@andyjevans
The PCBA I posted above is amplifier stage, not the power regulator.
I had a look at Rod Coleman's circuits. They are not really regulators, they are rectifiers with AC filtering. They probably to the job, but they are not very flexible because they are not really regulated so the Vout would change if you change the load.
I might end up using them or designing my own regulators.
By the way, my statement in the previous comment was incorrect: the amplifier that I am building will use a 4852 to drive the 300B, which is not DHT.
But I want to make the design modular so I can later on make a v2 with a DHT gain stage. I will probably end up splitting the PCB in 2 independent PCBAs, one for each stage, so I can swap the gain stage while keeping the same power stage.
I am relatively new to tube amp design, but I have long experience in electronics and I can have PCBs built very quickly and at low cost, so I don't mind building a few difference boards and play with them as I learn more about tube amplifier design.
The PCBA I posted above is amplifier stage, not the power regulator.
I had a look at Rod Coleman's circuits. They are not really regulators, they are rectifiers with AC filtering. They probably to the job, but they are not very flexible because they are not really regulated so the Vout would change if you change the load.
I might end up using them or designing my own regulators.
By the way, my statement in the previous comment was incorrect: the amplifier that I am building will use a 4852 to drive the 300B, which is not DHT.
But I want to make the design modular so I can later on make a v2 with a DHT gain stage. I will probably end up splitting the PCB in 2 independent PCBAs, one for each stage, so I can swap the gain stage while keeping the same power stage.
I am relatively new to tube amp design, but I have long experience in electronics and I can have PCBs built very quickly and at low cost, so I don't mind building a few difference boards and play with them as I learn more about tube amplifier design.
Last edited:
@marco,
No no no no! 😉 Rod Coleman has two parts, the reg rectification and filtering, which is what you saw, AND a second board with the real deal regulator. The schematic for that is proprietary. Don't tell Rod, but these things are dirt cheap for what they do and how they sound.
These have undergone intensive updates and optimizations through the years. You cannot design a better reg yourself. I guarantee it.
No no no no! 😉 Rod Coleman has two parts, the reg rectification and filtering, which is what you saw, AND a second board with the real deal regulator. The schematic for that is proprietary. Don't tell Rod, but these things are dirt cheap for what they do and how they sound.
These have undergone intensive updates and optimizations through the years. You cannot design a better reg yourself. I guarantee it.
I converged on this SET design from Andrea Ciuffoli, with the addition of fixed bias to both stages.
For the driver, I am planning to implement a filament bias topology through a high-current led to minimize the signal feedback through the cathode (see discussion).
The 300B bias will be configurable through a potentiometer on the PSU board.
Mod Edit: Removed schematic at User request
For the driver, I am planning to implement a filament bias topology through a high-current led to minimize the signal feedback through the cathode (see discussion).
The 300B bias will be configurable through a potentiometer on the PSU board.
Mod Edit: Removed schematic at User request
Last edited by a moderator:
Both stage with fixed bias?
I see only one, the power stage also with 1kohm + 1kohm of cathode of 300B!!!!!!!!!
Walter
I see only one, the power stage also with 1kohm + 1kohm of cathode of 300B!!!!!!!!!
Walter
Where is the fixed bias on first stage?
Then, the Rg3 is too high looking on data sheet of 300B.
I use 100k max without any issue
Last, on dc regulator is written “ don’t connect the psu to ground”
Both are connected !!
Andrea was a my friend ( I lost the contact) here in Roma
Many times I gave him some high level components
My opinion:
Every Dht sound better with ac supply on filaments ( and the life is longer)
Walter
Walter,
there is NO evidence that AC heating prolongs the life of the tubes. I don't even know how you would go about rigorous testing of that on the lifetime of a good 300B like a Takatsuki or WE. Plus you are at the mercy of the power company mains fluctuation.
there is NO evidence that AC heating prolongs the life of the tubes. I don't even know how you would go about rigorous testing of that on the lifetime of a good 300B like a Takatsuki or WE. Plus you are at the mercy of the power company mains fluctuation.
Thank you for the suggestions @waltube and @daanve
Regarding Rg_3, I updated the schematic and I will soon post the updated version of it. Thank you for the suggestion.
The "don’t connect the psu to ground” is a note to myself to leave the transformer winding floating to avoid ground loops. I'll change the note to make it clearer.
Regarding the first stage, technically you are right: that's LED bias. In practice, by using a large LED with very low differential impedance (dv/di < 0.5 Ohm) the ripple rejection of the ac component Ik_ac of Ik is very high that the effect of Vk_ac is almost neglegible.
Regarding the AC vs DC filament debate on DHT, I heard contrasting opinions about it. I think I will make it configurable through a switch to select between AC or DC and call some friends at home for a blind test.
Sei di Roma? Anche io sono Italiano, originario di Treviso ma vivo in USA
Regarding Rg_3, I updated the schematic and I will soon post the updated version of it. Thank you for the suggestion.
The "don’t connect the psu to ground” is a note to myself to leave the transformer winding floating to avoid ground loops. I'll change the note to make it clearer.
Regarding the first stage, technically you are right: that's LED bias. In practice, by using a large LED with very low differential impedance (dv/di < 0.5 Ohm) the ripple rejection of the ac component Ik_ac of Ik is very high that the effect of Vk_ac is almost neglegible.
Regarding the AC vs DC filament debate on DHT, I heard contrasting opinions about it. I think I will make it configurable through a switch to select between AC or DC and call some friends at home for a blind test.
Sei di Roma? Anche io sono Italiano, originario di Treviso ma vivo in USA
I simulated the coupling between the D3a out and the 300b in. With Rg=50k and C_coupling = 0.47uF, I would get -0.35 dB@1Khz, -0.8dB@20Hz and -3dB@7Hz which is not bad.
