I had read several places that it's a good idea to delay the turn-on of B+ to the plates until after heaters are on. Something about not hitting the anodes with high DC voltage until the heaters are working. But then more recently I've read the opposite, that you shouldn't delay B+, but rather should perhaps switch it on first. Not sure of the rationale given for this approach.
Can anyone (or lots of anyones) provide some thoughts on this? I just don't have enough experience or theoretical knowledge to make sense of these two options.
Thanks!
Can anyone (or lots of anyones) provide some thoughts on this? I just don't have enough experience or theoretical knowledge to make sense of these two options.
Thanks!
Hasn't really been my experience in 20 some odd years of designing tube kit, but it is not a bad idea nonetheless..
Cathode stripping is largely a myth at the voltages present in consumer gear, but there are certain types which are prone to this kind of damage including 6080, 6AS7G, 6336, 6C33, and derivatives - all low perveance types originally designed for voltage regulator service. They need to fully warm up before B+ is applied. (Some people use these in SE amps, and voltage regulators, but are more commonly found in OTL amps.)
Using a slow heating rectifier like the 5AR4/GZ34, 5V4, GZ32, GZ37, etc will usually suffice. Switching the raw B+ sometimes leads to problems with switch arcing, fried solid state rectifiers, and blown fuses.. (Switching the primary is usually fine)
Well, there's several hundred million TV sets and radios over the last century that did not feature this +B standby or delayed switching....
Tube life was fine with them....
Never done it myself, and not had any issues....
Tube life was fine with them....
Never done it myself, and not had any issues....
Running valves with hot heaters but no HT and therefore no cathode current can cause cathode interface problems, except for special valves designed to avoid this (mainly SQ or computer versions).
Running valves with warm heaters and full HT can cause ion bombardment of the cathode. Suddenly applying full voltage to hot valves can create circuit surges. You choose which damage mechanism you prefer.
A valve rectifier avoids the problem, except in the rectifier itself. Standby switches may be needed for expensive transmitter valves, but not for ordinary audio. They may do more harm than good.
Running valves with warm heaters and full HT can cause ion bombardment of the cathode. Suddenly applying full voltage to hot valves can create circuit surges. You choose which damage mechanism you prefer.
A valve rectifier avoids the problem, except in the rectifier itself. Standby switches may be needed for expensive transmitter valves, but not for ordinary audio. They may do more harm than good.
The root of this thinking incorrectly filters down from industrial and transmitting tubes and their use with voltages upwards of 1KV. Emission can be damaged and arc over can happen with some indstrial gas-filled rectifiers and thyratrons. Receiving tubes operating with potentials of around 500-600 volts are normally not subject to to these failures. This is old-school truth and still applies today.
However, because of the poor quality of todays foreign made power/output tubes, some limiting of applied voltage may be helpful in preserving them.
I built an RH84 amplifier, originally using a 5Y3 tube as a rectifier.
Changed that out to a 6087 tube, that gives me around a 20 second warm up.
Steve
Changed that out to a 6087 tube, that gives me around a 20 second warm up.
Steve
Many of these are $5 to $10 tubes. Assuming that instant-on B+ simply extends their life some, you would need to save quite a few of them to defray the cost of the slow start circuit.
My question came from my build of a SE Pentode (triode strapped) for which I have separate A and B iron (a Rat Shack 12.6v trafo for heaters, simple isolation iron with doubler for B+). I typically wait 10-15 seconds or so before throwing the B+ switch. I shouldn't have to worry about arcing b/c I'm switching the primaries. I included a switch for the B+ iron separate from the heater basically b/c that's what I had read about. But then I realized I really didn't know why I would do that, so my question.
Sounds a bit like it's not too important either way with the possible exception of certain tubes. Interestingly, I have a battery powered pre (Bottlehead Quickie) which has no HT switch (48v applied to anodes all the time). Power switch is for the 1.5v heaters only. No heaters, no conduction, so the B+ batteries don't wear out. But in this case, there's no sudden onset of full B+ (and it's only 48v), so no problems. I guess.
Thanks for all the great replies.
Sounds a bit like it's not too important either way with the possible exception of certain tubes. Interestingly, I have a battery powered pre (Bottlehead Quickie) which has no HT switch (48v applied to anodes all the time). Power switch is for the 1.5v heaters only. No heaters, no conduction, so the B+ batteries don't wear out. But in this case, there's no sudden onset of full B+ (and it's only 48v), so no problems. I guess.
Thanks for all the great replies.
