Perils of series heaters?

[hodad]

I have a Peavey VMP-2 tube mic pre. 8 12A_7 tubes with the heaters in series. The first tube after the heater supply is a 12AT7. A couple of times I've had problems with the tube in this particular position, & I'm wondering if the jolt from being first to the juice might be doing damage to the tubes.
I have swapped the same AT7 between the two channels, & when in the problem position the filament does burn brighter & light up faster.

Is there a potential for trouble here or am I looking in the wrong direction?

Thanks
Tom

 

[bcarso]

Current flow in wires and at at these frequencies is not like water gushing through an initially empty pipe. The pipe is full, and the water moves all at once. The last in a chain sees the same current as the first---position is of no consequence.

 

[CJ]

those little bottles have a rep for noise.
why series heaters?
Whats the heater violtage?
I would wire the tubes for 6.3, which takes advantage of noise cancelling, I think.

 

[PRR]

> in the problem position the filament does burn brighter

Carefully get your voltmeter in there and measure the voltage at each heater pin. It is almost certainly DC.

As mentioned: a simple series circuit flows the same current everywhere. Yet you say this one socket runs hot. Look around that and see if they are tapping current to go somewhere else than the other heaters.

 

[analag]

If the tube socket is made of plastic the possiblity exist where it might be shorting. Check the voltage at the pins of the socket and compare it to the other channel, if there is a difference....you know the rest. Is all the tubes connected to the same series heater string? If so check the socket better yet change it, since the shorting of the socket could mean only a break down in the high insulation resistance that should exist between the pins. I recently solved a problem similar to that by kicking out the plastic and inserting some ceramics. Hope that helped, bro.

 

[clintrubber]

A possible clue is in the datasheet. 12AT7 is ECC81, IIRIC.

Maximum cathode to heater voltage for ECC82 & ECC83 is 180 V (limiting value).

But for the ECC81 it's only 90 V !

So how high does that first tube ride with its heater above the cathode ?
The highest side will at least be 8*6.3V above ground, but when they're in series that'll be double AND exceeding the limiting value for the 12AT7, be it only a bit.
(and it's assuming a cathode of just a few volts above ground)

So it might at least be an influence - dunno if it's also the cause.


How things will be further influenced by the polarity I don't know. I mean, usually that Vkf will be swapped: cathode high above the heater, like in a cathode follower.
Here in that Peavey it might be unlike this: high riding heater and a cathode a few volts above ground.

If that box only uses 12AT7's it might be of use to put eventual cathode-follower tubes (let's hope they're grouped in the same tube) as the first tube in the heater chain.

Bye,

Peter

 

[bcarso]

That's an intriguing possibility Peter. I'd think, though, that it would have other gross effects on circuit operation.

 

[clintrubber]

[quote author="bcarso"]That's an intriguing possibility Peter. .[/quote]
I've no idea if this is really an influence, but if those heaters are done in 12.6V style then the first 12AT7 is likely too high above the cathode.
I understood that Fender happily made designs where tubes where used outside of their guaranteed ranges. But when a simple & obvious zero cost swap of place in the heater-chain could prevent an eventual problem with Vkf then Peavey should have seen this possibility.

But it's all speculations, is there a schematic around of this box ?

I'd think, though, that it would have other gross effects on circuit operation.

Sorry, how do you mean ? The influence of a large and either positive or negative Vkf ?

Regards,

Peter

 

[hodad]

thanks to everybody for advice. I was confusing when I said heaters were wired in series--Each individual tube is wired parallel, but the the tubes collectively are wired in series. I'll try to do some measuring soon & report back.

Tom

 

[clintrubber]

[quote author="hodad"]thanks to everybody for advice. I was confusing when I said heaters were wired in series--Each individual tube is wired parallel, but the the tubes collectively are wired in series. I'll try to do some measuring soon & report back.

Tom[/quote]

8 * 6.3V is less than the 90 V limit, so that possible influence won't be it. I'm curious though why the '81 (12AT7) has half the Vkf rating of the ECC82 & ECC83. Something related to the contruction ? Certain parameters ?

Bye,

Peter

 

[bcarso]

[quote author="clintrubber"][quote author="bcarso"]That's an intriguing possibility Peter. .[/quote]


But it's all speculations, is there a schematic around of this box ?

I'd think, though, that it would have other gross effects on circuit operation.

Sorry, how do you mean ? The influence of a large and either positive or negative Vkf ?

Regards,

Peter[/quote]

I mean, we are talking about LARGE currents (in order that some significant difference be made in the brightness and speed of lighting up of a filament. Unless the cathode is hard-strapped to common (again schematic needed here) that current would be wreaking havoc with signal and bias.

 

[hifizen]

Hmm... are you using the "A"-suffix version of the tube? ... eg: 12AT7A

My understanding is that most tube types were evolved into "A" versions as series heater connections came into common use, or had an A version specifically made for use in equipment with series heaters. Types with the A suffix are referred to as "controlled warm-up" or "controlled filament" types, and the heater is specially designed so that it's warm-up time falls within a narrow range. Since tube heaters, like lamp filaments, have a strongly positive temperature coefficient, if a non-A tube is used in a series filament string, it may hog a disproportionate amount of power during warm-up, causing excessive wearing of the tube, and possibly leaving it hotter than the other tubes and continuing to hog filament power.

Also (I haven't checked this, but I suspect it) the non-A versions may have lower filament-to-cathode voltage ratings, since they weren't intended for series use in a high voltage heater string.

 

[CJ]

The "A" can mean anything, from mechanical mounting differences, to improved of different ratings for heaters, plate voltages, off state leakage currents for tube computers, on and on.

The "A" in 12"A"T7 means amplifier, at least to RCA.

As far as the "A" in the suffix, don't know....maybe the search engine does.

 

[hifizen]

Oh, I was only referring to the A suffix, as in 12B4 vs 12B4A ... 12AT7 was a bad example, I suppose. As you say, the letters in the middle seem to be almost random - I figure a good portion of the tubes out there got that part of their name from tube designers' little black books. :wink:

But, looking through some data, the A suffix indicating controlled heater warm-up isn't entirely universal. I see some cases where an A suffix indicates improved performance, usually as part of a "GT" suffix... GT, GTA, GTB series, each indicating an evolution of the original tube design.

Anyway, controlled heater warm-up is definitely something required of a tube used in a series string. I was under the impression that this was most commonly employed in inexpensive 5-tube radios and such, where the cost reduction of eliminating the power transformer was a competitive necessity. So I'd expect to find controlled heaters on the radio tube types most often. I don't work with guitar amps much, but I'm still mildly surprised to hear of series heaters being used in one - maybe they're more common than I thought?

 

[ciminosound]

Controlled heater warm-up seems to be a standard feature in the "odd" heater voltage tubes, like the 4/5/7/9 volt variants.

 

[PRR]

> controlled heater warm-up is definitely something required of a tube used in a series string.

No, it's not.

Yes, a very few odd tubes did give trouble in series strings. Most tubes played well together. The original AC/DC radios used perfectly ordinary tubes not rated for series use, and worked well.

Also, most 12V tubes WERE aimed at series-strings. 2.5V and 6.3V tubes were mostly aimed at parallel operation. The standard US 5-tube radio was three 12V tubes, plus 25V, 35V, or 50V power and rectifier tubes. The 12A?7 and similar 6V/12V tubes are obviously aimed at both markets. And when such tubes became the BULK of the total tube market, most heaters in all tubes used similar construction and had similar warmup.

But someone found a failure and blamed it on unequal warmup. There was a rush to specify heater warmup rate and certify tubes. That is some (but not all!!) of the "A" suffixes. Truth is, that TV sets did many crude things to tubes, tubes failed, and not due to unequal heater warmup. It was a fad.

-------

> mildly surprised to hear of series heaters being used in {a guitar amp}

It's a Peavey, but not a guitar amp.

The logical reason (today) to go series is for DC heat. A 6V DC supply is wasteful due to the 0.6V-2V diode drop and too low voltage for standard caps which often stop at 16V. A 60V supply is more efficient and can use 100V caps well. Series DC heat makes tube-on-PCB possible: one wire between tubes (instead of two) and no buzz problem. AC parallel pretty much demands hand-wired flying heater leads (or a 4-layer PCB, which is way too costly).

-------

The heater-cathode voltage breakdown rating refers to a thin layer of baked clay between heater and cathode. If it's good, it's good; if it breaks down it is a heater-cathode short and usually a total failure of that stage.

The 12AT7 was always marketed as a VHF tuner tube. For various reasons, series-string systems always put the tuner at the bottom of the string. Heater-cathode voltage was just a few volts. Usually one heater pin was bent over and soldered to the chassis, as both heater return and RF shielding. While 12AT7 was used for many other things, apparently nobody ever asked for the generic 12AT7 type to be re-specified for higher heater-cathode voltage. Julie/Burr-Brown were using the 12AX7 at extreme cathode voltages so it did carry a high rating.

 

[hifizen]

aaaaaahhhh.... a little historical perspective goes a long way! Thanks for the clarification PRR. Curious that 'controlled' warm-up certification was just a fad. Then again, one can never underestimate specmanship. :roll:

> It's a Peavey, but not a guitar amp.

Whoops, did I say that? My bad. Should've re-read back to the beginning of the thread... series DC heaters in a mic pre does make sense.

The question that remains in the back of my mind is: are filament tempcos strong enough that a series-connected filament which heats up prematurely could establish a 'foothold' and then persistently hog a disproportionate amount of voltage? Or, will the filament string still tend to even out anyway?

 

[bcarso]

[quote author="hifizen"]The question that remains in the back of my mind is: are filament tempcos strong enough that a series-connected filament which heats up prematurely could establish a 'foothold' and then persistently hog a disproportionate amount of voltage? Or, will the filament string still tend to even out anyway?[/quote]

Now that is an interesting problem. I wish I had time to think about it in depth.

The empirical short answer must be no, at least not to a great extent, since otherwise disparate tempco heaters and/or different thermal mass heaters mixed together would have produced noticeable decreases in reliability. Maybe it would be hard to notice though---after all light bulbs have a much larger temp excursion routinely and don't burn out all that quickly.

There are a lot of variables, though, in framing the problem.

 

[43hertz]

Aah, finally a topic (old radios) I feel like I know something about!

While all of this discussion of series heaters is very interesting, particularly the potential current-hogging of the "first-to-glow", I'm pretty sure that this is not happening in the Peavey product in question. (Seems like Hodad's last post rectified his error in terms, or am I mis-reading.)

Is anyone aware of a currently manufactured product that uses series heaters (not DC). Maybe there are still some (?), but I thought those had pretty much gone away due to safety issues! (If you've ever grabbed the metal chassis of an old radio where the AC has got joined to the frame
you'll know what I mean!)

 

[PRR]

> Is anyone aware of a currently manufactured product that uses series heaters

The main reason would be when you have an appropriate voltage at zero cost.

I sure hope all the hot-chassis designs are long-gone. They can be perfectly safe when you have no electrical input or output, such as a radio: input is radio waves, output is sound, and you can wrap the whole thing in Bakelite. TV, same idea with light output too. You can couple 100MC, even 1MC, with RF transformers caps so small that 60CPS leakage is near zero.

There are surely some +/-12v hybrids running a couple 12V heaters in series. These days people are crazy for DC heat, but if you really search you may find someone who tapped the raw AC from the transformer.