F6 Amplifier

[lhquam]

Here is my guess at the F6 design, built with Semisouth R100s and the Jensen JT-123-FLPCH transformer.

The first scope photo shows response to a 20kHz square wave.
The second scope photo shows response to a 20Hz square wave.
In both cases, the top trace is the input after the input coupling capacitor and the F6 bottom is the output.

Posts 890 and 902 show THD measurements. Ignore the frequency sweep in post 890, it is corrected in post 902.

The R100 bias currents and output offset voltage are rock-solid stable. The positive temperature of the R100 means that thermistors are not needed for temperature stabilization. When cold, the bias currents start slightly higher than when at final running temperature.



I do not know how well this performs vs. Nelson's "official F6".
I haven't listened to it yet. I first plan build a second channel for stereo.

[Antoinel]

Originally Posted by banat

Antoinel

Here is the link to DIY Atma-Sphere M60 OTL Circlotron tube amp , there you can find the schematic ,altogether with many valuable blah - blah stuff , ( including mine blah - blah stuff ) : What tubes for a tube amp?

BTW , if you want to make bias network for F6 Amp in the same way as is done on SEPP OTL Tube amp from mine previous attached schematic than .....
It is just my personal opinion that is important thing to succesfully DIY F6 SS SEPP amp to find Right relative main ( 60 Hz) AC phase status between this two equal independent & floating bias secondary coils on main PSU transf.
By reversing relative AC phase of this two independent secondary bias coils you can worse or improve Amp output power stage (SEPP) PSRR significantly .
Same is valuable for relative relation between output power stage (SEPP) PSU secondary power coil AC main phase and this two independent bias coils AC phases .

Best Regards !

Thank you banat. I got it. Is the underlined saying thats the AC mains power plug [or maybe the mains power transformer] to Conceptual F6 needs to to be inserted/wired one way only, and not the other for maximizing PSSR?. This is an unexpeted find.

Best regards.

[Antoinel]

lhquam; Your schematic is simple and effective. No frills.

[Loudthud]

Originally Posted by lhquam

The first scope photo shows response to a 20kHz square wave. The second scope photo shows response to a 20Hz square wave. In both cases, the top trace is the input after the input coupling capacitor and the F6 bottom is the output.

Why is the output inverted with respect to the input?

[Zen Mod]

Originally Posted by lhquam

Here is my guess at the F6......

upper half bias doesn't need to go to gnd , but output rail

besides that , try to implement "those" few things in second iteration , then hear and measure difference

[lhquam]

Originally Posted by Loudthud

Why is the output inverted with respect to the input?

Oops, my mistake. I inadvertently had the bottom scope channel invert switch depressed. I had been using it to do a differential voltage measurement and forgot to switch it off.

[lhquam]

Originally Posted by Zen Mod

upper half bias doesn't need to go to gnd , but output rail

besides that , try to implement "those" few things in second iteration , then hear and measure difference

When connected to the amplifier output the offset voltage would not stabilize. When it is connected to ground it is totally stable.

My theory is the following:

• Let X be the bottom connection of the upper bias circuit.
• The capacitor C5 from the secondary to the gate stopper resistor integrates the current supplied thru the bias injection resistor R12. The voltage across C5 should be the offset voltage of the bias circuit plus the average voltage of X.
• The average value of the output should be zero. Therefore we can connect X to ground.
• This seems to work well, but depends on the RC time constant of C5 and R12 being large enough. However, spice simulations show better results at low power levels with X connected to the output.

[Zen Mod]

sorry

I neglected that you're having fixed voltage across LED's (unlike as in case of using just resistors)

in this case is certainly better as you drew it

[banat]

Originally Posted by Antoinel

Thank you banat. I got it. Is the underlined saying thats the AC mains power plug [or maybe the mains power transformer] to Conceptual F6 needs to to be inserted/wired one way only, and not the other for maximizing PSSR?. This is an unexpeted find.

Best regards.

If the F6 both bias DC sources is done in Unregulated way same as is done on that SEPP OTL tube amp schematic than is important to take care of the mutual phase relation between this three PSU transf. secondaries , regardless if that three secondary coils is on the common main power transf. or not ,
since the some minor rests of the main AC ( 60 Hz) ripple can appear on both independent bias DC lines and can further to be Amplified or Canceled on SEPP output power stage .
That thing depends directly from way how you connected ( phased ) three secondaries on the associated rectifiers .
By pure SEPP Amp or Julius Futterman SEPP Amp variants this Right main AC secondary phasing is important for unconditional stable Amp operation without of any output signal hum .

Best Regards !

[Antoinel]

banat: Thank you for the clarification. Best regards.