• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

When doubling output-tubes: What needs to be considered ?

I habe taken basically the schematic of the monobill II and applied the input section to a Ul-40s. i doubled the amount of el34. So, what are all the details to be considered now ? Each el34 gets its own coupling cap, its own neg. bias, its own grid resistors, basically everThing redundant. The opt is now 3k. The power x-former is large enough.

Input part from:

http://www.triodedick.com/monobill_2/monobill_2_schema_versterker.GIF

I ask myself:
- I believe Miller is neglectable, the el34 has only 40-50pf. Any other min. requirements on the driver regarding min. current ?
- are the grid res. for neg. bias adding up as min requirement or is each tube with 100k on the save side ?
- same question on the resistor for the ultra-linear tap: 1k enough like in the two tune scenario or double this to 2k each ?

What else ? What needs to be considered when going from one pair to two pairs ? My target is to have more power in class a, but not so much to have absolutely too much power. bass performance is as well not that important, bass goes separate through active subwoofers.
 

Koonw

Member
2013-04-09 9:37 pm
Here is a schematic of a commercial amp for parallel KT88 ulta-linear, you see how the bias for each output tube is implemented. Driver is cathode follower, output 200W, I get to hear it once in a dealer in capital city, it sounds very good. For 4 tubes per channel you can get about 60-100W with El34-b. I also built one 4x El34b for a friend before, output about 60W, the HT only 400V. The OPT is 120W from Bezdz
It's the entire schematic and components make up the sonic of amp, if you use another one it will has another sonic.
 
THX a lot...in principle bery similar to what I thought, but I have a few question where you can help me maybe:

- there is a resistor 100k as grid resistor for bias and a 100k resistor to ground in front of that...never seen this before ? Why 100k to ground with fixed bias ?

- the rc from Ht+ to UL-tap is for what exactly needed ? I guess against oscillations, but is that needed ?

- 100R resistor per tube to the UL tap seems to be low...is this by purpose ? Should I use instead 1k now as well 500 R if doubling the tubes ?


- the cathode follower...well for so kany tunes needed I guess, but a 6n6p with 13Ma should easily drive as diff. pair as well four el34, two pairs...or?


Best Regards

Frank
 
Last edited:

Koonw

Member
2013-04-09 9:37 pm
THX a lot...in principle bery similar to what I thought, but I have a few question where you can help me maybe:

- there is a resistor 100k as grid resistor for bias and a 100k resistor to ground in front of that...never seen this before ? Why 100k to ground with fixed bias ?

- the rc from Ht+ to UL-tap is for what exactly needed ? I guess against oscillations, but is that needed ?

- 100R resistor per tube to the UL tap seems to be low...is this by purpose ? Should I use instead 1k now as well 500 R if doubling the tubes ?


- the cathode follower...well for so kany tunes needed I guess, but a 6n6p with 13Ma should easily drive as diff. pair as well four el34, two pairs...or?


Best Regards

Frank

The other 100k is emitter resistor of 2n5401, the fixed bias voltage is same as emitter voltage despite connecting to ground, so the total load is still about 200k should be ok for coupling cap 0.22u.

Yes, judging from the time constant of RC 100k and 100pf of 100khz, it is to limit the upper frequency response.

For KT88 screen resistor 100 is ok, but for EL34 it's 1k because it is sensitive to screen grid current overloads, and since they're // tubes, it's necessary to make the (screen) supply very stable to prevent motor boating.

Yes, but for better music power, cathode follower or similar can better handle the grid current and faster blocking recovery at clipping, you can add some local feedback to 6n6 driver to make it more rigid in case it doesn't meet your need, maybe you never need so much music power and sound still good.
 
Last edited:
THX a lot...in principle bery similar to what I thought, but I have a few question where you can help me maybe:

- there is a resistor 100k as grid resistor for bias and a 100k resistor to ground in front of that...never seen this before ? Why 100k to ground with fixed bias ?

They drew this schemo a bit differently. R45-52 are the emitter load resistors of the bias network active decouplers.

- the rc from Ht+ to UL-tap is for what exactly needed ? I guess against oscillations, but is that needed ?

Could serve dual purposes: load down the natural ringing frequency of the OPT. Since this is a gNFB design, it could also serve to improve the phase margin at the high frequencies, or both at the same time. Series RC loading (a "Zobel" network) is preferred over a plain capacitor since the phase shift of a capacitor can approach -90deg as the frequency increases, and you don't want that. A series RC circuit limits how much phase shift can occur as the frequency rises.[/QUOTE]