Weird warmup question heater->B+ delay

[dmoose5835]

So, since I have the luxury of space and time, if I wanted to *measure* something to tell me when to hit the HT supplies, what would I watch for?

Assume I have a Raspberry Pi or Arduino set up to watch the heater current and bias supply voltage, what am I looking for?
I think it would be the heater current to peak and then flatten out and for the bias rail to reach its expected voltage. Is that correct?

I have the HT switching figured out - a solid-state relay in parallel with an EM relay should mitigate any arcing.

Cheers!

[Bergheim]

You mean how long to wait after power on before flipping the standby switch?

[dmoose5835]

Yes, but automate that using some measurement other than just time.

[Bergheim]

Well, the most obvious trigger would to be something that detects the presence of current in the tube, but that is obviously impossible because of the standby switch. Another method would be a temperature sensor that monitors the cathode temperature (precise measurement and shortest time to turn the B+ on, but tricky to implement), or the temperature of the whole tube (easy to implement, but inaccurate measurement and probably unneccessarily slow trigger time).

Have you considered the option to remove the standby switch entirely?

[R.G.]

I mostly agree with Bergheim. Have you considered simply removing the standby switch, or converting it to a switch mute on the power amp input? There exists a line of reasoning behind both of these.

Measuring something before turning on the B+ immediately brings up the question - how many things would you need to measure? Certainly the power tubes. Amps use one, two or four except for some rare outliers. Ideally you'd measure the temperature of the cathode on each in case one is slower. Optical measurement of the color of the cathode glow, but that gets tricky to implement; besides, different makers' tubes may have different glows.

One scheme I thought of was to measure the heater current to all the tubes. This starts out high when the power switch turns on, and reduces to a steady state when they're hot. There are chips that use Hall effect to measure current that could do this. you would read the peak then look for it to hit steady state.with

Frankly, time is probably the simplest solution. I'm guessing here, but a closely related issue is why you're doing this. Most people who dink with not putting B+ on cold tubes do it to prevent cathode stripping. I've read several studied that say that cathode stripping does not happen with B+ less than 600V or so. I'd have to go look up the papers again.

Another possible solution might be to just slow down the fill rate on the first filter cap with a current clamp. A MOSFET can be rigged to limit the current. If the limit is a little bit more tan the charging rate for the amp in full operation, this would ramp up the first capacitor voltage slowly at power-on/standby-off. A suitable circuit is at geofex.com

[Bergheim]

Said another way, if you remove the standby switch, the B+ will turn itself on precisely when the tubes are hot enough
Pretty much the only thing to watch out for is that all filter caps can withstand the unloaded B+ voltage.

[dmoose5835]

One scheme I thought of was to measure the heater current to all the tubes. This starts out high when the power switch turns on, and reduces to a steady state when they're hot. There are chips that use Hall effect to measure current that could do this. you would read the peak then look for it to hit steady state.with

Thanks R.G. - something like this is where I was headed, to maximize tube life by avoiding cathode stripping. I can hold off the B+ with a hybrid relay until the tubes are fully ready. Maybe slow down the B+ ramp with a bypassable series R as well.

[maxkracht]

https://valvewizard.co.uk/standby.html

"If standby switches really did have magical life-extending properties then we would expect to see it mentioned from time to time in proper textbooks and valve manuals, yet such discussion is conspicuous by its absence. Indeed, some texts explicitly exclude audio valves from discussion of similar pre-heating switches (mainly with regard to radio transmitters). Despite this, promotional materials glorifying the supposed effects of standby switches are often published (always on the internet and in non-technical magazines, never in real academic work) even by well-known guitar amp manufacturers, presumably with good-but-misguided intentions. So such myths become self propagating.

You may have heard of 'cathode stripping', which is a specious argument wheeled out by standby-switch obsessives. In its purest form, cathode stripping occurs when particles of the oxide coating are physically torn from the surface of the cathode when it is exposed to a powerful electrostatic field from the anode. This would happen if the valve is operated at saturation, without a usual space-charge of electrons to protect it. Fortunately, this effect does not exist in receiving valves, even when operated at saturation, because it requires an electric field strength of at least 4MV/m (yes, 4 million volts per metre!). No guitar amp ever comes close to this.
Another type of cathode stripping occurs when stray gas molecules in the valve become ionised by the electron stream. The positive ions will then be accelerated towards the more negative grid and cathode. If these manage to miss the grid then they may crash into the cathode, physically damaging its surface. The proper name for this process is cathode sputtering. Sputtering is a known problem in gas tubes and transmitting valves operating at kilovolt levels, near saturation. It doesn't occur to any significant degree in ordinary audio circuits. Note that even the RCA Transmitting Tubes Technical Manual No. 4, p65, states: “Voltage should not be applied to the plates or anodes of vacuum, mercury-vapor, or inert-gas rectifier tubes (except receiving types) until the filaments or cathodes have reached normal operating temperature” [My emphasis].
Receiving valves are the small kind used in radio receivers, i.e audio valves like those in guitar amps, in case you were wondering.

Cathode stripping should not be confused with cathode poisoning. Cathode poisoning refers to chemical –rather than mechanical– processes occurring at the cathode. There are several forms of cathode poisoning, including absorption of gas into the oxide coating, but the most pernicious type is the growth of interface resistance. When a valve cathode is fully heated but no anode current is allowed to flow for long periods of time (several hours), a high-resistance chemical layer can grow between the cathode tube and the oxide coating. This has an effect like an unbypassed cathode resistor; it increases noise and reduces the useful gain of the valve even though the oxide coating may have plenty of life left in it. This really does happen in receiving valves, and once formed it cannot be removed again.

The two main causes of valve ageing are natural barium evaporation from the cathode, and interface resistance growth. Barium evaporation continues as long as the cathode is heated, so an ordinary standby switch has no effect on this. But a standby switch does encourage interface resistance growth. In other words, the standby switch is more likely to shorten the life of the valves!

And there's more. Valve (vacuum) rectifiers should always be allowed to charge the reservoir naturally from cold. If the valve is preheated before the reservoir is allowed to charge, the valve will have to supply the full inrush current when the switch is finally thrown. This is called hot switching and causes sudden cathode saturation that can lead to catastrphic arcing inside the tube. Hot switching of rectifier valves was usually forbidden by valve manufacturers. Opening a standby switch can also induce a ghastly flyback voltage across the transformer winding, large enough to cause arcing in a valve rectifier (a precaution against this is to add ordinary silicon diodes in series with each anode of the valve to reduce the reverse voltage across it). Yet more reasons why standby switches are bad news."

[dmoose5835]

Neat! An actual explanation.