In the shematic of Tubelab there is no GNFB to the first stage, only the cathode feedback to the power stage. See the remark in the schematic about finding out what is the right way (polarity) of the secondary. I don't think it makes any difference for correct polarity/phase that Tubelab only connects the cathode capacitor to the secondary.
The following schematic is of a little amplifier I made, using output transformers I salvaged from two Philips EL3516 tape recorders in which cathode feedback was applied to the power tube. As you can see, the phasing of the cathode feedback to the power tube is opposite of that of the GNFB applied to the preamp tube (grounding takes place on the tap of the secondary). In that way both types of feedback are negative (but like in the amplifier of Tubelab, the polarity of the secondary has to be the right one, else both types of feedback would be positive).
The following schematic is of a little amplifier I made, using output transformers I salvaged from two Philips EL3516 tape recorders in which cathode feedback was applied to the power tube. As you can see, the phasing of the cathode feedback to the power tube is opposite of that of the GNFB applied to the preamp tube (grounding takes place on the tap of the secondary). In that way both types of feedback are negative (but like in the amplifier of Tubelab, the polarity of the secondary has to be the right one, else both types of feedback would be positive).
Last edited:
With due respect: IMHO it makes a lot of difference if only the capacitor or the capacitor and cathode resistor are attached to the OPT. I base this opinion on extensive modelling I've done with a 7591. YMMV with other tubes.
I didn't (mean to) state that it doesn't make any difference. I just thought that it doens't matter for polarity/phasing.
Thanks for sharing, looks like a nice little amp. I'm going to give this methodology a go in my next amp if the performance without GFB is unsatisfactory. Will try using just CFB first and add GFB if needed.
I spent a lot of time on my current amp testing the output transformer to determine the phase of the windings and correct configuration for CFB. The next one will be easy! Just wish I had a 16 ohm tap so the feedback factor could be a bit higher. As it is I only have 4 and 8 ohm taps and no option to add another winding (sealed toroidal xformer).
Thanks again,
Greg
You are right that it does not matter for polarity/phasing.
Basically a capacitor is high pass and with the speaker load it may give unwanted resonances. Plus I like the DC coupled to the secondary winding so the very lowest frequencies have feedback too.
Perhaps there is a reason tubelab uses 1000uF and has experienced instability with some OPT?
I had communication with the late Patrick Turner on this subject about if the primary is 5000 and the secondary 8. His response was "don't bother, it's not worth it". I'm using 3000 and 16 and added UL in the mix.