Hello,
I am designing a DC coupled amp and need all the voltage I can squeeze out of the PSU. I've used two paralleled 6X4 rectifiers (Rp=525Ohms per plate) and though I'd get a little voltage boost by switching to a single 6BY5GA tube (Rp=100Ohm per plate). However, there was no voltage increase. In fact, it might have dropped slightly from 312V down to 309V.
I am curious why the theoretically reduced Rp did not result in increased B+.
Thanks,
Jenya
I am designing a DC coupled amp and need all the voltage I can squeeze out of the PSU. I've used two paralleled 6X4 rectifiers (Rp=525Ohms per plate) and though I'd get a little voltage boost by switching to a single 6BY5GA tube (Rp=100Ohm per plate). However, there was no voltage increase. In fact, it might have dropped slightly from 312V down to 309V.
I am curious why the theoretically reduced Rp did not result in increased B+.
Thanks,
Jenya
That is because the voltage drop of the rectifier is larger in the 6BY5. If you want more voltage use silicon diodes. Only 0.7v drop with those.
I think i understand the confusion. Rp is not an equivalent resistance of the tube, but rather the source impedance. The voltage drop is current dependent and cannot be approximated by a resistance. Often, there is an appropriate graph in the datasheet.
http://www.fourwater.com/files/fullrect.txt
http://www.fourwater.com/files/fullrect.txt
Attachments
Squeeze the "tallest" B+ rail out by rectifying with high PIV Schottky diodes, whose forward drop is smaller than "PN" junction diodes. To be fair, the forward drop difference between Schotky and "PN" is tenths of a V. However, Schottky diodes don't exhibit a reverse recovery spike and are every bit as quiet as vacuum rectifiers.
Quiet they may be. Sounding even remotely similar they are not. If someone had gone to the trouble of using parallel rectifiers, they probably had a good reason for doing it.
Thank you all for replies.
Indeed, I was equating the voltage drop with Rp. I don't understand why that isn't a correct assumption and what Rp values are good for then.
It was me who went through the trouble of using two parallel 6X4. I guess I'll go back to using those.
The main challenge I am facing at the moment is that it is impossible to find a reasonably sized power transformer that has high secondary voltage but isn't rated for high power. I need 440-0-440 secondary but only about 100W. The only torroidal power transformer I was able to procure is a 400W monstrosity. This is for a headphone amp using 12b4 output tubes driven by 6fq7. The total draw is about 80ma. Is it correct to estimate the power requirement on the power transformer by multiplying the secondary voltage by the current draw, in my case 440V*80ma=35.2W?
I am debating about using two smaller power transformers in series.
Jenya
Indeed, I was equating the voltage drop with Rp. I don't understand why that isn't a correct assumption and what Rp values are good for then.
It was me who went through the trouble of using two parallel 6X4. I guess I'll go back to using those.
The main challenge I am facing at the moment is that it is impossible to find a reasonably sized power transformer that has high secondary voltage but isn't rated for high power. I need 440-0-440 secondary but only about 100W. The only torroidal power transformer I was able to procure is a 400W monstrosity. This is for a headphone amp using 12b4 output tubes driven by 6fq7. The total draw is about 80ma. Is it correct to estimate the power requirement on the power transformer by multiplying the secondary voltage by the current draw, in my case 440V*80ma=35.2W?
I am debating about using two smaller power transformers in series.
Jenya
Last edited:
Want current from tubes with less drop? Use dampers...
http://www.nj7p.info/Tubes/PDFs/Frank/093-GE/6DE4.pdf
http://www.nj7p.info/Tubes/PDFs/Frank/093-GE/6DE4.pdf
I was under an impression that 6BY5GA is a damper with the added convenience of having two diodes in one envelope.
get 2 smaller transformers 6.3v @ 10A and run the second one back to back of the first for your B+ supply. run the heaters off the 6.3v
2 25VA torroids from digikey run off the main transformer will work to yield 600v with a voltage doubler
( 440-0-440 will yield 622v with diodes )
Why so high a B+ for a headphone amp. 150 - 200v is plenty
2 25VA torroids from digikey run off the main transformer will work to yield 600v with a voltage doubler
( 440-0-440 will yield 622v with diodes )
Why so high a B+ for a headphone amp. 150 - 200v is plenty
Last edited:
I need a high voltage because this is a directly coupled design, similar to Loftin-White. Basically, tubes are stacked one over another. Assuming, each tube runs at about 150V plate potential, that requires 300V B+. I use choke input filter, so I loose many volts because of that after the rectifier. Plus loses due to DCR components of filtering chokes. Right now, I am getting 312V of B+ from a 400-0-400 power transformer.
Last edited:
312V seems low. Have you confirmed the 400-0-400V levels, and the load current? What is the choke henry and DCR?
You are at max PIV capability - have you added 2 series 1N4007 in series with each 400V winding as backup protection?
Why do you want to use a choke input filter?
You are at max PIV capability - have you added 2 series 1N4007 in series with each 400V winding as backup protection?
Why do you want to use a choke input filter?
I have not verified 400V. I'll do that later.
It doesn't sound too much out of an expected output voltage. Just using the choke input and taking into the account the 11V voltage drop across two parallel 6X4 gets 350V. Loses on two filtering chokes could account for the remaining 38V.
I found that choke input filters result in better sounding amps. Probably due to better voltage regulation.
Jenya
It doesn't sound too much out of an expected output voltage. Just using the choke input and taking into the account the 11V voltage drop across two parallel 6X4 gets 350V. Loses on two filtering chokes could account for the remaining 38V.
I found that choke input filters result in better sounding amps. Probably due to better voltage regulation.
Jenya
What is PIV? Is it peak voltage? The spec says that the tube has a typical operation of 450V with choke input.
I am not familiar with the trick of adding SS diodes in series with windings.
Jenya
P.S. For some reason, I couldn't find the "Edit" button to add this comment to my last post. I know that I used to see the "Edit" button, but now it dissappeared.
The Tung-sol datasheet I was looking at shows a derating starting at 360V, and extending out to 450V, for choke input, so 400V is starting to get near to max rated (depending on your local mains voltage).
Rectifier diodes going gassy is not too uncommon, so a modern day method of bullet-proofing that is to add silicon diodes in series with the valve diode. Peace of mind for some of us - especially as gassy diode faults have been known to knock out power transformers that aren't adequately fused.
Did you look at 'tuning' the first choke to reduce dominant 2nd harmonic ripple? That could be of some help, given that you seem to be chasing very low ripple by using additional filter stages.
Rectifier diodes going gassy is not too uncommon, so a modern day method of bullet-proofing that is to add silicon diodes in series with the valve diode. Peace of mind for some of us - especially as gassy diode faults have been known to knock out power transformers that aren't adequately fused.
Did you look at 'tuning' the first choke to reduce dominant 2nd harmonic ripple? That could be of some help, given that you seem to be chasing very low ripple by using additional filter stages.
PIV = peak inverse voltage. A graphic for the series SS diode technique is provided. I suggest UF4007s, instead of 1N4007s, on the hypothesis that less switching noise to begin with is better. The vacuum rectifier does block SS diode switching noise, no matter what.
Attachments
- Status
- Not open for further replies.