Mullard 5-20

[JDeV]

Originally posted by Tube_Dude
See if across the feedback resistor exist a capacitor of 100pF...if not ...put one...and keep as informed!

This cap is there, 330pF actually. On this specific amp., it is just the extra OT , else exactly same as org. Mullard 5-20.

[Tube_Dude]

And the value of the feedback resistor?

[JDeV]

Originally posted by Tube_Dude
And the value of the feedback resistor?

For 8ohm speakers, (7.5 on the org. schematic) Mullard recommend 5.6k for feedback resitor, mine is 5.6k.

[Tube_Dude]

Try to change the capacitor to 220pF and see in a scop if the oscilation increase or became smaler...

But i'm a litle ceptical as the diyer that assembled your amp may have comected same mistakes...and seeing it oscilating put the rc at the anodes of the EL34 for trying to stop the oscilation... that it still have!

Jorge

[JDeV]

Originally posted by JDeV
On this specific amp., it is just the extra OT , else exactly same as org. Mullard 5-20.

And as I said earlier, R24 and 25 goes directly to centretap on T1.

Why would there be 2 OT's ?
Anybody know of software with tube symbols?

[Tube_Dude]

I see two transformers...and i haven't drink nothing!

[JDeV]

How importand is the following , from this article:

"Design for a 20-Watt High Quality Amplifier
By W.A. Ferguson,* B.Sc.(Eng), A.C.G.I., Grad. I.E.E.
"Choice of Valves and Operating Conditions"
* Mullard Valve Measurement and Application Laboratory.

"A bus-bar earth return has been used with chassis connection at the input socket

I know my amp's earth busbar is only connected to the chassis near the bridge diodes.

[bournville]

JDev,

I'd hazard a guess that perhaps the first transformer has a relatively high output impedence - perhaps it was designed for a 70 volt line - and thus another transformer is required to drop the output for normal speakers. Of course, the second output tranny is not included in the feedback loop and this will affect the linearity of the response (apart from any mismatch on impedences by this setup).

There is another difference also - the first tranny is set up for normal loading, not UL (distributed loading) of the original Mullard, and that means the cathode bias resistors are wrong, too high really for this kind of working. If you want to sort out any instability you'd need to sort these things before worrying about bus bar connections.

(If all else fails to remove the oscillation, try small caps, a few hundred pfs from each of the EL34 grids to ground, or small caps, 1000pf at 1500vdc working, from each of the EL34 anodes to the centre tap of the first transformer. If the amp is deviating from the original Mullard 5-20 circuit then you cannot guarantee that the same methods used in the original will bring stability to your amp).

[JDeV]

Thanx to all for help. Bournville, you did it.
1st tried the "small caps to ground" idea. no help.
So disconnect OT2, problem gone. Connect fat 8ohm resistor, inject 100Hz test signal, appears perfect on scope. Pump up volume, start clipping at 2.4Vrms. So atleast it is working, no I must just get it up to the claimed 20Watts.

What must I do next to increase the output. At 12ohm I got it up till 4.2Vrms, still not 20Watt. Is it all because of the wrong OT, or some other factors as well?

[bournville]

Jdev,

Glad that you've got the oscillation under control, although I can't claim any credit for it!

It is interesting that the output appears to increase with a higher load. Looking back on the altered circuit, I was thinking that perhaps your Mullard had been adapted for PA work rather than normal audio work, which could mean that perhaps the transformer secondary has quite a high impedence (sometimes as much as 500 ohms) because PA speakers often have their own output transformers connected. Likewise, the normal loading as opposed to UL working could be an attempt to increase output at the expense of distortion, not such a problem in a PA amplifier.

In terms of checking, are you are sure of the HT voltage, bias voltage on the EL34 cathodes (check to make sure that operating conditions are not upset by 'leaky' coupling caps) and efficiency of the GZ34? Check the resistance of the output winding of the transformer - is it higher than you'd normally expect for a 3, 8 or 15 ohm secondary? If it is significantly higher it may be the high secondary of a PA transformer, which is why another transformer is connected to bring it down to 3 or 8 ohms.

If it does seem to be a 500 ohm secondary, check the feedback resistor - if it is a value for a 8 or 15 ohm secondary it will be way to low for a high impedence secondary, which would cause phase shift and oscillation in the amp.