• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tube flashing

Can someone please enlighten me on the following.

I have a preamp that uses 2 X 12AT7 and 2 X 12AX7. I just purchased some used Telefunken 12AX7 and when I switch on the preamp, none of the the tube flashes. Previously, when I was having Mullards 12AX7 in place, one each of the 12AX7 and the 12AT7 flashes. Why does this happen and does the flashing indicate the condition or quality of a tube?

Thanks.
 

ashok

Member
2002-06-06 4:43 am
3RS
I have several tubes that flash and other brands that don't.
I think the ones that don't flash were made that way. So a heater that flashes does so because of inrush current and localised heating. Once the coil heats up the resistance increases and the current falls and heating becomes uniform over the heater wire.

I would guess that heaters that flash will benefit from a soft start heater voltage , for longer life. The ones that don't flash probably are already tougher regarding this kind of faliure .
Cheers.
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
> soft start heater voltage , for longer life.

In theory. But in practice, not counting series-string TV sets, I have only ever seen two broken heaters.

One I put on a Variac and turned to double rated voltage. It glowed real bright, but did not burn out. I slid it up and up, and it burned-out at four times rated voltage in a few seconds.

One was in lab gear that I knew had been physically abused (thrown into and off of a truck). It acted like a mechanical break or broken weld, not a burn-out: it worked a few seconds when cold, but would not stay hot.

In light bulbs, and in high-voltage tubes that can't use oxide emitters, you do have to run the heater close to burn-out. But common oxide cathodes can run 2/3rd, even 1/2 that temperature. Filament life falls roughly as the 13th power of temperature. So if a light bulb can live 1,000 hours, a filament at 2/3rd that temperature can run 1/(0.66^13) times longer, or 200,000 hours. Few tubes were factory rated for even 10,000 hours, so filament failure is not limiting tube rated life, even if you occasionally get a little hotter for a few seconds. I have tubes in low-stress duty coming up on 100,000 hours and working fine, and I don't pamper their heaters.

OTOH, a slow-start scheme that fails may put 10V or 15V across a "6V" heater. Even this is not instant disaster, but you don't want that to happen for long.

I say enjoy the light.
 
Flash

Look carefully into the tube. On most 12A*7 type tubes there is a steel link that connects the filament of the two sections together. On some tubes the link is made out of filament material. It is this link that flashes on power up. I have also seen this on other dual tubes like 6U8's. I have also seen the link run across the top of the tube. There are some tubes where there is a considerable amount of filament material exposed outside the cathode. Since the cathode is cold on power up and acts like a small heat sink the exposed material heats up much faster and glows pretty bright during power up. This is a normal phenomenon for these tubes. If I remember correctly, I have only seen this flash in European tubes, Amperex I believe.
 
Filament life falls roughly as the 13th power of temperature. So if a light bulb can live 1,000 hours, a filament at 2/3rd that temperature can run 1/(0.66^13) times longer, or 200,000 hours.

As an aside, I remember taking a tour of Edison's winter laboratory in Ft. Myers, Florida, a few years ago. The lab was partially illuminated by dimly lit, orange-glowing antique light bulbs that, we were told, were from Edison’s day - still working all these decades later. I never heard the 13th power rule before, but that's a huge increase in life for a small decrease in temperature. That would explain Edison's bulbs.

Tubes in one location in my old ARC SP3A1 used to flash with terrifying brilliance, especially with some tubes. While I never suffered a failure due to this, it bothered me that the heater was undergoing such thermal shock and I worried about fatigue due to mechanical expansion-contraction cycles. It turns out that ARC combined heaters in a series-parallel arrangement, but lacking the needed even number of tubes, inserted resistors (whose values matched the fully hot heater resistance of a 12AX7). At turn-on, there was an imbalance in the series-parallel arrangement as the tubes started cold, but the resistor had a “hot” resistance. I ended up replacing the resistor with just the heater part of an older 12AX7, and tie-wrapped this dummy tube to a filter cap. No more flashing.

More recently, I have used voltage regulation with LM317s or equivalents, but I’ve implemented the simple slow turn-on circuit with an extra transistor, resistor, cap and diode as shown in the app notes. Seems to work very well. I too have never experienced a failed heater in audio amps, but I would prefer to be cautious.
 
The "flasher", captured

The "flash" has no bearing on the quality of the tube. I looked through my tubes and found this "flasher". I took several pictures attempting to time the camera and the power supply simultaneously. This tube is a Zenith 6GH8A, Made in USA. I don't think that anyone would accuse this of being a "quality tube". The flashing is due to the filament construction. The fact that the link between the sections is made from filament material instead of steel is probably a cost reduction move, not a quality move.

