I used 47R in each plate for current sharing, but I parallel them (4 tube diodes) because my amp idles at about 0.5A.
It's a Hammond 278CX... PSUD2 came up with 20R as Source Res... 424V
Well, there you are. You're using the vacuum rectifier as a slow-turn-on device, essentially. Which is a perfectly good use for it. You might want to double up those silicon rectifiers (2 in series, on each side), with 4.7 MΩ resistors in parallel with each too. Reverse voltage sharing the safe way. Causes them to last well past the life of the power supply.
Just saying,
GoatGuy ✓
And thinking even more specifically, there would be high value in putting a mid sized (say 47 µF at 450 V) “first capacitor” between the stacks of solid state rectifiers and the anode of the vacuum device.
With a 470 kΩ, 1 watt bleeder-resistor across it for safety.
It'd substantially lower the peak-i-ness of current flow thru the vacuum rectifier.
Which of course would extend its life significantly.
I might be inclined to add a 100 Ω resistor in series with each rectification leg as well, to limit inrush current some.
But that's just me.
Make sure the 100 Ω resistors are 5 watt devices.
GoatGuy
With a 470 kΩ, 1 watt bleeder-resistor across it for safety.
It'd substantially lower the peak-i-ness of current flow thru the vacuum rectifier.
Which of course would extend its life significantly.
I might be inclined to add a 100 Ω resistor in series with each rectification leg as well, to limit inrush current some.
But that's just me.
Make sure the 100 Ω resistors are 5 watt devices.
GoatGuy
I'm using series caps, 450v rated. Each cap has a 220k 2w resistor for balance, i believe these act like bleeder too ?
My initial plan was, to use some kind of wz68 rectifier, but build under, includes a 36ohm resistor. This will limit the peak current in the damper to ~1.07A, if i'm not mistaken, almost 1/3 derating of the peak current.
As i don't have much experience, would the 36 ohm, or let's say a 50 ohm, have any impact on the speed or dynamics? Or this is too small of a value to even bother thinking?
ureche, you state that you want to isolate ss diode hash, but I can't see how your target circuit does that compared to any other ss diode rectified power supply. If you don't want ss diode related hash exuding from all available exit paths (of which the B+ line is just one), then can I suggest not using ss diodes (shock horror).
You indicate you want to use the damper diode as a delay. Why is that? Are you trying to protect parts in your amp, and have special signage and a key lock on the power switch to avoid anyone turning off the mains switch and then back on quickly afterwards?
Did you measure your PT winding resistances to get the PSUD2 parameters for the transformer?
If you insert the damper diode before the first filter cap, then it has to handle twice the rectified current pulses compared to the more typical half-wave location. With respect to the diode datasheet not recommending use in power rectification applications, you have gone further away from that recommendation by doubling the stress on the diode. As suggested, if you place the damper diode in a filtered part of the circuit, it effectively sees dc current for which you have its 350mA max limit rating.
If you wanted to use a ss diode, then hopefully PSUD2 would have warned you about one 1kV PIV diode not being enough.
You indicate you want to use the damper diode as a delay. Why is that? Are you trying to protect parts in your amp, and have special signage and a key lock on the power switch to avoid anyone turning off the mains switch and then back on quickly afterwards?
Did you measure your PT winding resistances to get the PSUD2 parameters for the transformer?
If you insert the damper diode before the first filter cap, then it has to handle twice the rectified current pulses compared to the more typical half-wave location. With respect to the diode datasheet not recommending use in power rectification applications, you have gone further away from that recommendation by doubling the stress on the diode. As suggested, if you place the damper diode in a filtered part of the circuit, it effectively sees dc current for which you have its 350mA max limit rating.
If you wanted to use a ss diode, then hopefully PSUD2 would have warned you about one 1kV PIV diode not being enough.
The main use for the damper was to eliminate part of the ss hash. I read somewhere that, by adding a damper diode between ss rectifiers and first cap, the b+ will be somewhat isolated from ss diode hash. The delay is just a nice thing to have, i know delay is not needed, or at least this is what i have read, the cathode stripping happens on tubes with voltage over 1000v.
I have the pt specs from manufacturer, 25ohm secondary and 3 ohm primary. Psud calculates a ~35 ohm impedance.
This, i didn't thought off.
The picture posted yesterday does not show, it was a quick snip from google. But i'm using 2x5408 to increase PIV to 2kV.
Yes, of course.
BUT… don't forget, it is also serving as the SLOW-TURN-ON device, that valve.
And keeping the B+ probably 50 V lower than SS rectification would do
… by itself.
Just saying,
GoatGuy ✓
No, I mean right now both output caps are paralleled after the tube. Don't you mean to put the first filter after the UF5408 but before the tube?
Last edited:
Attachments
There is something i'm missing for sure, wouldn't the damper diode be affected by high cap 220uf ?
I wonder if the 4.7mohm resistors should have a voltage rating bigger than classical resitor of ~300v ?
It should be fine. In my experience you have to whip a damper diode pretty hard to have a problem.
That being said, the 47uF cap before the damper will be ready to instantly discharge into the 220uF cap once the diode starts conducting and has no current limiting whatsoever on it. You might hit it a bit hard there even if it is a brief moment in which this is the case because in steady-state operation the 100R resistors take care of current limiting. Normally I would say the diode can take it but it'll have a relatively cold cathode and weak electron cloud at that moment in time.
That being said, the 47uF cap before the damper will be ready to instantly discharge into the 220uF cap once the diode starts conducting and has no current limiting whatsoever on it. You might hit it a bit hard there even if it is a brief moment in which this is the case because in steady-state operation the 100R resistors take care of current limiting. Normally I would say the diode can take it but it'll have a relatively cold cathode and weak electron cloud at that moment in time.
Last edited:
Yes. Your analysis is spot on, less the word instantly.
Those darn vacuum bulbs with indirectly heated cathodes tend to come alive over a 0.7 to 2.0 second interval. Partially conducting until full conductance. Offset, of course, for the thermal inertia of the whole heater-wire-insulation-cathode-cylinder-oxide couple.
Usually the turn on is about 10 seconds. Can be more, can be less. But the dI/dit during the turn-on is quite long-period compared to the 120 Hz (or 100
Hz, continent) FWB pulse stream.
Just saying,
GoatGuy ✓
A 'hot' turn-on is the concern - even with additional 100R in series with effective 36 ohm winding resistance, the initial peak damper diode pulse is above 3A, and likely to subside only slowly if the rest of amp's valves are hot and loading the power supply, and the damper diode is doing 'double pulse' duty in that location.
[1] the OP (I think) said that only one HV vacuum rectifier tube was available
[2] If the stress is transposed to the silicon diodes, all the better
[3] Maybe I've not understood your question.
GoatGuy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.