Why? No reason why it shouldn't work well that I can see - what is it that concerns you?
Of course I think that current control for series connection is a good idea, and that ramped start-up for all valves is quite possibly valuable, if it can be achieved economically, and more importantly, reliably.....
Something to verify,
I think the original 12ax7's were designed to have their heaters connected in series.
But the 12ax7a versions, allowed to have the heaters connected in parallel.
Incandescent light bulbs, usually burn out when first switched on.
The same thing for the heater section of a tube.
BTW: When power is applied to a tube, the heater can draw more than double what the spec's state for current draw.
So for this reason, current regulation for the heater element prolongs the tube life
much better than voltage regulation.
I think the original 12ax7's were designed to have their heaters connected in series.
But the 12ax7a versions, allowed to have the heaters connected in parallel.
Incandescent light bulbs, usually burn out when first switched on.
The same thing for the heater section of a tube.
BTW: When power is applied to a tube, the heater can draw more than double what the spec's state for current draw.
So for this reason, current regulation for the heater element prolongs the tube life
much better than voltage regulation.
The US valve/tube numbering system approximately indicates the heater voltage in the first number - thus a 12AX7 has a 12.6v heater, whilst a 6L6 is 6.3v. The "12" series valves were apparently designed for use in car radios, with HT from a "vibrator" type supply.
The lightbulb analogy is worth considering, not because heater failure is common; as Enzo has pointed out, it isn't, but because of the underlying cause. Mechanical stress due to high thermal gradients in materials with none-zero temperature coefficients of expansion.
The glass of a valve envelope is very well matched for tempco with the material of the pins. But when there's a difference in temperature between the glass and the pins, there's still a problem.
This same principle will apply throughout the whole valve.
I can't begin to quantify this, but the slower a valve heats up and cools down, the smaller this effect will be.
The lightbulb analogy is worth considering, not because heater failure is common; as Enzo has pointed out, it isn't, but because of the underlying cause. Mechanical stress due to high thermal gradients in materials with none-zero temperature coefficients of expansion.
The glass of a valve envelope is very well matched for tempco with the material of the pins. But when there's a difference in temperature between the glass and the pins, there's still a problem.
This same principle will apply throughout the whole valve.
I can't begin to quantify this, but the slower a valve heats up and cools down, the smaller this effect will be.
Simon,
I suspect you are right - Valves (tubes) don't like thermal [or physical] shock.
Even Wine or Beer doesn't like it.
There have been tests on tubes - and it seems that using current regulation for the heater - instead of voltage regulation - prolongs a tube's life.
Hopefully going a slightly off topic is OK.
Have you seen any curves that describe how the transconductance of a tube changes with use/time ?
Thanks,
.
I suspect you are right - Valves (tubes) don't like thermal [or physical] shock.
Even Wine or Beer doesn't like it.
There have been tests on tubes - and it seems that using current regulation for the heater - instead of voltage regulation - prolongs a tube's life.
Hopefully going a slightly off topic is OK.
Have you seen any curves that describe how the transconductance of a tube changes with use/time ?
Thanks,
.
I mostly participate in pro audio forums, so there is something I need to be cautious of here. Here we have a mainly hifi membership. The needs of a guitar amp are different from those of a hifi sytem. (or a radar or a 1950s computer for that matter). SO someone bringing a hifi sensibility to the discussion may have different assumptions than someone like me.
In a guitar amp, we will be wearing out tubes a lot faster than anyone else would, and we expect to replace power tubes yearly or more often. Just as tires for your car (that is tyres for some of you I suppose) will present diffferent concerns if they are on mom's family car versus junior's drag racer. SO concerns over arcane things like cathode poisoning, or heater lifetime and such are really not an issue to us. The tubes will have ended their useful life long before those things show up. The hifi guy on the other hand will expect a longer life from his tubes and will be operating them in a much more polite manner. And in the case of that ancient computer with its roomful of thousands of tubes, well anything providing even a small incremental improvement in tube life matters.
Nigel, this particular whole amp series string is fairly recent even for Peavey, as far as I know. I can't think of anyone else doing it offhand, but there may be others. Preamp tubes, mainly 12AX7s, have been run in series by any number of amp makers for some time. I am a bit more aware of Peavey designs as I do more repair volume on their products than most others. They sell just tons of amps. They and Fender. Other brands don;t sell nearly as well. But a quick look shows 12AX7s in series across 24vDC in the Peavey Rockmaster preamp, and that was designed in 1991. So it is nothing new.
