Thinking about using a Tentlabs negative bias module in combination with Rod Coleman's Filament Supply, to fixed-bias and dc-power the filaments of a DHT.
The Tentlabs module would provide the -85V to bias the grid, while Rod Coleman's module would supply the 5v for the filaments (see schematic below).
I am thinking about how I could make the 2 work together..
The Tentlabs module would provide the -85V to bias the grid, while Rod Coleman's module would supply the 5v for the filaments (see schematic below).
I am thinking about how I could make the 2 work together..
Last edited:
Did you check out the tubelab SSE_II and the Bartola Valve 300b schematics?
Those would be good starting points.
Those would be good starting points.
Why do you use virtual ground for IDHT connected to cathode?
Filament elevating (heater potential -about- 20-40v over the cathode) impossible in this case.
IHMO using R.C regulator (and individual floating PSU and transformer) for IDHT totally unnecessary.
If hum occurs problems (in my practice only in phono stages), simple DC filament PSU (graetz, C-R-C) is sufficient.
This is my 300B amp (CCS loaded, LED biased D3a, capacitor coupled to source follower which is DC coupled to 300B, 5k:8, 10W 2.21%THD) AC filament feed D3a VAS stage output at 70V RMS (198Vpp).
No disturbing hum, no sidebands near carrier or distortion harmonics.
Filament elevating (heater potential -about- 20-40v over the cathode) impossible in this case.
IHMO using R.C regulator (and individual floating PSU and transformer) for IDHT totally unnecessary.
If hum occurs problems (in my practice only in phono stages), simple DC filament PSU (graetz, C-R-C) is sufficient.
This is my 300B amp (CCS loaded, LED biased D3a, capacitor coupled to source follower which is DC coupled to 300B, 5k:8, 10W 2.21%THD) AC filament feed D3a VAS stage output at 70V RMS (198Vpp).
No disturbing hum, no sidebands near carrier or distortion harmonics.
hi Marco -
The power stage
Notes:
300B pin 1 should be the Negative side.
I assume that the Tentlabs module has a connexion to 0V(anode supply), an that its vk_2 input has a low value resistor to this 0V power pin.
In this case, it will be compatible with the filament supply.
The Driver stage:
Will you use the heater current to bias a low-value cathode resistor? In this case it's a good use of the low noise of the V9 regulator; otherwise save the V9 for another project.
Thank you for the suggestions @Rod Coleman.
I updated the schematic to be sure I connect the - from the filament supply to the cathode reference voltage.
And to supply the "raw DC" to the filament regulator I am thinking about using this topology (one supply for each tube) which should produce quite clean 2A at 8V on the output. (I might leave space for a common mode choke at the output as well).
Is it an overkill?
Regarding the IDHT..I still haven't decided what fixed-bias circuitry to use and I am thinking about some sort of modular implementation where I can change the bias topology by moving some jumpers.
Assuming I don't use the filament bias for the IDHT, do you think at that point a super clean filament supply (ripple <1mV) would not make any audible difference ?
Thank you for your precious suggestions!
Marco
I updated the schematic to be sure I connect the - from the filament supply to the cathode reference voltage.
And to supply the "raw DC" to the filament regulator I am thinking about using this topology (one supply for each tube) which should produce quite clean 2A at 8V on the output. (I might leave space for a common mode choke at the output as well).
Is it an overkill?
Regarding the IDHT..I still haven't decided what fixed-bias circuitry to use and I am thinking about some sort of modular implementation where I can change the bias topology by moving some jumpers.
Assuming I don't use the filament bias for the IDHT, do you think at that point a super clean filament supply (ripple <1mV) would not make any audible difference ?
Thank you for your precious suggestions!
Marco
Marco,
I think having the 1085 reg defeats the purpose of the Rod Coleman approach. Better to get Rod's raw supply boards which are harmless crc filters and leave plenty of extra voltage to burn down to the 5V DC or whatever is required by your DHT.
I think having the 1085 reg defeats the purpose of the Rod Coleman approach. Better to get Rod's raw supply boards which are harmless crc filters and leave plenty of extra voltage to burn down to the 5V DC or whatever is required by your DHT.
