Feedback

[Syrinx1]

I'm building a lower volume variant of a classic guitar amplifier, and I'm keen to retain as much as possible the original circuits distortion characteristics. The stock amplifier circuit uses 2 x 6L6 in fixed bias push pull. My practice amp will instead use a pair of EL91's.

The original amplifier design uses the 4ohm OT tap to provide global negative feedback to the phase splitter, but for the EL91 valves, I only have at hand a transformer with one 15ohm tap.

The 15ohm tap would provide approximately twice the feedback voltage of a 4ohm tap.


Am I correct in thinking that because the drive voltage requirements for the EL91 valves is lower than that for the 6L6 valves that using the 15ohm output will somewhat correct for the lower drive voltage of the valves I'm using?

... or am I just going to end up with potentially too much drive for the EL91's but with less phase splitter clipping than if I instead obtained an output transformer with a 4ohm tap and split the anode load resistors of the phase splitter to tap off the required drive voltage for the EL91 valves?

[Ed_Dinning]

Hi, if you have the circuit to hand, see if you can identify the feedback resistor from the op trans. replace with double the value for the 15R tap.
There may also be a cap involved in the circuit. This is often "tuned" by feeding a 1KHz square wave in and adjusting the cap for best waveshape on the output.

If you look at the details of the Mullard 5-10 amp it shows how the value changes with designed op impedance.

Ed

[Syrinx1]

Thanks Ed,

But if I actually want to reduce the drive of the phase splitter so that it was better suited to EL91 valves, would using the 15ohm tap and the original feedback resistor value go some way to achieving this, or would it make the phase splitter gain too linear (remember I'm trying to retain the original distortion characteristics but adapt the circuit for use with EL91 valves as opposed to 6L6)


Thanks,

Jonathan

[Herald1360]

There seem to be two conflicting problems here...... negative feedback which will attempt to reduce distortion wherever it is generated within the parts of the amplifier inside the feedback loop and a requirement for the stage driving the output valves to go into clipping, presumably before it overdrives said output valves. The resultant sound will be somewhat unpredictable but may serendipitously result in a nice noise rather than an 'orrible one.



Don't many (most?) valve guitar amps eschew NFB anyway? I would think it would make the onset of distortion (nice or nasty) rather unpredictable. Maybe that's why apparently similar designs can sound so different musically.



Perhaps the easiest way to get phase splitter clipping at a lower output level than required for 6L6s would be to run the phase splitter at a lower HT voltage. Easy enough with a suitably decoupled dropper resistor in the HT line feeding the stage. Might need a fair amount of empirical effort to arrive at what you want soundwise.

[Syrinx1]

Thanks Chris,

I have considered doing this, and also substituting for a lower gain valve in the phase splitter. I think some experimentation is in order!

[Diabolical Artificer]

A different OPT will change any characteristics as will the OP valves surely, the end product is unlikely to sound the same. you could try reducing the anode resistors of the PS or as you say change an 83 for an 81 or 2, you'd be best drawing a load line and doing some calcs.

When you come to apply NFB, use a 10k pot instead of a fixed R, adjust to taste.

Andy.

[Paul JD]

Originally Posted by Herald1360

Don't many (most?) valve guitar amps eschew NFB anyway?

No - there are almost certainly more guitar amps that use global negative feedback than those that don't.

[ColinTheAmpMan1]

Originally Posted by Paul JD

No - there are almost certainly more guitar amps that use global negative feedback than those that don't.

There is a circuit that uses a series pot (~1 Meg) in the negative feedback loop to reduce the negative feedback. It can be called a "mojo" control in some circuits. Also, where a negative feedback loop exists, the negative feedback can be reduced to almost nothing by the use of a "presence" control. Such a control also reduces the negative feedback at the cathode of some earlier valve, but is somewhat more effective than the "mojo" control. You "pays yer money and gets yer choice".

Colin.

[Syrinx1]

Hi Colin,

Thanks - definitely something for me to look into!