| Sheldon |
I have a CT transformer and I'd like to make a sort of hybrid full wave rectifier with a ss diode and tube (indirectly heated) section in series for each leg. The main reason to do this is for the slow turn on, as I'd like that for the DHT output tubes. My PSUD sims show that I am right at the limit for PIV for the tube rectifier. If I were using only SS diodes, I could simply put them in series to half the PIV for each diode. But what happens with a tube diode and SS diode in series? Can I assume the the voltage drop will be significantly less for the tube than without the SS diode in series? Any thoughts here?
I could just make a full wave SS rectifier and then put the tube in series after rectification, but that would lose the possible benefit of soft shut off. Or I could take apart the transformer and rewire for a hybrid bridge (assuming the two ends of the center tap are within reach), in which case the PIV would be half of that for the full wave. But I'd rather avoid that if I can.
Sheldon |
|
|
| Eli Duttman |
Sheldon,
MAYBE SS diodes in series with the vacuum rectifier would work. What's the end to end voltage of the rectifier winding and which tube are you using? Changing the tube rectifier type seems a lot less problematic to me than series SS diodes or trafo modification. |
|
|
| Sheldon |
| quote: | Originally posted by Eli Duttman
Sheldon,
MAYBE SS diodes in series with the vacuum rectifier would work. What's the end to end voltage of the rectifier winding and which tube are you using? Changing the tube rectifier type seems a lot less problematic to me than series SS diodes or trafo modification. |
Thanks,
Transformer is 480-0-480. Should give me 600VDC. I'd have to fish out the leads and parallel the secondary sections to go with a bridge. I'm trying to use a 6X4 (yeah, I know it's marginal). I don't see many alternatives in a small tube (7 or 9pin) with two sections that would have ratings much higher for PIV. I could use a damper diode after rectification (6CK3, etc.), just for the warm-up. I don't want to go larger if I can help it, cause I don't have much room. Heck, I may try it and see what happens. If the 6X4 lives, it lives. If not, on to plan B.
Sheldon |
|
|
| Eli Duttman |
480-0-480 is WAY over spec for a 6X4.
6X4 Data Sheet
"Monkeying" with the trafo and a hybrid bridge seems best to me. |
|
|
| Sheldon |
Oops, my unedited reply was way off. I misread the data sheet. I don't see where things would be way off here. I would surely be on the limit for PIV, but with slow warm up, maybe not. One of the examples shows a 900V plate to plate supply. The only thing I can see is the plate to cathode rating of 450V. But couldn't that be dealt with by letting the heater winding float?
Sheldon
second edit: sorry for not thinking THEN writing.
third: oops, meant heater winding not cathode |
|
|
| Eli Duttman |
Sheldon,
Your trafo puts approx. 1358 V. on the 6X4's anode. The data sheet says 1250 V. max. That's a design center number; so, a little more with a choke I/P filter rates to be fine. However, your PSU is cap. I/P. I'm WORRIED about arc over. |
|
|
| Sheldon |
Yes, definetly on the edge. I'll see what the transformer holds.
Happy Holidays all,
Sheldon |
|
|
| Sheldon |
Actually, thinking about this a little more (danger, danger). It seems that with identical devices, the potential should be more or less halved if they are in series. In this case, with a SS diode and a tube diode, it would seem that the tube would see most of the inverse voltage drop. The tube is a slow cut off device and cannot conduct in the reverse direction. As I understand it, the SS diode has the property of a small initial reverse current, which is what can lead to the transient spikes that people are concerned with in these application. Therefore, it seems the the tube will control the cut off (hence the desireability of using it in a hybrid), and therefore see more of the reverse voltage potential across the two devices in series. Does that make sense? More importantly, it correct?
Sheldon |
|
|
| Eli Duttman |
Sheldon,
I've got an idea. It just may solve the problem. Put a UF4007 in series with each of the 6X4's anodes. Wire resistive voltage dividers across the combinations to the cathode. Instead of 50/50 equalizing, as is "commonly" seen, use 40/60. Let the 1000 PIV SS parts handle 40% of the voltage and the tube the remaining 60%. |
|
|
| Sheldon |
Thanks Eli,
That seems like it should get around the PIV limitations. But would the SS diodes now shut off normally, with the attendent reverse recovery spike? BTW, I'm a complete agnostic on the issue of whether or not that matters. I don't have the experience to have an informed opinion one way or the other. It is a practical way to get slow turn on with just one small tube, which is my main objective here. I'll still check the transformer to see if it would be easy to parallel the secondaries and go with a bridge, but if not, I'll give this a shot.
