I would imagine if executed with SemiSouth's you need no more than 2V for conduction.
Here is an update of my schematic and its Bode plot. You should note the 2 pole RC filter supplying the bias voltage for the upper output FET. This significantly improves the low frequency THD. The high frequency response shown agrees well with actual measurements. I cannot really test the low frequency rolloff. This implementation does not suffer from the low frequency oscillation problem that my previous build exhibited.
Also shown it the THD plot of the rewired circuit. It is interesting to note the at low frequencies the 3rd harmonic dominates; at high frequencies the 2nd harmonic dominates. The crossover it at about 700 Hz. (I really need to do a Bode plot of the harmonics separately.)
Also shown it the THD plot of the rewired circuit. It is interesting to note the at low frequencies the 3rd harmonic dominates; at high frequencies the 2nd harmonic dominates. The crossover it at about 700 Hz. (I really need to do a Bode plot of the harmonics separately.)
Attachments
You show the realtionship between V- and the lower bias circuit to be identical to the relationship between Vo and the upper bias circuit. Will this shown high symmetry give a highly symmetrical performance before loop feedback?
now choose Rs value & Iq , and calculate that in
ZM will allow at least 1mA through trimpot , and 3mA through zenner
in fact - ZM will use two 3mm greenies (meaning 2x1V93@3-5mA) in series instead - more shiny
and , after that ,ZM will use CCS-es - feeding voltage references , just for fun
why simple , when it can be complicated
Last edited:
Speaking of complicated. THD seemed to improve on Ilquam's version when he added the two pole filter. Does this suggest that a regulated supply might best current version in terms of performance. I was going to try 2 LED's today.
lhquam: Your refined circuit like its parent still has the following observations:
- The capacitor connected to the lead of the bottom transformer [circle] a dead short to AC ground [or V-] under all operating conditions.
- The lead of the transformer's upper winding [no circle] which is connected to a same-valued capacitor is not [never] at AC ground. The impedance from this upper transformer aiming at ground is the sum of the resistances and through the upper bias regulator to its official ground.
If one insists on exact matching, then the answer maybe found [or not] in the schematics of the old classic parent[ with bjts] of Conceptual F6.
Would you consider simulating the old classic design using LTSPICE as a comparative reference?
No, regs for bias and maybe FE. If 2 pole filter on bias circuit improved LF distortion, doesnt that suggest riplle/noise is getting through simple RC and causing distortion.
hell , no ;
CCS instead of 6k8 resistor(s) , in bobbo's sch
you'll make a big badaboom with sch in your pdf
see - zener is already reg ........... (shunt one )
LED also .........
feeding it with CCS is just making it better
audibly better - probably not
but , I can't resist - Babelfishing is my hobby
I am not sure what the relevance is for "AC ground". What is important are the bias voltages relative to the bottom of the FET source resistors. We are trying to maintain these voltages at constant levels under varying signal conditions.
Ok. It did seem funny that the Led's were reversed. Should take more time to think vs just copy😀