hello Bas,
My apologies, I didn't get a signal that there was a new post.
For V1 boards, R1=R2=9.1Ω 2W, or 4.7Ω 2W [all wirewounds] will run a 3A5.
My apologies, I didn't get a signal that there was a new post.
For V1 boards, R1=R2=9.1Ω 2W, or 4.7Ω 2W [all wirewounds] will run a 3A5.
I am working currently on the raw supply (without regs at the moment) and to understand what does what, I tried the weekend 80mH choke input (7.5V, 1,25A) vs. choke inpit with two split coils in differential mode.
Hmmm...interesting. It changes the sound. Brighter. Tighter. Stiffer. Maybe more resolution, but as well less warm tone...pick your poison I would say. Does anyone have a bit more experience with the differential setup va. normal PSU ? Any explanations why this sound that different ?
Hmmm...interesting. It changes the sound. Brighter. Tighter. Stiffer. Maybe more resolution, but as well less warm tone...pick your poison I would say. Does anyone have a bit more experience with the differential setup va. normal PSU ? Any explanations why this sound that different ?
I usually use this type of raw supply PSU for 10/801 tubes.
For me C-CMC-C give better sound, than C-R-C (lesser noise on R.C input).
I tried "classic" L-C type, but requires larger secondary voltage and the choke is too big ... and -for me- it doesn’t add that much to the sound, only a little different.
For me C-CMC-C give better sound, than C-R-C (lesser noise on R.C input).
I tried "classic" L-C type, but requires larger secondary voltage and the choke is too big ... and -for me- it doesn’t add that much to the sound, only a little different.
Attachments
I have a pair on my 300B set amp and its dead silent into a pair of 105db spl speakers and sounds amazing. Whatever it is it works.
Jeff
Sorry if this is an old chestnut but what is the opinion on using smps as raw supply for Rod's boards?
I use it only for decreasing output voltage of the row supply.
I have two box configuration: raw supplies in the PSU box and R.C. regulators and Ale's "gyrators" in the preamp box.
The latter (and the filament bias resistors) dissipating quite a lot, so sharing dissipation is necessary.
I assume you mean Mains/Line to DC power units.
If you get one that floats both + and - outputs fully, then it will function.
Many units with a 3-pin Mains plug have the PE connected to the output 0V (makes passing statutory EMC regulations easier).
But with DHTs running cathode bias, the cathode resistor is short circuited to PE if such units are used.
Units with a metal case will most probably connect their case to PE, and the outputs to the case with Y-class capacitors. This again makes an unexpected path across the cathode resistor and bypass cap., and I suspect it will degrade the sound.
2-pin mains feed units might still have an EMC cap between the mains-side of the transformer and the output, and this causes a leakage current to flow from the mains-switching node to the output, and through the filament - where it mixes with the anode current.
The remaining risk is that the leakage current runs through the system's audio signal cables, and gets into the DAC ground, on the way to PE.
Naturally, if you know about these risks, and can measure and sort the good units from the unsuitable designs, you may be able to get the performance somewhere near to a system that uses a normal EI PT, an R-core PT or a screened toroidal.
But it is impossible to predict whether a given unit is suitable. One has to keep in mind that a DHT filament supply has to float, at the Raw DC, and that is difficult for "SMPS" mains→DC units to achieve.
As a comparison, a typical EI 50VA transformer used for a 300B Raw DC for my regulators has a PRIMARY to SECONDARY capacitance of 50-80pF. This means a leakage current (at 240Vrms 50Hz input) of ca. 10µA at 50Hz in the output. Measurement and calculation of this current agree closely.
Meanwhile Mains-to-DC units have a typical leakage current in the milliamp region - and broadband noise with high levels around the switching harmonics, and at 100MHz to 400MHz from switching edges.
Constructors with good EMC analysis skills can diagnose and possibly work around some of these problems, for a suitable unit. But it seems to me that Split-bobbin EI PTs are mostly low cost, and give dependable performance. They also last forever, unlike packaged PSU units.
So that's the reason that I do not mention them in my documentation....
Thanks Rod, answers everything. Also highlights design details in the Raw Supply design notes. My 845 supplies have done great service and I now intend to go over them and implement more rigorously.
I use a such heavy transformers it makes the amps immobile which makes bench work difficult, so I thought of attaching a smps to the back. I think I will purchase the rated EI transformers.
much appreciated, martin
I use a such heavy transformers it makes the amps immobile which makes bench work difficult, so I thought of attaching a smps to the back. I think I will purchase the rated EI transformers.
much appreciated, martin
What are the current thoughts on the effect of the voltage gradient along the filament wire for high current transmitter tubes? I read some posts on that from a year or ten ago, any newer insights or experiences? I ordered the V9 for a pair of 811a's.
hello August, the voltage skew across the filament produces a skew in current density across the filament, but I see little to be worried about it. The triode curves of the 811A at high voltage are agreeably linear, regardless of the skew.
The anode current under normal conditions is a small fraction of the anode current:
I don't believe that any changes to the amplifier or operating regime of the triode are worthwhile, to accommodate the skew.
The filament runs both up and down the tube so I don't think it will affect plate dissipation much either. I do have a question why does it say max 25W push pull. With a 36W plate max I would have through you could get a lot more peak power out or have I made a mistake?