I have seen other tubes that "flash" and some were from well known European manufacturers. I don't believe that they are any better or worse than a non "flasher" from the same manufacturer.
 

Attachments

  • flash.jpg
    flash.jpg
    54.7 KB · Views: 505
I have a Mullard 12AX7 currently in use in a preamp that used to flash when it was first installed. The preamp provides it's heater power via a conventional silicon diode based rectifier (which of course applies it's power, pretty well stright away). After a few weeks of normal use -- it now doesn't flash anymore on startup. (Though still seems to work {i.e. sound} fine).

Now the interesting thing is that this tube was originally 'pulled' from tubed real-to-real tape recorder, which used tube rectification. Looking at it, it had appeared to have been well used. Yet still flashed after all these years when first installed in my preamp. But after just a few weeks of being fed from a non soft-start circuit, it looses its 'flash'.

Whether is has any sonicaly deleterous effects or ramifications for tube-life, I'm not sure. But whilst outright heater failures might be rare, there clearly seems from the above example to be *something* changing/wearing out here. Hence I now have some niggly concern (admitedly minor) about using non- soft-start cuircuits in any of my future designs.

Just thought I'd share to provoke any further thoughts.
 
In modern day language.....a flare up heater on pwr-up indicates NOS OR early manuf, i.e Darned good tubes......Some still work after 40 years) in my amps.
In saying this.......Don 't use toroid mains trannies for these tubes without soft start or inrush limit. The earlier generation of mains trannies were much softer on start-up for these tubes. i.e Bmax under 0.9 Tesla.


richj
 

fdegrove

diyAudio Senior Member
2002-08-21 1:20 am
Belgium
Hi,

The "flash" has no bearing on the quality of the tube.

As long as it's the heater flashing then, no.

The only reason you see this is because the ends of the heater is sticking out of the cathode sleeve and most likely isn't of the "controlled heater"warm up time type. (Usually comes with a A suffix after the reference, e.g. 12BH7A iso 12BH7)

Controlled heater types should carry an extended lifespan compared to their older brethren but other than the fact that the heater wires were coated at the ends to conceal the flaring, this was done mostly to reassure customers that there wasn't anything wrong with their units.

IOW, besides occasional manufacturing flaws, nothing wrong with a flashing heater at start up.

I'm thinking that maybe a thermistor after the rectifier could be a good idea.

When using silicon diodes this is always a good idea; protects tubes and filter caps from high inrush current.
On preamps you can even put them before the rects. if you like.
For heater circuits I wouldn't bother, they're slow to warm up as it is already.

Anyone tried this and care to report?

I use them whenever I can. Religiously so.

Ciao, ;)
 
Hi there..........Yes ...Frank I forgot about the A suffix after tube desig.......
Re thermistors...I've given up using them for inrush limiting on either B+ or AC simply because my supply is so rough where I am ....that I've lost many devices as the time-constant is too long before a surge reappears...then Zap....it happens so often that I have to design using S.S but If one lives in an electrically clean street environment, thermistors will work well.

Just a perspective where they don't work!

richj
 
fdegrove said:
For heater circuits I wouldn't bother, they're slow to warm up as it is already.
Well as per my concern on the previous page, it seems that something is changing/wearing on the tube heater side of things when being run in a non- soft-start silicon circuit. So hence I'm thinking that thermisting the heater side of things might be still worthy of consideration.


richwalters said:
Re thermistors...I've given up using them for inrush limiting on either B+ or AC simply because my supply is so rough where I am ....that I've lost many devices as the time-constant is too long before a surge reappears...then Zap....it happens so often that I have to design using S.S
:eek: gee Rich, that doesn't sound good at all!

I'm guessing that some kind of spike-protection beforehand hasn't helped you either so far?
 
I should have pointed out that my problems with thermistors arise because alot of later designed tube amps often use the hard low Z cap-input filtering using s.s rects. Quite a few late amp designs did away with the choke..as electrolytics caps got better.
The B+ choke input filtering is far better and softer using indirectly heater rectifiers with better power factor ....Little wonder why all the changes.

The orig GEC KT88-50 watt amp from 1950's uses an CZ6 type thermistor in the 500V B+ to avoid the surge from the 5U4...(a GZ34 would have been better bet but wasn't off the drawing board) .Since this amp works close to class A, I know that many users have scrapped the B+ choke (to save weight) and used a 355V tranny with a bridge feeding alot of uF and keeping the thermistor in the B+......this created more problems that if all was left originally. In this design 680uF on the B+ was required to keep 100Hz ripple within 6V p-p.....

So it stands...the modified circuit blew the thermistor more than the line fuse! So I use a power resistor/ switched out with relay.


alas richj