2V DC for heating of directly heated 4P1L. It takes only tiny tiny 320mA of current.

Try getting best possible output impedance for direct drive system driving headphones. You say you don't like low output impedance amps you've heard (SS amps); well, like I said earlier NOT ONE OF THEM had a linear voltage amp and a source follower driving the output (making the output impedance low).

Talking about the curves I commented;

It's a tube made for low powered radar pulse oscillator, or possibly small TV sweep duty. 6SN7 could be, or similar.

So it can pulse a few hundred milliamps? That's quite impressive for a tube, very impressive, but it's childs play for pretty much any FET costing 50 cents (I buy bulk).

Again, I would never ever ever use an SS device to do voltage amplification. That's another story. Just as I would never use a tube to control any meaningful amount of current in a "best results" amp.

We are currently testing what works best, not just in theory.

You see I don't dismiss anything that is found in real testing.

Click to expand...

It could be the language barrier, or whatever, but just to clarify in any case; all of the things I'm talking about have been succesfully built a number of times. I'm not sure by any means you were directing that at me even, but just to say it. There, it's said.

Also, of interest to me, is Soundstage...
They all have great separation,
But I found Soundstage to be the key in separating the best amps from the others.

Click to expand...

I completely agree.

I tease my friend who has a surround system that surround sound is for people with low fidelity.

Actually, driver tube selection is not even a choice,
As only one dualbtriode tube per channel makes it already up to the circuit to balance the two triode sections within ONE tube,

Click to expand...

Yes. It does this sloppily. Have the tail impedance be at least 1 megaohm, with the 220k plate resistors, and then you've got balance.

which are by default already close to matching being identical within the same tube,

Click to expand...

Not at all.

No no no. Not in any tubes that are actually made and sold. Sure they can be matched, but that's an anomaly that has to be laboriously looked for. Really well matched tubes cost a lot more. (Not all sold as matched are very well matched in actuality. A lot of dishonesty.)

+-10% is the norm. +-1% is like one in a thousand tubes manufactured.

Impressive! but wondering,
have you built a balanced tube amp?
Not many balanced amps out there.

Click to expand...

My original designs for the past three, maybe four years, have all been balanced. (I did do two SE amps as well, one this year, one last year.)

I've mainly built balanced amps for years.

I've also built balanced direct drive ESL amps, two of them. The DIY ESL elements need a lot of work, but the amp was fine. Direct drive was nice at the time, but later I achieved the same things other ways.


I catch your drift, I'll make a new thread for general tube theory. There's a lot of forum sections here, could somebody tell me where it might be smartest to put it?

I catch your drift, I'll make a new thread for general tube theory. There's a lot of forum sections here, could somebody tell me where it might be smartest to put it?

Click to expand...

In the DIY section most likely.

mrcurwen said:

In soviet russia you don't choose the cap, the cap chooses you!
In future, I propose only use cap of very cheap price and horrible aesthetics. No pretty boy caps anymore!

Click to expand...

 
For the record, my three first coupling cap replacements were russian :
K75-10 (0.33µF/500V) ... nice upgrade from the horrific SCR-PA 0.22µF MKP
K73-16 (0.47µF/630V) ... I really liked those ones, very lively
TF-3 (0.22µF/600V) ... but I probably didn't have the time to break them properly
 
And to be honest, none of those could not stand a chance against the Jupiter Copper/Wax, the difference is just cruel.
 
 
But the K75-10 ar still in my amp, as WCF cap, bypassed by small K72P-6
And I have plenty of SGM-4 in my new cap banks, alongside some WIMA FKP-1 and other stuff
Oh ! I didn't mention the K71-7 I still have to test sometime ...
 
 
So, yes, I would tend to agree with you. No need to break the bank to have more than decent caps.
But the best (Duelund, Jupiter, AN, V-cap, ...) are in an another league, pretty much like the transformers from what I have understood (think "amorphous" ...)