@grataku I think you are right..
I started with "let me draw a PCB for a simple filer" and then from there the step to throw in a regulator and and inductor was short 😅
But you are right.. a linear regulator is an overkill and dissipates too much power
I started with "let me draw a PCB for a simple filer" and then from there the step to throw in a regulator and and inductor was short 😅
But you are right.. a linear regulator is an overkill and dissipates too much power
It's not that a linear regulator is overkill per se. The 1085 regulator might as well be a noise generator. You will actually have more noise with that regulator than you would with a sufficiently sized crc filter.
Hi Marco - yes, please use CRC for the Raw DC.
Another note:
Cgnd (220n): this capacitor is effectively in parallel with the Cathode bypass capacitor. But if it is mounted in the Raw DC assembly, some audio current will run from the cathode, and up to the GND point where the cap is located. So it's best to avoid his large cap.
But if you would like to experiment with radio frequency decoupling of the Raw DC, you could try a 470pF FKP capacitor in series with 3 - 10Ω Resistor. This will be enough for mains-power RF suppression in most cases. Confirm with listening test.
Another note:
Cgnd (220n): this capacitor is effectively in parallel with the Cathode bypass capacitor. But if it is mounted in the Raw DC assembly, some audio current will run from the cathode, and up to the GND point where the cap is located. So it's best to avoid his large cap.
But if you would like to experiment with radio frequency decoupling of the Raw DC, you could try a 470pF FKP capacitor in series with 3 - 10Ω Resistor. This will be enough for mains-power RF suppression in most cases. Confirm with listening test.
Thank you for the suggestion @Rod Coleman , I will update my schematic as you suggested.
In the meantime I had an idea for biasing the IDHT.
I would use filament bias but with a large LED on the cathode instead of a resistor. Why?.. because with a large LED, the differential resistance dV/dI can go as low as 0.5 Ohm @320mA (see the graph below), which is more than an order of magnitude lower than the 8Ohm resistor I would need to obtain the same Vk_bias.
That should give me a very quite Vk_bias, and minimize the feedback through the cathode.
Have you ever tried it before?
In the meantime I had an idea for biasing the IDHT.
I would use filament bias but with a large LED on the cathode instead of a resistor. Why?.. because with a large LED, the differential resistance dV/dI can go as low as 0.5 Ohm @320mA (see the graph below), which is more than an order of magnitude lower than the 8Ohm resistor I would need to obtain the same Vk_bias.
That should give me a very quite Vk_bias, and minimize the feedback through the cathode.
Have you ever tried it before?
I have not tried it yet. But if you build it just as you describe, it will be easy to compare the sound against an unbypassed resistor.
LEDs should be protected from domestic lighting, as the led works in reverse, to some degree, and lighting can be flickery.
LEDs should be protected from domestic lighting, as the led works in reverse, to some degree, and lighting can be flickery.
Thank you for the suggestion @Rod Coleman, maybe I could install one of those EMI shields to cover the LED from external light.
And yes, I am going to use a jumper (or a dip switch) to quickly switch from LED bias to R bias so I will be able to do a blind comparison.
I am very excited about this project, I hope it will sound well
And yes, I am going to use a jumper (or a dip switch) to quickly switch from LED bias to R bias so I will be able to do a blind comparison.
I am very excited about this project, I hope it will sound well
LED biasing is very impressive, I used it usually... but only with -almost- constant current (CCS or gyrator loaded stage).
The current is limited (few-few ten mA).
I use -very- old 5mm diameter LEDs.
The red has the lowest dynamic impedance.
The current is limited (few-few ten mA).
I use -very- old 5mm diameter LEDs.
The red has the lowest dynamic impedance.
Below I reported the schematic of the solution I am thinking to use. That LED has an impressively low dR= 0.5 Ohm.
I am also leaving the option to use a resistor instead of the LED (by moving the jumper J1) to compare the 2 solutions.
I am also leaving the option to use a resistor instead of the LED (by moving the jumper J1) to compare the 2 solutions.