Use a damper diode in series with the HV supply between filter caps. They can take the voltage and the current for most applications. There is a 10volt drop at 250ma and they come on line as the rest of the tubes heat up preventing spiking in the power supply.
No, not if the B+ supply is less than about 1,000 volts. What it is hard on is the filter caps. Typically what happens is the B+ comes up very fast, many seconds before the heaters heat and cause the power tubes to draw current. So the B+ supply is very high with no load and may reach the voltage limits of the filter caps.
Fender and others added "standby" switches to their amps because the caps were under spec'd. Nothing to do with the tubes. Back in the day (mid 50's) filter caps were expensive and the B+ switch allowed them to save a few bucks and use caps that were spec'd for a few less volts.
Last edited:
I use a 5AR4 which provides sufficient delay. They fail very quickly otherwise - I made this mistake with the 6C33 and had several arc over during warm up - often fatal for the tube in question. The 6336 is expensive and equally fragile when cold.. 😀
You know?
That IS odd for you to have issues with the 6C33 arcing on warm-up...
--I'm curious why....
My OTL's all use the 6C33, there's no +B delay and never had any issues, either a fixed-bias circlotron at 200V +B or the auto-bias Futterman running +- 150V rails...
The Only Ever time I had issues was with two tubes that I hadn't burned in.
--Lesson learned with That though!--My fault for not treating the new tubes properly....
They arced about an hour or so after first switch on, not on warm-up though...
I prefer delay circuits to allow the cathodes to come to temp. These days, power tubes are too expensive to risk not doing so. I prefer choke input supplies, so inrush current isn't a problem. I like relays, and arcing can be avoided by using a zero-crossing thyristor circuit to switch the HV before the relay closes. Or use a mercury relay
Damper diodes do make nice turn-on delay devices. The current draw during their slow warm up will probably cause them an early demise, but they're cheap enough.
Damper diodes do make nice turn-on delay devices. The current draw during their slow warm up will probably cause them an early demise, but they're cheap enough.
Its absolutely amazing that the electronics from years ago actually survived not having a delayed B+ supply. Among them are my RCA Radiola Model 62 and my Atwater Kent Model 55 and model 60 radios from 1928 not to mention a pair of Wurlitzer mono block amps from the early 40's. Is it possible that these designers from the past knew more than we do or is delayed B+ really needed? Is cathode stripping as common as some are lead to believe? I was told this by a bright individual on this forum years ago and that was as long as signal was not present when the device was turned on meaning the volume control turned up and the device warming up trying to produce signal that no damage was done. I have practiced safe signal precautions since then and have suffered no problems. When faced with a situation of high B+ and a rectifier with no delay upon warm up I do use a damper tube to soften the B+ inrush or I will add additional diodes to the rectifier tube so that they take the brunt of the in rush load.
My own experience puts me in the "no need to delay" camp also. I've been using 6C33C and EL509S in OTL amps for some years, without using B+ delays, and never had any trouble. It is hard to believe, especially at the low voltages used typically in an OTL (about 150V), that anything untoward could happen. I did include a 2-minute delay in the first OTL I built, but if anything I could imagine it puts more of a strain on the tubes because of the turn-on surge than anything they might suffer because of immediate application of the B+. It certainly put a strain on my nerves, the first time I switched on, waiting for the B+ to kick in and not being sure what would happen!
Chris
Chris
Absolutely with you there on all points!
Personally I think that +B delay on certain types like Cathode-Bias power-amps is not a good idea, as the instant that +B hits those fully heated cathodes, they are going to be passing a huge surge as the cathode by-pass caps charge up to establish the 'auto'/self bias point....
As I said, several hundred Million TV sets and radios of the last century CANT ALL be Wrong!
--There's loads still working round the world, and I know many will have the original tubes in too...
+B delay only needs to be done on transmitter PA stages running a kilovolt or so and above, or for non biassed RF PA stages that rely on input signal to bias up properly, where without drive, excess anode current can easily destroy the tube....
I'm definately in the 'Non Delay' camp...😀
So I am.
And don't forget that the link caps in the grids had to charge up too 😱
I've seen a self biased pair of 6550 with gigantic (more than 1µF Russian PIO) grid link caps driving more than a full ampere for some fractions of second.
Bah !
Yves.
You can use a "Quench-Arc" from Mallory -- it allows the "spark" to be coupled capacitatively between the plates of the switch or relay.
With respect to the hundreds of millions of TV sets -- that's why RCA, GE, Sylvania, Philco etc. made TV Sets and tubes! Most of those I took apart to build my ham gear in the 1960's had 5U4GB rectifiers.