Pre amp or power tube flash?
22 responses Add your response
Try lightly tapping the tube with a pencil eraser while it is on the tester, to make sure the tube doesn't have a marginal short circuit.
Also, it's possible that something else failed in the amp that caused the tube to flash. For instance if a coupling capacitor connected to the grid broke down, it might have put a large positive voltage on the grid, causing the tube to flash.
I think I see where Stevecham is going with his question - Amperex (Holland) small signal tubes will 'flash' when first powered up, and it's totally normal with those tubes.
On the other hand, I've had power output tubes 'flash' on me, blow a fuse, and test fine afterwards (all EL34's for what it's worth) but I've learned to think of my tube tester as a rough way to weed out obviously failed tubes, not a final arbiter of good tube vs. bad tube. A lot of my tubes have tested well, but were noisy/fuzzy or microphonic or had less output than a tube used in the opposite channel that tested exactly the same.
Have you seen those old Upscale Audio ads that say "we don't use your amp as a tube tester!"? Don't use yours as one! If an output tube flashed, don't risk putting it back in your amp, regardless of what your tester is telling you.
Just my opinion, of course. Good luck!
don't use your amp as a tube tester?! how COULD you?
Well lets see... the tubed flashed in the equipment. The tube was pulled and tested ok in the tester.
Put the tube back in the equipment, maybe flashes again but this time damages an electrical component or trace on a circuit board.
Yep maybe the tube was bad after all..... Hell of a way to find out, use the amp for a definitive test.
Is it possible for a tube to flash, blow the fuse in the amp, and then test ok on a tube tester? Or is my tube tester broken?
May not be the tube that caused the fuse to blow.... Might be though....
Thanks everyone very much for your comments. It was a power tube, coincidentally an EL34, as mentioned by Steveaustin. I did try the tapping test and saw no flicker in the testing lamp, and no, I did not put the tube back in the Amp after I tested it. Ive replaced the tube and fuse and have 10-15 hours on the amp since replacing. So it does not appear to be a problem inside the amp, at least a problem that manifesteds itself right away.
Just to take it to the next level though, this is the fourth of eight EL34 Winged C tubes that has blown in the amps (theyre monoblocks) in the last two years. The tubes started blowing within the first 200 hours of usage. The tubes still have less than 1000 hours of service and all test in the good range on my tester. The tubes were purchased from a very reputable dealer. The tubes were all matched. The amps are Conrad Johnson Premier 12s. The tubes have blown in both amps. Ive taken the amps back to Conrad Johnson twice for testing (I live in the area). Each time they say no problems with the amps, just a bad tube.
Anyone have any ideas?
I'm not a CJ expert (in fact I've never owned one) but as I recall the CJ Premier 12 was originally designed around a 6550 tube, not an EL34. It also has a manual bias system where you turn a pot until a LED turns on. If I'm correct so far read on....
Amps of this type come with a bias arrangement that anticipates the use of certain tubes which would exclude others. For example many, if not most, will bias 6550's at around 50ma. But smaller tubes such as an EL34, E34L or 6l6 and 6l6 GC performs optimally at a much lower number, ranging from 25ma to 35ma (for example the latter for the E34L and the lower numbers for EL34s which would be around 30ma). This is general and amps/tubes will be different considering amps design and the extent that you select the bias and are willing to lower or increase bias to get the best tone, etc.
If your amp has a bias circuit which assumes 6550 type power tubes, as well as KT88s and KT90's, then you would be biasing your EL34's far to hot. This would cause early death, shorting, etc.
I do not know how you could circumvent this unless you could manually bias the tubes using a meter with real numbers or by getting CJ (or a good techie) to alter the bias arrangement so that when properly biased according to the LED your bias will match the requirements of smaller power tubes.
I hope that helps you a bit. I may be way off base.....:-)
BTW when you read Jim McShane' post of AA (posted above by Jea48) be sure to read it in context, not just superficially by looking at the words 'bad' and the yellow highlighting. McShane has always spoken highly of SED EL34's and that has not changed, although he recently indicated he had recieved some negative comments on NEW SED 6550's which suggests a new quality control propblem.
A tube tester won't apply the full B+ voltage testing a tube. Even the top testers won't if the user doesn't know all the operating parameters of your amp.A lot of newer amps drive tubes a lot harder then they used to back in the tube only days. Vintage amps also drive tubes harder now do to the higher line(outlet)voltage than we had years back.That might explain why good NOS tubes may be failing sooner than they should be.If a tube does flash in a amp,I wouldn't want to reuse it even if it wasn't at fault,assuming a part in the amp failed causing the tube to flash.
Thanks Newbee for pointing out about the yellow highlighting of the word "bad".... The yellow highlight is the nature of the beast apparently in the AA program when doing a search.
I entered "Bad EL34 Winged C tubes" so the word bad will be highlighted in all found posts.
