I went ahead and modeled the power supply in PSUD2.
I was not getting the labeled voltages with the tube rectifier. I was getting 284 Volts at the 82uF cap (B+)
The schematic indicated 328V.
When I ran the sim with 1N4007 diodes, I got about 320V B+.
Try PSUD2. We can help. Your output tubes can be modeled as a 67mA current Tap. And your front end can be modeled as a 4.7mA Current tap.
FYI, you can "tune" the supply voltage by adding or subtracting capacitance from the "1 uF" position.
PSUD2 predicts that .01uF in that position will give a b+ of 293V while a 10uF in the same spot gives about 430V B+.
After you get it working, measure the B+ and tune the voltage to within 10% or less by adding or subtracting capacitance from the first capacitor.
Your choke should be fine. What is the maximum voltage rating?
HTH
Doug
I was not getting the labeled voltages with the tube rectifier. I was getting 284 Volts at the 82uF cap (B+)
The schematic indicated 328V.
When I ran the sim with 1N4007 diodes, I got about 320V B+.
Try PSUD2. We can help. Your output tubes can be modeled as a 67mA current Tap. And your front end can be modeled as a 4.7mA Current tap.
FYI, you can "tune" the supply voltage by adding or subtracting capacitance from the "1 uF" position.
PSUD2 predicts that .01uF in that position will give a b+ of 293V while a 10uF in the same spot gives about 430V B+.
After you get it working, measure the B+ and tune the voltage to within 10% or less by adding or subtracting capacitance from the first capacitor.
Your choke should be fine. What is the maximum voltage rating?
HTH
Doug
I never 'heard' mercury rectifiers myself, but I guess the 'hard switching' is due to the fact that the mercury vapor switches back and forth between the gaseous and the plasma state with every cycle. I can imagine that it ignites and extinguishes quite abruptly, thus generating a lot of hash. We don't get this in true vacuum diodes, for obvious reasons.
I modeled the psu in psu2 and got the same result as Dougl, only way i could get the voltage up to 328v was to reduce the load current.
If i used a 120ma load it needed 2uf to bring the voltage up to 330v, if your b+ is 328 volts now it will run hot for sure if the 5y3 is just changed out. changing the 1uf cap to 0.5uf would lower it enough to keep things in check, you could then fine tune it by changing the 0.5uf cap up or down until the voltage is where you want it.
If i used a 120ma load it needed 2uf to bring the voltage up to 330v, if your b+ is 328 volts now it will run hot for sure if the 5y3 is just changed out. changing the 1uf cap to 0.5uf would lower it enough to keep things in check, you could then fine tune it by changing the 0.5uf cap up or down until the voltage is where you want it.
I found a website once that had current vs. time or voltage vs. time for various solid state diodes during switching. I guess it would be neat to see the same plot for a mercury rectifier.
Interestingly, one of the plots compared a fast recovery diodes with a standard and I would probably choose the standard. I don't know what the fast recovery diode was, but it had a nasty sharp reverse recovery (but it was fast).
I am currently using Fairchild Stealth diodes as they supposedly have a fast and soft recovery. I also measure leakage inductance on the secondary of my power transformers and calculate out RC snubbers to eliminate any possibility of excitation of parasitic resonant circuits there. As a result, I have noticed a decrease in output noise from my Maida regulators. In fact, I can no longer measure anything AC on the output of the regulators with the equipment that I have. Before I had little spikes at a rate of 120Hz.
I found the same with SS rectifiers.
Had to add a small cap across the diode cathode to ground to kill switching spikes.
The use of the power supply filter input capacitor to regulate the output voltage is a neat and simple way of doing it, with one disadvantage. It could place matters somewhere between normal and choke filter output on an output voltage curve which is quite sensitive to load current variations, resulting in poor regulation.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.