 |
|
Techniques To Maximize Power Tube Life
By David C. Gillespie
Republished (with permission from AudioXpress), here is the original (and only) article that directly addresses the causes contributing to reduced life expectancy and arc events in audio output tubes, and what can be done about it.
|
|
|
|
|
Originally published 7 years ago, this article was the first to identify the necessary solutions to maximize power tube life in vacuum tube audio amplifiers. Since then, the solutions presented have been publicly validated by countless diyers seeking to resolve these issues in their amplifiers. The article is written for all levels of diyers, and is applicable to new and old equipment alike.
|
|
|
|
|
|
With screen stability resistors, high gm tubes maintain stability, in spite of negative factors (discussed in the article).
|
|
|
|
Reader Comments
| Posted by
Dave |
January 18, 2016 - 07:23 am | |
Hi Mark -- Depending on the unit, you're absolutely correct. The voltage rating of all relevant components should always be considered wherever modifications are being contemplated. Thank-you for mentioning that point. I did not get that specific in the article as it was offered as more of a concept piece than specific how to instructions for every application. But the point is worth drawing special attention to, so thanks again for bringing it up!
Dave |
| Posted by
Mark |
January 14, 2016 - 07:09 pm | |
| When using the two relay solution that you show in this article. Isn't there a concern about the B+ exceeding the voltage of the Power supply capacitors when the relays are open during start up? |
| Posted by
Dave |
December 05, 2015 - 07:51 am | |
Hi Peter -- From a pure theory standpoint, carbon composition resistors get recommended for stopper service because they do not have the inductance associated with them that film type resistors can have. From a practical standpoint however, the effects of any such inductance is a moot point at the frequencies that vacuum tube audio amplifiers operate at. Still, in theory, they are the best type to use in this application, so for control grid stoppers, they would get the nod.
For Screen Stability service however, the very real and practical qualities of metal film resistors make them the clear choice in this position. With the high voltages that a screen grid operates at, it always invites the possibility of collateral damage should a fault occur. In this application then, a carbon composition resistor needs time to heat up, overheat, and finally burn up to potentially open the circuit. However, a metal film resistor will open rather instantly by comparison, announcing the fault with a distinct "pop", and clear the overload much quicker. This quality gives metal film resistors a definite advantage for Screen Stability applications, as well as in all power supply positions in general.
Thanks for the interest -- I hope this helps!
Dave |
| Posted by
Peter Capobianco |
December 04, 2015 - 08:40 pm | |
Hi Dave,
It has been my understanding that carbon comp resistors are desirable for grid connections, like grid stop resistors. Do the screen stability resistors in your article need to be carbon comp?
Thanks.Peter |
| Posted by
Bill |
August 26, 2015 - 12:59 pm | |
I'm a new owner of a Ming Da MC-5S. What with 10 power tubes, you can imagine I'm most keen on life cycle management. I found the Audio Express article most interesting.
Tubes seem to fail in a number of ways. Heaters burn out, if only rarely. Gas accumulates. Arcs happen. Physical deformation/breakage within the assembly. And, finally, the cathode emission simply degrades over time.
It seems larger power tubes are more prone to these failure modes, than the smaller signal tubes. A signal tube might last 10K hours, but the author was only able to achieve a 2600 hour average on the power tubes (vs something in the 300 hour range).
A massive and apparently statistically valid improvement, to be sure. Sadly, It took years to demonstrate and I suspect there may be other routes to a premature failure as well.
Opinions on the net are all over the place. For example, "slow start is pointless because cathode stripping doesn't happen." Ok, but, perhaps the slow startup allows the assembly to heat more evenly and thus reduces mechanical stress due to different expansion rates within the assembly.
Another is "standby is bad (for all sorts of various reasons)". But, perhaps, keeping just enough heat on the getters helps reduce gas accumulation. It may well be bad, for any of the assigned reasons, but which "bad" is worse?
Some say run the tube within spec, and you'll be fine. Others, including some manufacturers, say that life can be extended by derating the tube (close to 70-80% of specified max power, for example). But, how low is low? And, of course, there will likely be sonic implications too, but how bad is "bad"?
Of course, the answers are always along the lines of give it a try, YMMV, and good luck with that. Sadly, the author's investigation took years to do and a required a certain level of precision. We really should thank him for that. I, for one, am factoring his findings into my units. Its very cheap insurance.
I wish we had similar studies on some of the other failure mode possibilities (and potential trade offs), even if not scientifically flawless but simply well considered and documented.
Alas, 10, or so, years is a long time. |
| Posted by
Dave |
April 08, 2015 - 03:25 pm | |
Hi Dak -- The operating conditions of Class A single ended cathode biased designs are such that they do not promote output tube arcing. Therefore, you can use whatever approach you wish to use to achieve the appropriate drop.
Good luck with your project!
Dave |
Add your comments here...
|
|