Does the Jupiter Copper/Wax caps require a operating temp of 80°C to sound best? 

john57 said:

Does the Jupiter Copper/Wax caps require a operating temp of 80°C to sound best? 

Click to expand...

The Jupiter are in another league..
A SERIOUSLY huge jump.

Imagine, thinking you have a really great cap in a mundorf silver/gold/oil..

Then you switch to this Jupiter copperfoil cap,
And suddenly the mundof sound unnatural and artificially sweet in treble?

The treble and whole sound of this cap is so natural and clean, it is a seriously transparent cap.

Edit*
I am only describing the coupling cap position.

Not the WCF position where the Mundorf are my second choice.

Also I feel jupiter in the WCF position is not needed and a waste of money there, base on past descriptions of WCF sonics I posted .

maxx134 said:

Very nice!
have you tested it yet?

Click to expand...

No! I tried to test it about a week ago and my DAC didn't work, maybe something to do with changing over to windows 10. I can't replace bad tubes or test it further atm because I'm still in temporary accommodation. So I still have to wait for another few weeks until I'm in my new permanent home unfortunately .
 
But I will give my impressions as soon as possible.

I hate having to make these clarification posts,
But here we go again. .

coinmaster said:

I'm not sure if you understood what I meant. Go back to MrCurwens explanation of force balancing.

Click to expand...

Yes I understood, but I agree the force balancing won't work in our amp.

coinmaster said:

Quote:
"Since this thread is turning into a complaint thread about bias without giving suggestions.."

Click to expand...

we reading the same thread? It's Are the opposite of that.

Click to expand...

Sorry, but I called it as I saw it...

No values,
just shedding light on general design limitations.

The driver stage cathode resistor issue & bias issues are common amp design compromises brought to light without any solutions.

I didn't say it was a bad thing.
Shedding light on any weaknesses is a good thing, as it makes us think more,

So we have more things to look over and fine tune.
For instance, this helps to narrowing down what are the best choices for driver tubes to match the circuit.

So far we dealt with our limitations by otimization of the circuit with higher tolerance quality matched components.
And Mods that are sonically verified.

coinmaster said:

"..you should let the people on the forum decide what mods they want to do instead of dictating it. "

Click to expand...

Ugh!
HUGE MISUNDERSTANDING!!


Now YOU'RE the one reading differently!


Since when have I actually told anyone WHICH mods to do.???

In fact it's the opposite..

I enthusiastically accepted SonicTrance bias mod..!
Even to impementing similar into my own amp..!

THEN,
Redge78 posted SonicTrance mod on first page..(!)


You want to do any mod on this amp go right ahead and be my guest!

You want to specifically suggest a mod,
then go right ahead be my guest!

Don't confuse things.

coinmaster said:

"The only mod suggested so far that is easily doable and explainable is the source follower grid drive mod.
Let people decide individually what they want to do.."

Click to expand...

Easy?
Where would you come off a "common cathode tail"?
You would have to eliminate the balanced input stage,


Just because I have my doubts on implementation.
If it will benifit ME,
over what I currently have..
Which is a top quality over-spec jupiter copperfoil as coupling cap of highest clarity...
Doesn't mean it I disagree it will be working perfectly in other designs.
So sorry if your misunderstanding me.

Also,
I am not stopping anybody..
I only stated I am backed up on mods..
I have 3 other things to do.
So Who is going to try this "easy" mod?

There are only two others here that might try it besides me.
You wanna help with specifics & answers to speed it up?..
Instead of complaining about me?

coinmaster said:

I'll believe what I hear.

Click to expand...

At least we can agree on this.

mrcurwen said:

No pretty boy caps anymore!

Click to expand...

Haha!
I gave up some of those "pretty" caps for "Vitamin Q" caps in my WCF cap area.
I go for the sound over the looks,
Yet in these amps, the pretty white caps are actually very inexpensive .

mrcurwen said:

Well you do at the moment. Several.

Click to expand...

