Hi all
I tried to make a class A amp which would sound better than my current power triodes amp, about 15 to 20 watts.
I tried first ZV9, had a nice amount to problems to solve and finally got it right and could listen to it. I found out that theoretically it was a nice piece but IMHO really forcing the coupling cap sound weakness on the final result. Also, lots of cooling needed, making it complex and cumbersome. So away it went.
Than I tried Zenevo, sort of clone with some npn/pnp drivers. Results were disappointing, with agressive sound. Away it went.
I have now finished an F4 and use a one only N mosfet / P mosfet pair, the textbook IRFs. Before I connect it to the speakers I find that the output offest drifts from +50 to - 50 for some hours and eventually steadies around +/- 20 mV.
My question is : is that issue known and what would be the answer ? I precise that the offsett when cold is as high as 1.5 V+/-. Not good for speakers !
I feed the F4 with opamp driver, with much better signal caracterisitcs than the BOZ, BOSOZ , Sym BOZ I all tried. They get hot, have limitations in signal amplitude and all show to have an output signal which crumbles down when interfaced to the F4.
Hence, here is my second question: the static input impedance of a F4 with FET inputs is what you decide with the entry Res. But the dynamic input Z is FAR lower (some 8 to 10 K) and that makes a mess with the drivers. None of those drivers have a signal which can handle easily this inp Z, according to all the tests I tried. When not connected they are all perfect or almost perfect, trouble starts when there are plugged to the F4 (or the others I made, they all had the same behavior). How come this point is not really ever mentionned in all F4 or others class A litterature ?
Thanks for your opinions !
I tried to make a class A amp which would sound better than my current power triodes amp, about 15 to 20 watts.
I tried first ZV9, had a nice amount to problems to solve and finally got it right and could listen to it. I found out that theoretically it was a nice piece but IMHO really forcing the coupling cap sound weakness on the final result. Also, lots of cooling needed, making it complex and cumbersome. So away it went.
Than I tried Zenevo, sort of clone with some npn/pnp drivers. Results were disappointing, with agressive sound. Away it went.
I have now finished an F4 and use a one only N mosfet / P mosfet pair, the textbook IRFs. Before I connect it to the speakers I find that the output offest drifts from +50 to - 50 for some hours and eventually steadies around +/- 20 mV.
My question is : is that issue known and what would be the answer ? I precise that the offsett when cold is as high as 1.5 V+/-. Not good for speakers !
I feed the F4 with opamp driver, with much better signal caracterisitcs than the BOZ, BOSOZ , Sym BOZ I all tried. They get hot, have limitations in signal amplitude and all show to have an output signal which crumbles down when interfaced to the F4.
Hence, here is my second question: the static input impedance of a F4 with FET inputs is what you decide with the entry Res. But the dynamic input Z is FAR lower (some 8 to 10 K) and that makes a mess with the drivers. None of those drivers have a signal which can handle easily this inp Z, according to all the tests I tried. When not connected they are all perfect or almost perfect, trouble starts when there are plugged to the F4 (or the others I made, they all had the same behavior). How come this point is not really ever mentionned in all F4 or others class A litterature ?
Thanks for your opinions !
show me exact schm of your F4 , and pictures
there is no way that input buffer of F4 will trouble poor opamp
anyway , how'z that tubehead is trusting to OP , to be good enough ?
there is no way that input buffer of F4 will trouble poor opamp
anyway , how'z that tubehead is trusting to OP , to be good enough ?
Hi Zenmod
the schematic is the regular one from NP in his 'F4 prod manual'. the input buffer does NOT trouble the Opamp, but all the discrete Bozs and others I tried.
The triode set is based on 4 pcs 6080, it was published some years ago by Hiraga. Really gorgious. IMHO can compare with 300Bs (especially chinese modern ones).
Before going back to make a triode preamp with a medium power buffer, I will first solve the offset drift issue. I am very surprised that the F4 input needs a low Z ouptut source.
It can't be a matter of matching the IRFs since I chose to use only one pair (N+P) per channel. The drift was the same with one or two (N+P) pairs.
ImageShack® - Online Photo and Video Hosting
the schematic is the regular one from NP in his 'F4 prod manual'. the input buffer does NOT trouble the Opamp, but all the discrete Bozs and others I tried.
The triode set is based on 4 pcs 6080, it was published some years ago by Hiraga. Really gorgious. IMHO can compare with 300Bs (especially chinese modern ones).
Before going back to make a triode preamp with a medium power buffer, I will first solve the offset drift issue. I am very surprised that the F4 input needs a low Z ouptut source.
It can't be a matter of matching the IRFs since I chose to use only one pair (N+P) per channel. The drift was the same with one or two (N+P) pairs.
ImageShack® - Online Photo and Video Hosting
An externally hosted image should be here but it was not working when we last tested it.
- Status
- Not open for further replies.