Variac Minimum Voltage

[ululufut]

Hi All,

I hope I have selected the right category for this question, and that an obvious answer doesn't already exist elsewhere. I did a keyword search on the forum and didn't find anything. My question is of the "what" and "why" category. The first is a practical question: To the best of my understanding, the lower the source voltage at which one operates a tube amplifier, the faster the tubes wear out. Assuming this is true, how quickly - in general terms ("very" quickly, "after a while", etc.) will I be frying tubes if I am cranking, say, a plexi amp, and running at 70 V? More generally, my question is, how low can I go on the voltage before I am inviting unnecessary risk?

My "why" question is this: It is obvious to me why it is hazardous to operate an amp above its intended source voltage. Why, though, is operating an amp at low voltage also risky? I assume that filaments won't heat up as much, but it isn't clear to me how this would negatively affect the usable life of the tube. Also - although I can imagine it might be true - I don't understand how running at low voltage might cause other circuit hazards, such as unintended function of filter capacitors.

I would appreciate any insights on this, and I apologize in advance if I have selected an improper forum category. In short, I want to run my amp at low voltage for some recording; I don't want to damage the amp, and I am interested in understanding the technical aspects what the hazards are. In terms of what I will understand in a technical answer, I am by no means an expert, but I have built a few semi-original amp circuits, and I can read schematics. Thanks!

[acheld]

OK, I know I'm the the most knowledgeable person here, but I had not heard (and can't imagine why) that operating your amp at low voltages will harm it.   I'm skeptical that this is true -- what is your source for this?

It is certainly true that an amp will not sound right if operated with too low of a voltage.  All of your bias voltages will be off, probably not predictably.

What effect are you trying to achieve by recording at low voltage?   I suspect there are better ways to achieve your desired outcome than to run the amp off of a variac.

[ululufut]

Thanks very much for your perspective. EL34 World is the only forum on the internet where I would consider perspectives on this - and anything else regarding amp circuits - to be reliable. Thus far, your opinion, and the opinion of a relative of mine who designs integrated circuits for a living, are essentially the same.

I've looked up info on using variable transformers hundreds of times online and never found a reliable answer. Looking for insight on this topic is just about the easiest way to find misinformation on the internet. Even a major international retailer has an article online saying this:

"Let [the amp] warm up at normal voltage for 20 to 30 minutes. Then begin playing your amp at a moderate volume while SLOWLY turning down the output voltage of the variac."

I can't see why lowering the voltage needs to be done slowly; when you turn an amp of, the source voltage drops to zero nearly immediately.

As to why I would want to use a variable transformer to shape tone, I acknowledge it's just one of multiple ways to do it. The only tube amp I have at the moment that isn't on my shelf waiting for me to finish assembling it is a 20W Marshall SV20H plexi head. It doesn't have MV, so I need to turn it up or run pedals in front of it. It was just very recently that I said to myself "Hey, I have a variac in the garage - if I'm trying to get a convincing plexi tone at a lower volume, why not try it?". It just happened to work really well for what I wanted. After tweaking endlessly with this pedal and that, I turned the two volume knobs up, lowered the voltage to about 70 volts, and voilà! Instant plexi tone and response. I don't like the sound of traditional load boxes very much. I could try more modern IR stuff, the Ox, Rockcrusher, etc. but they're expensive! If the variac works terrifically well in the studio and doesn't present any real hazard, why not use it?

[dwinstonwood]

...I know I'm the the most knowledgeable person here...

Wait, I thought I was.

[Cirrus]

Using a variac, you're reducing all the voltages in the amp, including the heaters. If you drop the heater voltage below a certain point, the valves wear out faster when in use.

In an ideal world, you'd just reduced the high voltage being applied to the plates, and if applicable adjust the bias voltage as appropriate.

[jordan86]

Using a variac, you're reducing all the voltages in the amp, including the heaters. If you drop the heater voltage below a certain point, the valves wear out faster when in use.

In an ideal world, you'd just reduced the high voltage being applied to the plates, and if applicable adjust the bias voltage as appropriate.

This is the reasoning I have heard as well. It’s more an issue of your heater voltage not getting too low. And the number that gets thrown out is ~90v.

[mresistor]

There are people like EVH and Robin Ford that have not just lowered AC voltage to an amp they have performed like that.  They call it the brown sound, I think they are lowering it until they get the sound they want and I don't think the amps suffered any damage.  I have no idea of the exact voltage they were operating at though, and 70 VAC input seems a tad low, but then again I haven't gone much below 110 myself.
If you've got the money and tubes  wearing some out is not a factor really.
A cathode biased amp would be better wouldn't it?

[acheld]

OK, I know I'm the the most knowledgeable person here,

Ooops!   

Well, you know what I meant  . . . 

[ululufut]

Using a variac, you're reducing all the voltages in the amp, including the heaters. If you drop the heater voltage below a certain point, the valves wear out faster when in use.

In an ideal world, you'd just reduced the high voltage being applied to the plates, and if applicable adjust the bias voltage as appropriate.

Thanks - yes, I now remember heater filaments being the main issue. I am still uncertain about why operating filaments at lower voltage|temperature causes them to wear out more quickly - I guess knowing that it does is understanding enough. I did throw caution to the wind yesterday evening and recorded some tracks with the voltage lowered to 70 V. After struggling for months to get a recorded distorted tone that sits well in the mix, this worked like a charm. I don't think I'll be making a habit of it though!

[PRR]

To get lots of electrons from a cool cathode you can "rip them off" with high voltage. This mechanically damages the oxide.

It is a real concern on 25,000 Watt TV transmitter tubes, where they must use miserly metals and they must work everything to the raw edge. It is a non-concern for our tuner and preamp tubes; the oxide has lot of reserve. For our ~~25W power tubes... I dunno.

For recording: for some people, a couple hundred dollars of used-up tube may trade for a million dollars in album sales. Ah, there's no albums anymore?

[trobbins]

I think the hazards to tube life relate to a variety of cathode related functions as there are a lot of technical concerns that are best managed by sitting within 10% of the nominal heater voltage.  The heater allows the cathode to operate at a temperature where it maintains an adequate electron cloud to provide anode current demands, but also to isolate the cathode from damage due to stray molecules in the vacuum (which the getter won't neutralise as efficiently if the getter is colder), and to maintain the cathode surface health (as there is complex chemistry going on to maintain the health of the surface).  RCA 1962 book does a deep dive into valve technical operation, and there may be some pertinent sections that point to such issues with lower heater temp, although you may need chemistry and physics 101 refresher courses as preparation.

Certainly the anode and screen idle voltages will fall with mains voltage reduction, and a base level look at valve curves and operating points show that signal swings get smaller before hitting grossly non-linear regions, and idle operating points may get noticeably 'hotter' or 'colder' and so the onset of clipping to one side gets more sensitive - but that is somewhat dependant on the actual circuit and valves being listened to, so maybe not a generic outcome.

There are also some vintage 'valve reliability' references that may point to references or example reports of actual damage, but all too often the discussion may just be subjective as to hazards.

[dude]

What about a VVR, reduces the voltage to tubes but heater voltage stays where it belongs..? A VVR set up to lower “both” power tube voltages and preamp voltage, might give you the tone you’re looking for…