A diode is a diode , don't compare it with other tubes . The internal resistance is low , the energy stored in a filter capacitor is way higher than in bypass cathode cap ...
All tubes are diodes, with grids. If a rectifier does not benefit from a delayed HT when working in the harshest conditions, then it is dubious to claim an amplifier tube needs it for less harsh conditions.
Yes, yes and yes! I've always been more than sceptical about the idea of dedicating a rectifier tube as a slow start device. As it's cathode basically isn't anything different from those of the receiving tubes, cathode wear due to the inevitable filter capacitor charge current surge and uneven warm up with small hot spots conducting earlier than other surface areas simply gets swapped to the rectifier.
I also follow Merlin's scepticism on the origin of the phenomenon reported by the OT. With a cold cathode, a vacuum tube is a quite good isolator. What should happen within it, unless the electrode potential, nay, field strength, is that high that arcing might be ignited (which in turn is quite unlikely in small signal tubes with high value plate resistors)?
Best regards!
Slamming full B+ to a heated tube might be bad for the tube. Think of coupling caps
which are discharged when B+ arrives, one side will adopt to the drivers voltage,
typically a jump of several hundred volts, the other side should be kept at zero ( or negative bias). When B+ arrived the bias side will have a large positive puls to the grid.
This is in my opinion a very strong reason NOT to have delayed B+ or using a standby
switch at start of an amp.
which are discharged when B+ arrives, one side will adopt to the drivers voltage,
typically a jump of several hundred volts, the other side should be kept at zero ( or negative bias). When B+ arrived the bias side will have a large positive puls to the grid.
This is in my opinion a very strong reason NOT to have delayed B+ or using a standby
switch at start of an amp.
Slamming full B+ to a heated tube might be bad for the tube. Think of coupling caps
which are discharged when B+ arrives, one side will adopt to the drivers voltage,
typically a jump of several hundred volts, the other side should be kept at zero ( or negative bias). When B+ arrived the bias side will have a large positive puls to the grid.
This is in my opinion a very strong reason NOT to have delayed B+ or using a standby
switch at start of an amp.
That is a very good point. For convenience I've usually included a B+ standby switch on the experimental amps I've built. So on the clone of the Electrovoice Circlotron with fixed bias I noticed huge currents in the KT88s as the standby switch was activated. I measured some, the scope traces should be somewhere here in my notes from about 20 yrs ago. I'll take a look, might get lucky & find the evidence.🙂
But not a problem with cathode bias, the bias comes on immediately altho the current may be a little higher for a few mS. Depends on the bias resistor & cap time constant.
which are discharged when B+ arrives, one side will adopt to the drivers voltage,
typically a jump of several hundred volts, the other side should be kept at zero ( or negative bias). When B+ arrived the bias side will have a large positive puls to the grid.
This is in my opinion a very strong reason NOT to have delayed B+ or using a standby
switch at start of an amp.
That is a very good point. For convenience I've usually included a B+ standby switch on the experimental amps I've built. So on the clone of the Electrovoice Circlotron with fixed bias I noticed huge currents in the KT88s as the standby switch was activated. I measured some, the scope traces should be somewhere here in my notes from about 20 yrs ago. I'll take a look, might get lucky & find the evidence.🙂
But not a problem with cathode bias, the bias comes on immediately altho the current may be a little higher for a few mS. Depends on the bias resistor & cap time constant.
All this recent chatter over surges and such with tubes... why?
Some might think that once a rectifier tube is powered, there's some sort of dangerous "slam" effect with current when charging filter caps - there isn't any danger, actually.
Not unless you load the damn rectifier with rediculously large filtering past its design level.
(the more uF is better capacitor syndrome)
What you have to understand is...... tubes are very much a "forgiving" creature, much more tolerant than solid state devices are.
It takes a couple of "cycles" of AC voltage for a capacitor to come up to fully charged, it's not some "instantanious" thing happening immediately.
Yes, there's a higher than normal operating current during that time, but with each AC cycle it reduces substantially.
There is also a loading of the power transformer to take into account during this time.
I wouldn't go nuts over worrying about all this, because a properly designed tube circuit is trouble-free for most of its life - as was/still is, my 18 tube stereo console, and millions of others.
Some might think that once a rectifier tube is powered, there's some sort of dangerous "slam" effect with current when charging filter caps - there isn't any danger, actually.
Not unless you load the damn rectifier with rediculously large filtering past its design level.
(the more uF is better capacitor syndrome)
What you have to understand is...... tubes are very much a "forgiving" creature, much more tolerant than solid state devices are.
It takes a couple of "cycles" of AC voltage for a capacitor to come up to fully charged, it's not some "instantanious" thing happening immediately.
Yes, there's a higher than normal operating current during that time, but with each AC cycle it reduces substantially.
There is also a loading of the power transformer to take into account during this time.
I wouldn't go nuts over worrying about all this, because a properly designed tube circuit is trouble-free for most of its life - as was/still is, my 18 tube stereo console, and millions of others.
Don´t forget an important detail:
a cathode self biased tube, by definition, is a self current limited device, which is safer.
Higher than normal current through the cathode resistors automatically biases grid more negative so self limiting current.
Yes, there is a small time constant until cathode cap is charged to cathode voltage, but energy involved charging a cap to, say, 25 or 30V is WAY smaller than that involved in charging one to 300/400V
First the Short Answer. For the curious the details are in the pdf attached from the article published Aug 2004 in AudioXpress magazine.🙂
A solution is possible through the use of a pair of high voltage power FET’s, in this case IRF840’s. I had originally tried using a single FET but found that at turn-on there was more than 500 volts applied to this part of the circuit. That would not be good for reliability since the IRF840 is rated for 500 volts max from source to drain so that there are now a pair of FET’s (Q1 & Q2) in a series connection.
A solution is possible through the use of a pair of high voltage power FET’s, in this case IRF840’s. I had originally tried using a single FET but found that at turn-on there was more than 500 volts applied to this part of the circuit. That would not be good for reliability since the IRF840 is rated for 500 volts max from source to drain so that there are now a pair of FET’s (Q1 & Q2) in a series connection.