Potato / potato..
No,
I don't have any SS device "IN" my signal path.

mrcurwen said:

Yes. It does this sloppily.
Have the tail impedance be at least 1 megaohm, with the 220k plate resistors, and then you've got balance.

Click to expand...

So you admit It works...

I wish your general suggestion was applicable in our amp,
But as Coinmaster pointed out,
the issues are PSU voltage limitations before those 220k plate resistors.

mrcurwen said:

"Schade is not applicable in any way here."

Click to expand...

I will take it a step further and say not much in these past topics thus far is applicable here.

mrcurwen said:

"They most surely are not. Depends a lot on what you are doing."

Click to expand...

Agreed I was too general.
Specifically, I am referring to this amp for now.
So my point and links count.

mrcurwen said:

I tease my friend who has a surround system that surround sound is for people with low fidelity.

Click to expand...

Oh!
I never thought about that!
Haha good point

mrcurwen said:

NOT ONE OF THEM had a linear voltage amp and a source follower driving the output (making the output impedance low).

Click to expand...

Don't get me wrong.
I actually find this idea VERY interesting.
I have not looked at other hybrids to see whats the common implementations, but this sounds unique.
I'm a hobbiest who is learning all the time so it's interesting to see new things.

mrcurwen said:

I catch your drift, I'll make a new thread for general tube theory. There's a lot of forum sections here, could somebody tell me where it might be smartest to put it?

Click to expand...

To clarify;
The only "drift" I am making is to clarify whether ideas presented are applicable to this amp.

They are nevertheless interesting and others like your posts and no one is telling you not to post.

mrcurwen said:

My original designs for the past three, maybe four years, have all been balanced..

Click to expand...

Now your talking...
Have you compared yours sonically to the usual popular top amps around?

There are a few popular amps in threads here like the "torpedo" and "bottlehead" amps,
which users can build kit or order, but
Those amps are only upper mid-tier amps..
Not TOTL type..
I personally heard them in quiet settings so I know.

I have not heard the Feliks-Audio ELISE amp yet and that would be interesting to hear as I love all tube amps.

So far I haven't heard anything to pull me away from this amp in its current highly moded form.


Do you have kits? Or just build?
That would be a better thread to start.

john57 said:

  Does the Jupiter Copper/Wax caps require a operating temp of 80°C to sound best? 

Click to expand...

To answer that specific question ... NO
 
Actually, you should try to be as far as possible from the "max temp" of the cap.
A generic rule of thumb ("Arrhenius law") is that you double the life expectency of each 10° decrease from the Max Temp. Originaly, this law is for chemical caps, but you can never be too careful.
For a cap given for 5000H @80°, you can expect to keep it for 10000H @70° or 20000H @60° ... 

The driver stage cathode resistor issue & bias issues are common amp design compromises brought to light without any solutions.

Click to expand...

There are solutions, we discussed them.
 

Now YOU'RE the one reading differently!

Since when have I actually told anyone WHICH mods to do.???

Click to expand...

 
Here:

coinmaster said:

 

"At the very least the source follower mod on the grid seems easily doable without annoying complications and should probably be one of the first things looked into at this point, especially if we want to test the "good caps are better" theory."

Click to expand...

Click to expand...

It will probably be one of last things attempted,
Only because we are backed up with focusing on optimal tubes and bias now, and finishing overdue mods.

We are currently testing what works best, not just in theory.

Click to expand...

Easy?

Click to expand...

Yes, the source follower mod would require a few tiny components and a wire, the other proposed mods would require many many times more effort to implement due to the bulk and complexity and the layout.

Where would you come off a "common cathode tail"?
You would have to eliminate the balanced input stage,

​

Click to expand...

Those sentences were not in good english, mind repeating?
 

Just because I have my doubts on implementation.
If it will benifit ME,
over what I currently have..
Which is a top quality over-spec jupiter copperfoil as coupling cap of highest clarity...
Doesn't mean it I disagree it will be working perfectly in other designs.
So sorry if your misunderstanding me.

Click to expand...

