New to power supply design and have what might be a dumb question...
According to the 5AR4 datasheet, it appears that the maximum AC voltage per plate could be as high 550V or even 600V depending on the DC output current per plate. I've read in several places on several different forums that the max AC voltage of a 5AR4 is 450V. Am I reading the datasheet wrong?
Thanks!
According to the 5AR4 datasheet, it appears that the maximum AC voltage per plate could be as high 550V or even 600V depending on the DC output current per plate. I've read in several places on several different forums that the max AC voltage of a 5AR4 is 450V. Am I reading the datasheet wrong?
Thanks!
"Am I reading the datasheet wrong?"
No you're not. For a "real" 5AR4 550 to 600 volts is OK. And by real I mean a NOS American brand or a quality European make like Amperex, Phillips or even Telefunken. And for these the peak inverse voltage is well over 1KV. However, since there are now many low quality Chicom and Russky tubes available, that level of voltage is risky and sometimes asking for premature failure by fireworks. Levels of 450 volts is more of a typical operating level used now as a max for those garbage unites. (Yes I'm prejudice) That's why putting diodes in series with the plates is a good practice.
No you're not. For a "real" 5AR4 550 to 600 volts is OK. And by real I mean a NOS American brand or a quality European make like Amperex, Phillips or even Telefunken. And for these the peak inverse voltage is well over 1KV. However, since there are now many low quality Chicom and Russky tubes available, that level of voltage is risky and sometimes asking for premature failure by fireworks. Levels of 450 volts is more of a typical operating level used now as a max for those garbage unites. (Yes I'm prejudice) That's why putting diodes in series with the plates is a good practice.
Ah ok. That makes some sense as all of the datasheets I looked at were of an era... I have a NOS Matsushita 5AR4 that I'm planning on using with a PT with an output of 420V AC. From what I've heard, these were made either using Mullard equipment or under Mullard supervision. Should be alright, ye?
Great! And in the case of a choke input, directly to the choke?
Wish it were mine! Found it online a while back.
Yes, either the capacitor or the choke.
Point is, the only changes are adding a diode in series with each HV secondary transformer lead.
Unused rectifier socket lugs are tie points. Cathodes of the diodes go toward the capacitor or inductor.
Point is, the only changes are adding a diode in series with each HV secondary transformer lead.
Unused rectifier socket lugs are tie points. Cathodes of the diodes go toward the capacitor or inductor.
Last edited:
The "yellow sheet" mod. Too bad someone didn't copyright it, they'd have some extra money. ;-)
5AR4 Maximum Plate Voltage depends on:
DC Load current.
Capacitor Input filter capacitance value (not just the maximum spec).
In some cases, series R to each plate lead (includes DCR of 1/2 secondary winding, plus series resistors if needed).
Choke inductance, Henrys.
Proper use of 5AR4 is not trivial.
Obey all the rules or there will be trouble.
Ask me how I know (I had one arc, due to ignoring one specification).
DC Load current.
Capacitor Input filter capacitance value (not just the maximum spec).
In some cases, series R to each plate lead (includes DCR of 1/2 secondary winding, plus series resistors if needed).
Choke inductance, Henrys.
Proper use of 5AR4 is not trivial.
Obey all the rules or there will be trouble.
Ask me how I know (I had one arc, due to ignoring one specification).
And just to be clear... the function of the series diode approach as described above is to avoid a host of upstream and downstream problems if the rectifier fails, correct?
I feel that this approach plus a thermistor to limit inrush current and two fuses - fast and slow blow - on the mains should provide a good degree of protection for the power supply. Can you see any downside - besides additional complexity - to this belt and suspenders approach?
1N4007 diodes have peak inverse voltage rating of 1000V, Right?
PIV
If you have a 400-0-400VAC rated secondary (when loaded), then the loaded voltage to the rectifier is 400 x 1.414 = 565V
When the secondary is unloaded, and the tube rectifier is unloaded (low voltage drop), there will be at least 565V B+ at the first cap.
(Before the output tubes warm up, the first B+ capacitor will have about 565V).
The peak inverse voltage will be 565V x 2 = 1130V.
If the tube rectifier fails, the 1N4007 diodes will also fail.
Use 1500V Peak Inverse voltage rated diodes, not 1000V rated diodes
Cheap is not Good, Good is not Cheap . . . Frank Reps in Sound Practices.
PIV
If you have a 400-0-400VAC rated secondary (when loaded), then the loaded voltage to the rectifier is 400 x 1.414 = 565V
When the secondary is unloaded, and the tube rectifier is unloaded (low voltage drop), there will be at least 565V B+ at the first cap.
(Before the output tubes warm up, the first B+ capacitor will have about 565V).
The peak inverse voltage will be 565V x 2 = 1130V.
If the tube rectifier fails, the 1N4007 diodes will also fail.
Use 1500V Peak Inverse voltage rated diodes, not 1000V rated diodes
Cheap is not Good, Good is not Cheap . . . Frank Reps in Sound Practices.
I'll be using a 420-0-420VAC PT with a choke input so it should be a bit less voltage-wise, but still easy enough to go with a higher-rated diode.
Thank you and amazing... I was just reading this!