EL34 PushPull UL Amp

[Tesla88]

Hi !I'm new on this forum ,i'm from Italy so sorry for any mistake.

I bought a new pair transformers for a PushPull Amp with Ultra Linear connections .
Here you find specs fot the trasformers:
TU052

Primary Z :4300 ohm
Primary Rdc :220 ohm
Primary Inductance :40H
Current :120mA
Pout rated 35W RMS

I think 4300 would be Ok for a Pair of 6CA7 - EL34 tubes in Push Pull , the UL connections are at 43%

What circuit would you suggest for best results?
What about Dynaco ST70? Any better ?
Thanks

 

[DF96]

Mullard 5-20? The main snag with that is that it has too much gain for modern sources. The Lundahl site has a modified version.

 

[prairieboy]

Check the various board options at Classic Valve Design - Welcome!. I've always found them easy to deal with, and the products are good.

 

[Tesla88]

uhm , in the Mullard 5-20 schematic i don't like the ECC83 as driver so much ...

 

[DF96]

People often say that. Often because they have heard someone else say it. If you substitute a lower mu valve then you will need to add a CCS tail to the LTP, or use different anode resistor values to restore balance.

 

[artosalo]

Dynaco ST70 is a good circuit.
I would modify it by adding both bias- and drive balance adjustments.

 

[vinylkid58]

uhm , in the Mullard 5-20 schematic i don't like the ECC83 as driver so much ...

Check out the http://www.diyaudio.com/forums/tubes-valves/161702-opus-5-0-modern-mullard.html thread as a reference.

jeff

 

[roline]

IF you want to use less feedback the 6550 and kt88 tubes can drive the 4.3k Ra outputs easier. The 6L6, EL34 and 6P3S-E needed more feedback. There are thousands of ST70's with the EL34's if you want to copy that feedback scheme.

 

[planet10]

Check the various board options at Classic Valve Design - Welcome!. I've always found them easy to deal with, and the products are good.

I've been living with the proto of Gregg-the-Geeks DynaMutt and can highly recommend it.

dave

 

[Eli Duttman]

Has anybody paid attention to the fact that the O/P trafos lack the magnetic headroom needed to cope with a GNFB induced low freq. error correction signal?

I suggest the OP scan the archives for Bandersnatch's "exolinear" style circuitry, which does not place the O/P trafo inside a NFB loop. FWIW, I like having the O/P "iron" inside a NFB loop, but what has already been acquired will not support that type of circuitry.

 

[artosalo]

Has anybody paid attention to the fact that the O/P trafos lack the magnetic headroom needed to cope with a GNFB induced low freq. error correction signal?....

Would you clarify this, please ?

 

[roline]

The output iron was not the issue. If you use a cathodyne driver or LTP driver with stones and a .2u or .3u coupling cap, massive amounts of feedback are not necessary. Be sure to have a solid power supply. I like a diode bridge to large cap -L- large cap to a slow start pass tube diode to large cap. That is the B+ in the KEG-EAR amp.
4.3k iron prefers to be driven by 6550 and KT88 tubes. I tried 6P3S, 5881, 6P3S-E, EL34, 6L6GC and they were wonderful with 6.6k but fell short driving the 4.3k iron at volume. They required lots of feedback to do so. So there was another weakness. Also 6.6k Ra iron will give you more class A output with overall lower power out. 4.3k gives you more power out but less class A operation.
One advantage of the LTP is the use of local feedback plate to gate (220k) on the output tubes some call Schade feedback. I tried to drive 4 ohm speakers on 8ohm taps that would change the 6.6kRa iron to look like 3.3kRa iron. the little 6GV5 sweep tubes did not even notice the difference.
Strong recommendation would be to use a LTP to drive the output tubes with 220k plate to gate localized feedback to drive the 4.3k iron.....

You might want a little GNFB if that at all.

 

[Tesla88]

Hi guys , thanks for the replies

People often say that. Often because they have heard someone else say it.

I think a 12AX7 in a cathodyne configuration would be ok to drive a pair of EL84 or tubes quite easy to drive , to drive a pair of EL34 i think you need more driving capacity , es 12AU7 , 6CG7 , 6SN7 ...

I've a question about Dynaco ST70 , the 7199 tubes are rare and expensive , is there any kind of tube I can use instead? Es. one pentode for input like EF86 and one triode (6cg7?) for the drive section?

What do you think about this kind of circuit :
http://www.triodeel.com/achpwill.gif

On evening i will start bench testing the Acrosound schematic ,but i would like to test the Dynaco also, so any replies would be great!

Best regards

 

[DF96]

12AX7 is rarely seen as a cathodyne PS; more often as LTP. EL34 are fairly easy to drive. Remember, if you are using global NFB (as most of the classic EL34 designs do) then you need some HF rolloff to stabilise the loop. If your circuit does not naturally include this then you have to add it somewhere.

I think some people have had success by replacing 7199 with ECF80/6BL8? Might need rewiring and a few component changes.

 

[rongon]

I remember Van Alstine using 6GH8A as a sub for 7199, after rewiring the tube socket connections. Isn't 6U8A also a close equivalent to 6GH8A?

 

[artosalo]

...I've a question about Dynaco ST70 , the 7199 tubes are rare and expensive , is there any kind of tube I can use instead? Es. one pentode for input like EF86 and one triode (6cg7?) for the drive section?...

I can suggest one especially well working tube/cathodyne circuit; Soviet 6F12P
triode-pentode.

The schematic is this:

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

I have built several amplifiers with this and the performance is very good.

 

[scott17]

I use this circuit. It has been well criticized. I know that DF96 has seen this before. It works very well for me.

 

[payitforwardeddie]

Mullard 5-20? The main snag with that is that it has too much gain for modern sources. The Lundahl site has a modified version.

That's what I'm building right now. Don't hold your breath though, I might have them finished by Christmas 2013.

 

[Tesla88]

I'm going to test the Williamson circuit , i've built a pratical test bench to try different configurations , like Dynaco ST70 with modern driver using 6922 tubes , or a combination of EF86 + 5814.
This evening i'll start with the Williamson , just to test Output trafo and power .

Has anybody paid attention to the fact that the O/P trafos lack the magnetic headroom needed to cope with a GNFB induced low freq. error correction signal?

I Don't understand this , Output trasfo are rated for 35W rms , this is the nominal power of a EL34 UL configuration , size of OP trasformer core is same as a 200-250VA , really big and heavy.

 

[Tesla88]

Some test made with Williamson circuit using 6CG7 for input and driver, sounds great and output is around 40W rms (B+ 410V - 220mA at clip , ) , i get most 2nd harmonic distortion at medium levels . Bias is about 30mA each tube , el34 JJ

Output xformers are good sounding , really linear

I reduced the global NFB , getting more gain and more balanced sound , with original NFB i was getting a too "loudness" kind of sound.

Next step :
- substitute input tube with 5814
- try a triode cinfiguration for output tubes

Cheers