I have no clue whether it will sound better either I have Duelund Cast caps which are considered the best of the best of the best. Still though, after doing some reading and rereading some grid driving posts it does seem to be proven that the source follower mod can improve transients greatly, whether that is applicable here with high end caps in place I don't know. In either case if the theory behind it is correct then it wouldn't surprise me if it worked given the abysmal current the input stage runs at.
 
 

You wanna help with specifics & answers to speed it up?..

Click to expand...

I'm waiting to see if there is interest, also MrCurwen said he would go into it more if there was interest. I'm a ways off from testing it, I've got other priorities and the waiting game on cash and delivery is creating more of a delay than I had hoped for my build.
 

Instead of complaining about me?

Click to expand...

Where's the fun in that? 

 

I don't have any SS device "IN" my signal path.

Click to expand...

Well, technically the regulator is solid state and it most certainly affects the sound due to inherent impedances on the signal if nothing else. I would rather go with a shunt regulator due to the supremely low impedance they have, in either case solid state is helping rather than hurting here unless you want the warm, "old school" tube rectifier/regulator sound.
 

I wish your general suggestion was applicable in our amp,
But as Coinmaster pointed out,
the issues are PSU voltage limitations before those 220k plate resistors.

Click to expand...

There is hope, a current source should pose effectively the same impedance as a 1 meg resistor with the added bonus of a horizontal load line.
Plus a gyrator load which negates the issue of the power supply voltage being too low and also provides super high impedance.
I already have a nearly complete PCB board for the both of them
I chose a different variation of gyrator then MrCurwen but in general it's the same.

It wouldn't be too difficult to implement assuming you could find the space(it's smaller then it looks). You would need to create your own -15v supply though. I could make it so you can plug it in to the B- of the amp but I don't see a safe way to prevent disaster in case of some sort of malfunction putting -100v on the cathode. 
 
If there's interest I can create a PCB model for the 15v supply as well using the 2 fet design MrCurwen was talking about, or you can just buy an IC for it which will be much smaller and is way cheaper. I wouldn't recommend perf-boarding it, I fried many a component with those accursed things and they are a P.I.T.A to get working.
There's no way to fit these inside the case unfortunately, at least I don't think so.
 
This PCB can be used on the output stage as well which gives the tubes CCS tail and gyrator load which should keep both tubes on a horizontal load line, unlike simply using a CCS tail which seems like it would unbalance the current between the tubes.
 
Also you can fully traverse the load line horizontally and vertically with it although you would want to be able to read voltage/current levels while doing it and traversing the load line of a WCF is more effort then it is worth over just picking a good spot on the load line.
 
Once I have the file ready I can upload it and you send it to https://oshpark.com and spend $30 or something for a few copies. I've used them a few times and they make really nice PCBs.
 

The only "drift" I am making is to clarify whether ideas presented are applicable to this amp.

Click to expand...

That depends on the individual. They are all applicable if you put in the effort.

Nice ideas. So maybe we could make a list of potential mods together with supposed benefits, difficulties in implementing and so on to give everyone an idea what to go for?

coinmaster said:

There are solutions, we discussed them.

Those sentences were not in good english, mind repeating?

Click to expand...

Coin
you didn't answer me.
Instead you say it's not good English.
Since when are you an English teacher?

1-Yet again you didn't admit your wrong about slandering me.
2- your trying to "grasp straws" by highlighting my use of the word "we"..
The word "we", could mean me and my dog...

3- Your blatantly avoiding my points about the driver stage.

You cant implement this without loosing the "common cathode" connection, responsible for keeping the balanced operation.

You forget this is a balanced amp?

Basically I see you would have to separate the cathodes for that "idea" so there is no balanced driver stage.

Thanks for the info in that post for the boards,
But that doesn't make you right.

It be nice to have some help from you for a change without trying to be right all the time.

QUOTE :
"Oh, well if there enough interest. ."

Click to expand...

That's not the point.
The point is if it is viable to even try.