Preventing "Mullard flash"

Guest · 1049

0 Members and 1 Guest are viewing this topic.

Deke609

  • Guest
on: January 25, 2020, 07:35:34 AM
Someone posted a link to the following video on AA and I thought it might be of interest to people here who use Mullard or Telefunken driver tubes that flash on start-up:
The videomaker suggests that, based on his experience, filament flash leads to premature failure.  He shares a couple of designs for slow heater warm-up that solve the problem.

A couple of people on AA have suggested that filament flash is benign. I don't know enough to have an informed opinion, but the theory that flash could lead to premature failure makes some sense to me: if only a certain segment of the filament gets momentarily white hot on start-up, I would expect in the long run that segment to fail before others. That said, I have no idea whether other parts of the tube would be expected to wear out long before a flashing filament failed. 

cheers, Derek

[edit: corrected link so that the video starts at the beginning]
« Last Edit: January 25, 2020, 07:40:18 AM by Deke609 »



Offline Tom-s

  • Sr. Member
  • ****
    • Posts: 492
Reply #1 on: January 25, 2020, 11:45:33 AM
I use European small signal tubes 99% of the time.
Philips/Mullard/Valvo/Siemens/Dario/Pope and many more Philips factory tubes, they flash.
The Telefunken i’ve used (and many more tested) generally don’t flash on power on (funny how the guy in the vid mentions those as tubes that flash).
GEC and Marconi labeled  made in England, they flash. Whereas Marconi/Ates/Fivre from Italy generally don’t.
I still have to experience my first premature tube failure as a result of this.
Even special “longlife” variety of these small tubes have the flash right out of the box(“NOS/NIB”).
Philips at some point in time was the worlds biggest manufacturer of vacuum tubes.
They must have known what they were doing. And (from what I‘ve read online) there’s no general consensus that the Philips/Mullard/GEC tubes are worse than American tubes due to short service life.
I’ve also discussed this question before with a service technician that worked for Philips in Eindhoven for 45+ years and was a ham in his own time (and repaired radios for a local radio museum). He felt there was nothing to worry about with this flash in small signal tubes.

Big power tubes may well be a completely different story.

Edit: what I’m trying to say is that I feel the vid is offering a solution for something that’s not really a problem.
« Last Edit: January 25, 2020, 12:47:04 PM by Tom-s »



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #2 on: January 25, 2020, 03:31:25 PM
I have some issues with what's presented in the video.  Most importantly, the tube he's messing with has a manufacturing issue (as he mentions).  One side flashes very brightly (not the Mullard flash that I'm used to seeing) and the other side doesn't.  The other immediate issue is that he speaks in the beginning of "large tubes" having filament failure.  I would presume these were something closer to a thoriated tungsten transmitting triode perhaps, which is a lot different than a 12AU7. He then moves to the 5751, but it doesn't have a heater flash...


Next on the solution, so you put a 27 ohm resistor in series with the heaters that you can short with a switch.  If you forget to throw that switch, you're either putting only 2V into the filament half (just using his numbers for half a 12AU7).  If you leave a Crack like that with the power on, the 12AU7 won't draw the plate voltage down and you will risk damaging the rest of the amp.  You could also destroy the cathodes in the 12AU7 in the process because they are not adequately heated. 

My last issue is that I have piles of used Mullards that flash, but I can't ever remember throwing one away that had an open heater.  This is just anecdotal of course. 


Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Deke609

  • Guest
Reply #3 on: January 25, 2020, 07:18:28 PM
Many thanks Tom and PB - this is good information and news.  So I'll chalk it up as "not a real world issue".  I do like, though, his simple auto-delay schematic using a fet. I'm sure it's quire common, but it was new to me.  Looks like a handy circuit.

cheers and thanks, Derek