In a guitar amp, we will be wearing out tubes a lot faster than anyone else would, and we expect to replace power tubes yearly or more often. Just as tires for your car (that is tyres for some of you I suppose) will present diffferent concerns if they are on mom's family car versus junior's drag racer. SO concerns over arcane things like cathode poisoning, or heater lifetime and such are really not an issue to us. The tubes will have ended their useful life long before those things show up. The hifi guy on the other hand will expect a longer life from his tubes and will be operating them in a much more polite manner. And in the case of that ancient computer with its roomful of thousands of tubes, well anything providing even a small incremental improvement in tube life matters.
Nigel, this particular whole amp series string is fairly recent even for Peavey, as far as I know. I can't think of anyone else doing it offhand, but there may be others. Preamp tubes, mainly 12AX7s, have been run in series by any number of amp makers for some time. I am a bit more aware of Peavey designs as I do more repair volume on their products than most others. They sell just tons of amps. They and Fender. Other brands don;t sell nearly as well. But a quick look shows 12AX7s in series across 24vDC in the Peavey Rockmaster preamp, and that was designed in 1991. So it is nothing new.
Good points Enzo, though in fact the increase in life achieved by Tommy Flowers on the Colossus was up to about three times the expected figure, which did stick in my memory.
It's precisely because guitar amps get through tubes so fast that this interests me. I fully accept that actual heater failure is rare - I never did finish the christmas-tree-lights-out-of-dead-valves project, a house move consigned them to the garbage. My dead-bin is growing again though, so maybe this year...they were series connected too, come to think of it.
The physical properties of at least some of the materials used in valve construction, like hardness, ductility etc change a fair bit over the big range of temperature change that valves experience. So it seems possible that over the first 100C of temperature rise a valve might have a different sensitivity to the speed of that change, compared to over the last 100C. And of course some bits get hotter than others.
I'm trying to phrase this in non-partisan terms as far as possible, as I'm interested enough in this to be willing to lose the bet to learn a little.....
In the meantime it seems that the only reliable information applicable to guitar amps is.................
Never turn them off!
well, I claim no particular expertise in the matter other than experience. In the guitar amp world there are always debates about whether to have a standby switch and where to put it and how it is to be implemented, and in those someone always brings up cathode poisoning or stripping, and other tube life things. There may be something to it, but in my experience, no partiicular approach to amp circuits seems to make more of a tube-eater than any other style. Like guitar strings that will continue to function long after they have lost their tone, power tubes wear out and lose their tone long before they get weak emissions or suffer other failures. And unlike hifis, guitar amps are thrown into trucks and hauled down the highway time after time, not to mention the combo style amps have the speaker in the same box as the amp and her tubes, vibrating away.
For an obvious start, the inability to unplug valves and leave it working
There just seems no advantage to it?, and the valve series used aren't even intended to be used in that way - hence the problem over their different heater currents.
There are specific valve series intended to be used in this way, the P series (300mA) and the U series (100mA), the E series (as commonly used in guitar amps) are designed for 6.3V parallel heaters. The double triodes (ECC8x) have the major advantages of centre-tapped heaters, so you can wire them in various ways (it might be sad, but I can still remember the pin connections on the double-triodes).
You could just as well soft-start parallel heaters as well. but is it really worth it? - the technology is all but obselete, about 100 years old, and hasn't needed soft-starting in all that time.
Admittedly valve heaters do VERY occasionally go O/C, and soft-starting 'may' help to reduce that - but I feel the soft-start circuit is likely to be far less reliable than the heaters it feeds.
I made a case for it earlier. A common problem in guitar amps is the high current connections for 6.3VAC burning up on the inter-board wiring and even the transformer wires plugging onto the board. By running them in series like that, they need only a 0.9A connection, and the resulting higher voltage winding can be lighter.
Commercial amps are not designed for running with tubes missing. Many do run that way, but they are not designed with that in mind. When they want to disable a tube, they turn it off via the active elements, not the heaters.
Those particular tubes may not have been designed for series strings, but it works fine and causes no problems, at least in the examples we see.
Commercial amps are not designed for running with tubes missing. Many do run that way, but they are not designed with that in mind. When they want to disable a tube, they turn it off via the active elements, not the heaters.
Those particular tubes may not have been designed for series strings, but it works fine and causes no problems, at least in the examples we see.
Proper construction prevents that happening anyway, so essentially it's a penny pinching measure by shoddy manufacturers
Series/parallel connections have got to be a bad idea, where one heater going O/C will cause the other parallel ones to do the same (so three valves blown where just one failed). Even if one of the 0.9A valves fails, then the amp is completely dead - one output valve failing in a conventionally wired amp will leave it still working for the gig (at least to some extent).