Sheldon |
|
|
| Eli Duttman |
Sheldon,
UF4007s have a small reverse recovery spike, but the tube is faster than the "sand". So, the spike is blocked by the 6X4. The same holds true in a hybrid bridge. |
|
|
| dhaen |
| I've seen schematics where a SS bridge is followed by a single tube diode. Presumably this amounts to the same thing for both noise and delay. |
|
|
| Sheldon |
Like this: http://tinpan.fortunecity.com/saint...imer/211-se.gif
And Eli, you been holding out on me?
"FWIW, the "book" on Schottky diodes is that they should not be stacked. However, a Schottky diode and a FAST PN junction diode can be stacked. The combination works well. The Schottky diode blocks PN switching noise and the junction diode protects the Schottky from overvoltage. Thanks go to AA "inmate" Volt Second for the stacking synergy concept."
http://www.diyaudio.com/forums/show...0599#post610599
How about a Schottky and Fast PN in series on each leg of the full wave, bring the legs together and just put the 6x4 paralleled, in series with the rectifier output? BTW, I checked the transformer and it looks like the tap source is well buried, so changing the secondary configuration may not be so attractive.
Sheldon |
|
|
| Eli Duttman |
Sheldon,
I'm CHEAP. UF4007s cost 15 cents each. Resistors don't cost all that much. Check the price of 1200 PIV Schottky diodes. Also, you have limited space. The 1200 PIV Schottky parts come in TO220 cases. The 6X4 will block SS switching noise. Why make things more complicated than they have to be? |
|
|
| Sheldon |
| quote: | Originally posted by Eli Duttman
Sheldon,
I'm CHEAP. UF4007s cost 15 cents each. Resistors don't cost all that much. Check the price of 1200 PIV Schottky diodes. Also, you have limited space. The 1200 PIV Schottky parts come in TO220 cases. The 6X4 will block SS switching noise. Why make things more complicated than they have to be? |
Fair enough. Cheap or not, it makes no sense to spend more when you don't get more for the application.
I'm now going to reveal the depths of my ignorance and make you sorry you replied at all. I'm not visualizing how the tube will block the switching noise here. If it's just a tube and ss diode in series, the tube will not reverse conduct so neither can the SS diode. But if the tube is controlling the shut off, it seems like it will see perhaps the full reverse PIV. In this configuration, the would still be useful but really only as a fail safe device. If I use resistor dividers with the SS and tube, the SS diode can share the potential but then will see some reverse current (maybe less than normal?) and so will have some spike?
I'm not saying that this is not worth doing. In fact, I'll probably try it. I'm just trying to better understand all the issues.
BTW, would it be worth measuring the reverse voltage at the between input and middle and output for a tube and SS diode in series, with a reverse DC voltage? Would that give some idea about the potential at various points, or (and I'm thinking it's likely here) is the dynamic case so different from the static case that the test would be pointless?
Thanks,
Sheldon |
|
|
| Eli Duttman |
Sheldon,
To be stone cold certain the PSU is QUIET, add a RRSF between the rectifier and the power trafo. |
|
|
| Sheldon |
| quote: | Originally posted by Eli Duttman
Sheldon,
To be stone cold certain the PSU is QUIET, add a RRSF between the rectifier and the power trafo. |
Quiet is good. But you missed the part about ignorance. That wasn't false modesty. RRSF?
Sheldon |
|
|
| Eli Duttman |
Sheldon,
RRSF = reverse recovery spike filter. It was devised by the late John "Buddha" Camille. Jump over to Audio Asylum and search the Bottlehead Forum archives for RRSF. There's plenty in the "vault". |
|
|
| Sheldon |
Thanks again,
Sheldon |
|
|
| rcavictim |
| quote: | Originally posted by dhaen
I've seen schematics where a SS bridge is followed by a single tube diode. Presumably this amounts to the same thing for both noise and delay. |
This is a good trick if you use a slow warmup rectifier like a TV damper diode, 6AX4, 6AU4, etc as the tube with a SS bridge to provide a B+ delay, but I do not see how the tube iin this configuration is going to do anything at all to stop bridge diode switching pulses getting to the rest of the PSU. It seemed to me that this aspect was not clearly addressed in this thread yet.
If I'm not seeing this correctly could someone illuminate me? |
|
|
| mllum |
Sheldon,
Would you please update all of us on which solution you eventually settled on? I am particularly interested in whether solid state diodes like the uf4007 can be wired in series with the tube rectifier to extend the PIV value.
If you implemented the connection of the uf4007 in series with the plate of the 6x4 with resistors to balance the voltage, would you please illustrate the same with a schematic.
Please excuse my ignorance as I am truly a newbie at this.
Thanks. |
|
|
| Sheldon |
Here's what I ended up with:
http://www.diyaudio.com/forums/show...3559#post883559
No balancing resistors. I've had no trouble with this one. Interestingly, the second of two identical amps gave some slight cracking/popping noises the first few times on start up. Some arcing in the rectifier tube? Have not heard it since, and the amps have performed flawlessly.
Sheldon |
|
|
|
