NFB in valve power amp.

[awkwardbydesign]

I have an old Velleman K4000 that I am modifying, and I have a query about the NFB circuit.

In the original K4000 circuit the 8 ohm tap is used, and in the later K4040 the 4 ohm tap. The only other relevant change I can see is the resistor in the circuit (R13 in the K4000, and R68 in the K4040. Same resistor, just different numbers).

1K5 in the K4000, and 820R in the K4040, and different transformer taps used.

Does it matter which tap is used, as it is really fiddly to change them, it means unscrewing the back panel, unbolting the transformers, wedging them up off the wires and resoldering the solid and stiff wires. I have to use a 100w iron to get enough heat into the binding posts, and if it doesn't need to be done I would prefer not to do it.

I am also following an upgrade (I hope!) posted in DIYAudio to convert to parametric (?) phase splitter, which means changing to K4040 spec too!

While I am at it I am throwing fairydust resistors and caps at it, but cheap ones. (Anyone remember Mavis the fat fairy in Willo the Wisp? Magic, but common!) I'll post a thread in DIY later, with pics, but I would like to avoid letting the magic smoke out.

Here are the cicuit diagrams.





And the modification.

 

[SergeAuckland]

Changing output taps ina valve amplifier changes the output voltage and so the open-loop (i.e. beforee feedback) gain from input to output. So, when the feedback is taken froma different tap, it needs a different value of resistor to make closed loop (i.e. after feedback) gain the same. If the resistor wasn't chanhed, then both the gain would be different, but more importantly, the stability margin between when the amplifier changes from being an amplifier to being an oscillator would also change.

Valve amplifiers with feedback unlike SS amplifiers, are always a compromise between too little feedback, which gives higher distortion and output impedance or too much feedback which gives instability.

S

 

[Valvebloke]

As Serge says, provided you use the right value resistor (820R on the 4 ohm tapping or 1k5 on the 8 ohm) it shouldn't matter much which tapping you use. What impedance are your speakers ? I have heard it said that if you take the feedback from the tapping that the speakers are not connected to then in the event of a speaker short-circuit you won't be so likely to induce damaging instability in the amp. I haven't seen a numerical justification for this though and as I write this I can't see why it would obviously be the case. Maybe I'm missing something. Speaker shorts will, I hope, be rare though. So I wouldn't personally let this influence my choice of which tapping to use.

VB

 

[awkwardbydesign]

Thanks to both of you for that. Changing the feedback resistor as you change taps seems so obvious now! I wonder why they bothered when they changed from K4000 to K4040 specs.

As for speaker impedance, I don't know at the moment. When the amp is finished I will be sorting the speakers, so it could be anything. I will certainly bear it in mind though.

This amp was for sale, but when I brought it down from the loft and used it, my wife said she had always liked it! So it's staying, even if I gut it and put a different circuit in. As long as the valves are the same shape she'll be happy, and WAF is priceless.

To be honest, I would think something like 6SN7s would drive the EL34s better, but I don't know enough to redesign it. Obviously! When I built it in 1992, I bought the transformers from Velleman and made everything else, just borrowed a kit and copied the pcbs, monoed the PSUs and left out the soft start. Cost about £250 instead of £750, so I'm now spending a few pounds to update it a bit.

Fingers crossed. :nerves:

 

[Tenson]

Keep the valves and heater circuit. Put a good quality SS amp circuit inside

Just what is the modification doing and why is it better?

 

[awkwardbydesign]

a good quality SS amp

Oxymoron.

 

[awkwardbydesign]

Just what is the modification doing and why is it better?

Raising the voltage available from the phase splitter, making it more linear, and lowering the LF-3dB point. Hopefully it WILL be better.

 

[Tenson]

making it more linear.... Hopefully it WILL be better.

Careful, you might make it sounds like a SS amp!

 

[awkwardbydesign]

Careful, you might make it sounds like a SS amp!

:shock: :nup: