Downside to Thermistor delayed filiment warmup?

Is there a downside, besides heat, to using thermistors to soften the shock to tube filiments as they warm up? I don't see it used much, but it would seem like a good idea.... unless it is noisy, unreliabe or causes some other problems.

Teacher?

Tod

I assume you mean heaters. Hardly anybody uses naked filaments.

Is there an actual reason to worry about start-up shock? The VERY few heater failures I have seen related to huge series-string TV sets, and gear that was heaved into a dumpster before I grabbed it.

Thanks, PRR.  Just thinking how expensive tubes are getting, mostly... I'd like to magically make them last forever

So you don't think start-up stresses are that big an issue?

Tod

No, it's not a big issue.  I truly believe, as usual, most people want to dick around with an overly-complex solution to a nearly non-existent problem, because it's 'possible'.  I shall continue to fail to understand how a 3 terminal regulator is a better solution than a variable power resistor.  I certainly expect the regulator to be the failure prone device.  I have regularly turned on more than 50 tubes a day for more than 10 years, and have never, in more than 20 years of tube gear experience, seen a filament fail on start-up.  In that same time, I have had to repair various DC filament schemes that failed; diodes, caps, regulators.  I have never had significant problems that required use of DC filaments, other than the Gates SA-39 limiter.  I have multiple 90 dB tube preamps with AC filaments that have no obvious hum problems.  Heck, not even DC biased AC filaments.  It all comes down to layout, and anything else is a band-aid IMHO.  As PRR has stated elsewhere, most tubes will take an enormous overvoltage/overcurrent situation for a reasonable amount of time without giving up the ghost.    The 'fix' for certain noisy tubes is to over-voltage the filaments so they repurpose stray electrons in the vacuum.  If a tube can take this for extended periods, it can certainly take the 20-30 seconds of unregulated AC warm-up.    The only place I'd even think of worrying about this would be with really expensive unobtainium tubes (300Bs, 348A's, 45's, 1603's, etc), which one shouldn't bother with in the first place unless wealthy and willing to throw bales of cash into the fire for the sheer luxury of it all. 

> how expensive tubes are getting

I recall what tubes cost in my youth, also I was just looking at a 1938 price list.

As far as I can tell, tube prices have kept-up with the price of beer.

> magically make them last forever

Making a beer last forever might be nice....

NEW-Tube prices seem to have risen a little faster than the price of gasoline or beef, a LOT faster than the price of resistors (which dropped 1970-2000), a lot slower than the price of cigarettes, a little slower than the price of cars.

On average, new-tube prices relative to your other expenses and vices have not changed much in decades.

> magically make them last forever

Heater failure is usually not why we discard a tube. Yes, I have heard of brand-new tubes apparently suffering heater failure in the first hour; these were just made bad, nothing you can do. We discard 12AX7 because we throw the amp in/out of the Econoline so hard the little bits loosen up and rattle microphonically. We discard 6L6 because we run 500V to get 50W out of a pair of 360V tubes good for 36W (or, again, because they were made bad). This actually makes sense to professional musicians: maximum racket in the least weight and up-front cost is worth regular re-tubing. Stage-amp tube-costs come to around 10 cents an hour, less than the cost of strings, picks, and shirt-laundry. Less-over-stressed tube applications will have even lower costs, and hopefully higher hourly value.

> really expensive unobtainium tubes (300Bs, 348A's, 45's, 1603's, etc)

Right. If you "must" use an HK257 (large sweet power pentode) in your Champ, you have to fight for the very few ever made 50+ years ago. Like buying a Packard, except you can re-ring and paint-up a Packard, we don't rebuild small bottles.

Hi Doug,
Can you explain more about this :
" I have never had significant problems that required use of DC filaments, other than the Gates SA-39 limiter "
What is so special with this unit ?
Cheers,
Guy

The GR tubes tend to hum like a motherf$*#$r if there's any imbalance whatsoever.  The last versions of the SA-39B implemented DC filaments for the GR tubes only, as a fix. 

Thanks, Doug, for the answer about the Gates.

Also, I have read that using DC for heaters shorten the life of tubes.
Do you know if this is true ( and why ? ), or just a legend ?
Best,
Guy

Don't know.  I think the whole problem is grossly exaggerated.    Other than guitar amps, which get moved a lot and thrown around, I have seen so few tubes fail that I cannot say I've seen enough evidence to say.  Of all the pieces I've restored for myself and for others over the last decade, I've rarely seen dead filaments.  If tubes are bad, it's because they are at the end of their emission life (decades of use), or noisy.  Tubes that light up fine, but virtually don't pass signal anymore, or rectifiers that can no longer produce DC.  But glowing just fine.  I've replaced more truly bad transistors than I have truly bad tubes.....and I work on more tube gear than transistor.    I don't know what it all means, other than that there are a ton of unsubstantiated rumors and old wives' tales about tubes to be digested, if one wishes to do so.