Like the other poster above, I also just connect the two heater windings of the Toroidy transformer direct to the heaters, one winding for each board. It is perfectly quiet, but I have not grounded the heaters or referenced them to ground in any way, and I do need to do that.
The cathodes of the driver tubes are at -60Vdc, so I suppose elevation is not that important? Or is there a good case for doing that?
I was thinking just a capacitor to ground from one of the wires of each winding. Is there anything to be gained referencing with a 100R hum pot instead?
What you are measuring is falsed by the instrument input impedance: a 12AX7 biased with the cathode at -60Vdc would have zero current flowing through.
I usually keep driver's cathodes around +2V, while with standard BH-EL34 configuration they are too low at around +0,8 V, in grid current zone for a 12AX7.
I use something around 1Meg to 100k||1u from my B+ that is 460V getting around +42Vdc to elevate heaters.
Hi @zintolo , thanks for the reply. Yes, I had not thought through what I was trying to say there. What I was thinking was that there are no tubes with cathode potentials significantly above ground, e.g. due to direct coupling, so elevation is a nice to have rather than necessary.
What voltage rating do you use for the 1u cap?
I was worrying that if I implemented a solution like the one you describe, there are two separate ground connections due to the two heater windings/circuits, and maybe a source of hum due to a loop?
What voltage rating do you use for the 1u cap?
I was worrying that if I implemented a solution like the one you describe, there are two separate ground connections due to the two heater windings/circuits, and maybe a source of hum due to a loop?
@OldHector the "1Meg to 100k||1u" is in parallel with the B+ PSU capacitor, grounding there.
I use the pcb version without external PSU board (MK I, highly inspired to the EL84 version), but if you have it you can simply use one elevated reference for two ore more winding. No heater current flowing through the voltage elevator.
Having 40V across that cap you can use a 63V film cap, but I use 100V film caps as I have many.
I use the pcb version without external PSU board (MK I, highly inspired to the EL84 version), but if you have it you can simply use one elevated reference for two ore more winding. No heater current flowing through the voltage elevator.
Having 40V across that cap you can use a 63V film cap, but I use 100V film caps as I have many.
One file that I need to find again (meybe others have done it as well) is the drilling template for supports and tubes, as they are not centered into the pcb.
If someone else has it (specifying the pcb version), it would be helpful for many. If I'll find I'll post it.
If someone else has it (specifying the pcb version), it would be helpful for many. If I'll find I'll post it.
Hi Zintolo,
Wanted to reply regarding these changes after spending some time listening on my main system. They do drastically change the sound, in a few words: more controlled. The amp is generally a bit quieter, I'm having to turn the volume up higher than I did previously and still find I want to turn it up further. One thing which impressed me when I first powered on this amp with my main speakers was the great sense of dynamics, I think that has been lost a bit in favor of a more solid-state sound. Overall the amp now exhibits this (solid state) sound profile with a bit of warmth. I do have a nice solid state amp already, so going to look at reverting some of the above mods. Please chime in if anything specific should be reverted first, or any other changes that might be recommended.
I did see that you are running a higher B+ voltage, of 460V. Mine is at 385V. Wonder if this might be cause for what I'm hearing.
For sure, yes. First thing, go back to the original current through the phase inverter.
Changed R18 to original value of 680ohms. I am getting 1.2V across R18. Dynamic quality I remember is restored, however finding bias doesn't hold well. The same with balancing R5. In both cases I adjusted initially, then again after it played for few hours. Then off, on after a day or two. Checked values after it played about 30 minutes, not where they should be ( getting 1.2V between C1 & C2 inner legs), bias about 10ma lower than where they should be, and 10ma difference between the two boards.