In commercial power supply design, there's a thing called inrush current. This has to do with tripping the mains breaker when a unit is plugged in or turned on. If that action happens to be coincident with the peak of the AC...
From an engineering perspective, whatever design allows less stress to a devices constituent components is better. However in a commercially available product, that which may be better isnt always necessary if you can get away with it. As has been pointed out, numerous examples exist where they got away with it many times - decades - past the warranty period.
If I was making something myself, for myself, I'd delay the B+ to correlate with the time it takes the tubes to come up to full cathode emission. I'd do it in such a way that the rectifier sees little, if any, current stress upon power on. Further, I'd make it infinitely resettable; you can cycle AC power indefinitely and it never quits going through the entire the power up routine. But that's just me.
From an engineering perspective, whatever design allows less stress to a devices constituent components is better. However in a commercially available product, that which may be better isnt always necessary if you can get away with it. As has been pointed out, numerous examples exist where they got away with it many times - decades - past the warranty period.
If I was making something myself, for myself, I'd delay the B+ to correlate with the time it takes the tubes to come up to full cathode emission. I'd do it in such a way that the rectifier sees little, if any, current stress upon power on. Further, I'd make it infinitely resettable; you can cycle AC power indefinitely and it never quits going through the entire the power up routine. But that's just me.
In some cases the inrush current is maximised by switching on at the AC zero crossing. It depends on whether the impedance presented to the mains is inductive or not.
Delaying HT can sometimes merely shift the current surges from 'switch on' to 'delay on'. You have to analyse the actual risks, and balance running valves with inadequately warm heaters vs. valves with hot heaters but no cathode current; both situations can do harm, but you usually have to choose between them.
Delaying HT can sometimes merely shift the current surges from 'switch on' to 'delay on'. You have to analyse the actual risks, and balance running valves with inadequately warm heaters vs. valves with hot heaters but no cathode current; both situations can do harm, but you usually have to choose between them.
Phew, we learned KUB anti-aircraft moblie system designed back in 1950'Th, and it had delay time network for magnetron tubes. They really needed that, unlike audio power amps where abrupt application of B+ to already conducting tubes causes charge of coupling caps through control grid currents.
I was just reading at a couple of sites that suggested adding bypass resistors to the standby switches to allow some current to flow through to start charging the capacitors: Amp Mods and The Valve Wizard It got me wondering about combining something like the delayed start in the diyaudiostore with a resistor bypass.
I know I could use a thermistor but I like the idea of avoiding another heat generating component and after the delayed start switched on the resistor wouldn't generate any heat.
Last edited:
Either a thermistor or a resistor will generate heat.
A thermistor may better manage the initial rise. Either way, they are off when bypassed by the switch (or relay contact).
Last edited:
Applying some resistance in PS works miracles for ramping up B+unlike audio power amps where abrupt application of B+ to already conducting tubes causes charge of coupling caps through control grid currents.
Did some more searching about the cathode stripping effect and found the attached. Sounds plausible.
An argument you often hear is 'I own a 1962 radio with original tubes and it still plays fine'. But these old radios had tube rectifiers and therefor build-in high voltage delay. So this is really an argument showing the benefits of high voltage delay ...
The reason for my high-voltage delay unit article published in AX. I should add that since it only delays the B+, any negative bias voltage required would be present when the B+ comes on, so no issues there either.
It also has the benefit of switching on the secondary voltage at a zero crossing, limiting the switch-on surge.
Jan
An argument you often hear is 'I own a 1962 radio with original tubes and it still plays fine'. But these old radios had tube rectifiers and therefor build-in high voltage delay. So this is really an argument showing the benefits of high voltage delay ...
The reason for my high-voltage delay unit article published in AX. I should add that since it only delays the B+, any negative bias voltage required would be present when the B+ comes on, so no issues there either.
It also has the benefit of switching on the secondary voltage at a zero crossing, limiting the switch-on surge.
Jan
Last edited:
Applying some resistance in PS works miracles for ramping up B+
Even better, I once made a voltage regulator that waits for tubes to start drawing currents, then brings it up gradually.
www.wavebourn.com • View topic - My new toy: BC-348
Yes, for magnetrons and directly heated transmitting tubes with high anode-cathode current density.
...do you call selenium rectifiers also "tube rectifiers"?
It's silly to pick apart subjects about this "cathode stripping" stuff, B+ delays, etc...
As for my lovely 1963 console, and it being powered by a directly-heated 5AS4A rectifier tube..... Once turned on, the power supply is up to voltage in 3 seconds, and the rest of the tubes (17 of them) are operating fully in 14 seconds.
This has been going on with this console since 1963 - Fifty-Six years!
As I mentioned, the original RCA tubes, date stamped 1962, are still fine, tested, and I'm certainly not concered with this "cathode-stripping" crap.
Nor should anyone else be, unless you allow the internet gods to make you nuts over it.
I've worked on restoring hundreds + of "vintage" consumer equipment - radios, stereos, etc.
1929 era, up into the 1960's...
And in most cases, the original tubes were still in their sockets, and testing fine.
As for the need for "delay" discussed here, I suppose much of the "new breed" of tubes are so crappily made today that they need to be handled with feathers and gloves.
Because quality-made tubes are robust creatures.
As for my lovely 1963 console, and it being powered by a directly-heated 5AS4A rectifier tube..... Once turned on, the power supply is up to voltage in 3 seconds, and the rest of the tubes (17 of them) are operating fully in 14 seconds.
This has been going on with this console since 1963 - Fifty-Six years!
As I mentioned, the original RCA tubes, date stamped 1962, are still fine, tested, and I'm certainly not concered with this "cathode-stripping" crap.
Nor should anyone else be, unless you allow the internet gods to make you nuts over it.
I've worked on restoring hundreds + of "vintage" consumer equipment - radios, stereos, etc.
1929 era, up into the 1960's...
And in most cases, the original tubes were still in their sockets, and testing fine.
As for the need for "delay" discussed here, I suppose much of the "new breed" of tubes are so crappily made today that they need to be handled with feathers and gloves.
Because quality-made tubes are robust creatures.
OK Iam late in this bus, but to keep things well done and safe in the amp one will run in his own home a builder that worth his shoes must build a amp with the best standard of the old age: two power on switches, this is especially useful to take a longer tube life from expensive rare tubes.
Switch 1 will power on only the heaters, it must have its own Soft Start, 30 seconds or more. After a few minutes of heating the filaments usually 10 minutes (its minutes not seconds) or more when the tubes no more are cold, the owner will manually activate the SWITCH 2.
Switch 2 will power on the HV grids and Plate etc it must have its own Soft Start 30 seconds or more. HV supplies never can be released with the tubes cold, otherwise tube life will be shortened and even whisker noise can develop in carbon plates. After more 10 minutes of heating the grids and plates the amp will be on operate condition.
Tube manufacturers dont inform Soft Start time and filament heating time as they want to sell you more tubes in the shortest time possible, a rare case it is know is the 6C33, its filament heating minimum time are 120 seconds (minimum).
My two cents.
Last edited:
I haven't found any data on the desirability of delaying the neg bias, if used. But the neg bias must be present when the B+ comes on, for obvious reasons.
So I would say, switch on everything except B+ at the same time, then switch on B+ after a delay.
Jan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.