"even the more heated discussions"
Penny has finally dropped, better late than never I suppose!
yes, if one tube heater fails, then either they all go dark or one of the preamp tubes is gone and the others will run hot. However, the VAST majority of tube failures do not involve the heaters. I haven;t seen an open heater 6L6 in decades. I am sure it happened to someone somewhere yesterday, but for most, it just isn't an issue. When tubes do fail, it is generally power tubes shorting between elements. That takes out high voltage fuses most times, and so heater issues become moot.
In an amp like we are discussing, all three of those preamp tubes (I include the phase inverter for convenience) are needed for amp function, so the only case in which a dead tube would leave the amp playable would be a failed power tube. And that leads us back to the almost never happens condition.
I have seat belts and air bags in my car in case I am involved in a collision. Perhaps I should also install ceiling air bags in case a circus elephant tries to stomp my car. Sorry, that was flippant. My ceiling bags might protect me from a tree falling on my car. That actually happens now and then. But not very often, the recent wind storm we had over here being an example.
As to penny pinching, well, one man's penny pinching is another man's not wasting money. Molex connectors have a current rating that limits their use. A manufacturer of amps could use multiple molex connectors, or find some different larger connection materials, sure. Or they could simply do this. And by doing this they allow themselves to use the common connectors they use for everything else.
In an amp like we are discussing, all three of those preamp tubes (I include the phase inverter for convenience) are needed for amp function, so the only case in which a dead tube would leave the amp playable would be a failed power tube. And that leads us back to the almost never happens condition.
I have seat belts and air bags in my car in case I am involved in a collision. Perhaps I should also install ceiling air bags in case a circus elephant tries to stomp my car. Sorry, that was flippant. My ceiling bags might protect me from a tree falling on my car. That actually happens now and then. But not very often, the recent wind storm we had over here being an example.
As to penny pinching, well, one man's penny pinching is another man's not wasting money. Molex connectors have a current rating that limits their use. A manufacturer of amps could use multiple molex connectors, or find some different larger connection materials, sure. Or they could simply do this. And by doing this they allow themselves to use the common connectors they use for everything else.
FWIW, and as the original proponent of current regulation for heaters in this thread:
Out of the uncounted but certainly many hundreds of dead valves that have gone through my hands in the last 30 years or so, I think I've noticed perhaps three or four with open circuit filaments. The rock'n'roll christmas lights idea grew out of my noticing that the filaments usually still worked.
Switch-on and switch-off stress on the electrode structures and the glass-metal seals contributes to valve failure. What portion of guitar amp valve failures are caused by this is unknown - I've known guitarists who it didn't apply to at all, their failures were more that the envelope melted.......
Returning to the OPs question about using a regulator to feed this particualr heater chain, one idea is this:
Use three LM317T regulators, wired as current regulators, in parallel, each carrying 0.3A.
For each regulator, to get the required 1.25v drop across the resistor in the output path, downstream of which the reference ("adj") terminal is connected, R = 4.17 ohms, so 3R9 + 1R2 is close enough. 0.6w metal film should be fine, the 3R9s dissipate 0.351w.
Three regulators to ease heatsinking and increase reliability, the silicon works out cheaper than fancy heatsinks. Should be cheap and reliable, even into a dead short, disspation per regulator is 10w, so heatsink for that and it should be fine.
Out of the uncounted but certainly many hundreds of dead valves that have gone through my hands in the last 30 years or so, I think I've noticed perhaps three or four with open circuit filaments. The rock'n'roll christmas lights idea grew out of my noticing that the filaments usually still worked.
Switch-on and switch-off stress on the electrode structures and the glass-metal seals contributes to valve failure. What portion of guitar amp valve failures are caused by this is unknown - I've known guitarists who it didn't apply to at all, their failures were more that the envelope melted.......
Returning to the OPs question about using a regulator to feed this particualr heater chain, one idea is this:
Use three LM317T regulators, wired as current regulators, in parallel, each carrying 0.3A.
For each regulator, to get the required 1.25v drop across the resistor in the output path, downstream of which the reference ("adj") terminal is connected, R = 4.17 ohms, so 3R9 + 1R2 is close enough. 0.6w metal film should be fine, the 3R9s dissipate 0.351w.
Three regulators to ease heatsinking and increase reliability, the silicon works out cheaper than fancy heatsinks. Should be cheap and reliable, even into a dead short, disspation per regulator is 10w, so heatsink for that and it should be fine.
Like I said in post #29
O/C heaters are VERY rare, I've probably seen more O/C heaters on CRT's than valves? - usually where just one of the three heaters goes O/C so you lose one colour totally. Obviously you can't read it as O/C as they are paralleled, but by careful observation you can spot one has gone out.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.