How would you improve this EL84 SE circuit?

[4season]

This little amp by Tecon Audio looks nice, but I'd like to improve the sonics, which seem kind of lean and constricted to me. Can anyone suggest changes? If possible, I'd like to stick to running the tubes as pentodes, but am open to suggestions.

[bacon665]

The cathode on both of these is attached to the ground
Im not sure where they are getting the power for the heaters but...
I will suggest a set of opposing diodes (in series) to keep ac from the input and anywhere else its rejected to ground from entering the heater/cathode and causing harmonics distortion ect ect.

Maybe you could get more suggestions if you spit out what you would like suggestions on changing?

And the whole reason the 6BQ5 was developed was to remove the need for a triode preamp...

Also i dont know what this is powering BUT
if you figure out a replacement for the resistor next to C14 you can raise it to 6 volts then use a second 6BQ5(EL84) in triode mode and increase your power... (slightly)

[aardvarkash10]

source device is? Speakers are?

Simplest first step is to try pentode / ul / triode options on the output tube. Don't need ot change anything and easily reversable.

[richwalters]

The cathode resistor in the o/p stage: I read 15 Ohms. Surely a misprint ? This is way too low; should be around 470R.
The bandolier of C7-C10. This isn't a switchmode power supply. A class A amp shouldn't need all this and a single can of accumulative value should be more than adequate.
The screen resistor of the 6BQ5. I would never omit this..A 270R or higher in series is the done thing.
Can't see what priorities the designer had in mind:but for true pentode, a bit of global nfb may sparkle....but a redesign.

richy

[jjman]

Can someone teach me the function (and "name") of the 470k on the plate of V2 ? Some sort of limiter or anti-oscillation widget?

[chrish]

Originally posted by jjman
Can someone teach me the function (and "name") of the 470k on the plate of V2 ? Some sort of limiter or anti-oscillation widget?

I am just an apprentice here, some of the wise masters will be able to give some advice, but as far as I can work out, it appears to be some form of 'plate to plate' feedback. I believe the RH84 uses a similar feedback mechanism. Here we go, RH84 link , might be worth a look for this...

Cheers,

Chris

[Sheldon]

That amp is UL connected. The 470k resistor provides some plate to plate feedback. You might want to try it pentode connected and play with the value of the feedback resistor. Also, that 10k load on the driver seems low. Look at this for reference: http://www.diyaudio.com/forums/attac...amp=1239702584

And search on RH Amp for examples. Many builders here like it.

Sheldon

[Michael Koster]

Originally posted by Sheldon
That amp is UL connected. The 470k resistor provides some plate to plate feedback. You might want to try it pentode connected and play with the value of the feedback resistor. Also, that 10k load on the driver seems low. Look at this for reference: http://www.diyaudio.com/forums/attac...amp=1239702584

And search on RH Amp for examples. Many builders here like it.

Sheldon

That's my take on the circuit also.

And I think Sheldon's suggestions would indeed be interesting experiments. Pentode mode could give you more output power depending on the current OP and Zpri (which is hard to determine because the value of the EL84 cathode R is in error). Changing the 470K reedback resistor would give you control over output impedance etc. in pentode mode and you may not need UL feedback in addition.

Edit:
The 5755 is only good for 1W plate dissipation. You would need parallel sections and/or an increase in the plate load resistor value. 10K must be a mis-print. You would need 470K or greater assuming one section in order to operate at below .7 mA and 140V

If you change the feedback resistor you may need to change the plate load resistor also to keep below 1W.

It will be interesting to see if you get closer to the sound you're looking for. A lot will depend on the speakers you're driving.

Michael

[Wavebourn]

Use the same EL84 for driver, but pentode connected, otherwise you are getting huge triode distortions loading it on a parallel feedback.

[agent.5]

Originally posted by Wavebourn
Use the same EL84 for driver, but pentode connected, otherwise you are getting huge triode distortions loading it on a parallel feedback.

Do you think it may be better to just eliminate the 470K feedback all together, if 5755 has to be the driver tube as the circuit sits on a PCB and major surgery is difficult although possible?