Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: astronomicum on July 07, 2026, 02:11:11 pm
-
Univox used a form of feedback circuit to increase the cathodyne preamp and cathodyne gains in a number of their models. I have a U152 on the bench. I have attached the U-152 schematic where I have circled the circuit in question. I would imagine a few of you have seen this circuit before. I have documented the AC gain of this circuit on varying Univox models noting a significant AC gain increase of 50% or better in the models using a 3K or larger cathode resistor producing a signal gain averaging better than 100. I am measuring a signal gain of 105 in this particular amp.
Deciding to study this circuit more, I became puzzled by the fact that this circuit feeding AC from Anode to Cathode through the two caps would be negative feedback with, normally, an expected observation of a reduction in gain. What I had not looked at before were the AC signals at both the Anode and Cathode on a dual trace Oscilloscope (see attached pic). What I am seeing are signals appearing to be in phase :w2:. Just to be sure what I was seeing was accurate, I did the same test on a normal gain stage and the scope does show the signals 180 degrees out of phase as expected.
Am I right to deduce that the negative feedback is so strong that it actually flips the cathode phase? Would there not be a phase shift in the feedback signal after passing through two caps? What else I am missing?
-
This seemed like an interesting problem so I decided to model it. I can confirm the anode and cathode of the first triode are in phase, but I don't come close to understanding it. It seems to get negative feedback from at least three places: degenerative feedback from the unbypassed cathode, AC feedback from the split-load node at the anode, and AC feedback from the cathode of the following stage. Though I didn't model it, it looks like the Univox also gets global negative feedback from the OT secondary.
The technique of connecting the split-load node to the cathode of the following stage reminds me of Merlin's "bootstrapping" technique: https://valvewizard.co.uk/dccf.html (see the last paragraph), which I also don't begin to understand.
-
This seemed like an interesting problem so I decided to model it. I can confirm the anode and cathode of the first triode are in phase, but I don't come close to understanding it. It seems to get negative feedback from at least three places: degenerative feedback from the unbypassed cathode, AC feedback from the split-load node at the anode, and AC feedback from the cathode of the following stage. Though I didn't model it, it looks like the Univox also gets global negative feedback from the OT secondary.
The technique of connecting the split-load node to the cathode of the following stage reminds me of Merlin's "bootstrapping" technique: https://valvewizard.co.uk/dccf.html (see the last paragraph), which I also don't begin to understand.
Passaloutre, Thank you for the confirmation. I can confirm that there is also a gain increase in the cathodyne so your observation on the bootstrap appears to be valid. I have measured about a 10% increase in gain (closer to 1) in this arrangement.
What about there being phase shifts in the AC feedback? The feedback from the split load is going through two caps, the feedback from the cathodyne cathode through one. Would there not also be the same shift in the bootstrap? Going to investigate in a bit.
Edit: Update. Not seeing a shift on the scope.
-
It's bootstrapping, from the cathode back to the anode split load.
It's positive feedback.
As Merlin describes, it increases the effective value of the anode load resistor.
The cathode to cathode feedback is negative, probably intended to improve linearity.
-
It's bootstrapping, from the cathode back to the anode split load.
It's positive feedback.
As Merlin describes, it increases the effective value of the anode load resistor.
The cathode to cathode feedback is negative, probably intended to improve linearity.
Thanks. I can see the boot strap on the cathodyne and the lower value of the anode load resistance used demonstrates the effect of the bootstrap increasing the effective value of the anode load. It would seem to be way out of balance with the 43K (50K ll 330K) anode and 101K cathode resistance but it is in balance (within 1V). Drifting resistors over many years does effect balance and I find I need to adjust values from time to time to bring them back in balance.
Any insight into the flipping cathode phase on the gain section? By all intents and purposes, the arrangement is producing positive feedback like the bootstrap but using negative feedback to flip the cathode phase?
One other observation is that whatever "is" happening with the gain stage cathode, it is apparent that there is some distortion produced as evident in the cathode wave form. Looking at the oscilloscope traces, the anode looks nice and clean but we are looking at a 45V wave form vs. a half volt trace from the cathode so it is not apparent in the anode trace and most likely makes little difference in the scheme of things.