And yes as you pointed out Jim does not say all EL34 Winged C tubes are bad.
As Hifi mentioned: most testers do not apply full operating voltages(B+) to the tube being tested. Tubes, under actual operating conditions, will occasionally either short, draw high screen current or open a connection internally. In the event of a short or high current draw; your circuitry may suffer some consequences. The only way to minimize the odds of that occuring is to burn in the tubes at full voltage and cull the weak ones prior to their use, as is Upscale's practice with their power tubes, ie:(http://www.upscaleaudio.com/rare/tpm.htm).
Rodman, funny you should mention Upscale because thats where these tubes were purchased. I guess its the luck of the draw, and in my case, my draw was aces over sevens.
And Newbee, you have an interesting premise. I should have mentioned though that my amps are the Premier 12 XS, which are configured to use EL34s. So I assume the bias circuit is designed around EL34s, but Im going to ask Conrad Johnson that exact question. Its also interesting that the last time I was at Conrad Johnson, I bumped into Lew Johnson and asked him what he though might be going on. His first question was what type of speakers are you using. When I told him Quads (63s), he commented that they (Conrad Johnson) had some previous problems with customers that used Quads and their amps. He said that the Quads, under certain circumstances might draw too much current. This all sounded a little strange to me (I was always told Quads and tubes were a good match) but Im not an electrical engineer and he was talking a little over my head, so I may have misunderstood. Since then Ive taken the Quads out of service, and Im using dynamic speakers. But the problem has continued.
Thanks again for everyones comments. Any other thoughts would be very much appreciated
Jaffeassc, "so I assume...." You know what they say about what happens when you assume........:-)
Next time you talk to CJ ask them what the bias setting programed into the amp actually is. After they tell you ask them what the bias setting is in the 6550 model. Then find out if they actually checked the bias in your amps with your tubes in it as opposed to just noting the the tubes and the little LED were working.
BTW, some tube amp are designed to push their tubes very hard to meet high power spec's. Yours may just be one of those and underly the reason for CJ's statement about it having problems with a speaker that required more current, current their amp couldn't deliver.
BTW I drove Quad 63's with SED EL34 in a SFS 80 amp which only put out 40wts with this tube. I set the bias at 30ma and drove them off a 3.5ohm tap (internally configured). I never had a problem with tube reliability or tube life and I drove them hard enuf that they 'rattled' a tiny bit.
CJ's comments about the 63's could be accurate about their needs but I really don't under why a high quality mono amp couldn't drive them. As I recall they have a minimum impedance of about 3.5 ohms in the bass - any respectable tube amp should be able to handle that, especially a Quad which you can't drive that hard anyway. Since you have the same problem with dynamic speakers I suspect the CJ's may have a problem. If you have a techie nearby you might ask him to check and see what the bias actually is.
But, as I said in my first post I'm not a CJ authority.
I've had tubes flash, cause damage, and not only measure perfectly fine, but go on and lead a typical, long, quiet suburban life.
Culling tubes that test poorly in their infancy is certainly wise practice. While tubes that get through boot camp obviously will lend a higher degree of confidence, let's not be lulled into a false sense of security. Tubes do hiccup and fail, and sometimes collateral damage results. The old investment adage "past performance is no indication of future results" rings true with vacuum tubes as well.
"How old was the tube that failed? Have you contacted Kevin in regards to the tube's failure (and possible warranty)? Did the same thing happen more than once? You mentioned, "the problem has continued
No I havent contacted Upscale the first tube that failed was probably 4 6 months old. As I mentioned before, this is the fourth of eight tubes that has failed. I should have taken notes, so I really dont remember when the tubes failed, which positions failed, etc. I do know it has happened in both amps.
"(why I ask)".
Why the problem continues is what I am wondering as well.
"Next time you talk to CJ ask them what the bias setting programmed into the amp actually is. After they tell you ask them what the bias setting is in the 6550 model. Then find out if they actually checked the bias in your amps with your tubes in it as opposed to just noting the tubes and the little LED were working."
Now that I know a little more specifically what to ask, I will. Thanks for the advice. I do know that they checked the bias in the amps with my tubes the last time they were in for service. Thats how they found the problem (the amp was blowing fuses and my tester was not detecting a bad tube).
"If you have a techie nearby you might ask him to check and see what the bias actually is."
Actually, Bill Thalmann, whom I believe may have designed the Premier 12s when he worked for CJ, also has his shop in my area. So I may take it to him the next time it happens.
One of my first jobs was at a electronics tube,and some other supplies such as tv antennas,picture tubes,etc. dealer.In my grade schools years,I tested tubes for service techs,and the public.They had about 5 testers,and some were better than servicemen had.While working there,I tested literally in the thousands,maybe tens of thousands of tubes.I've found the opposite of what they state to be the case.Yes,an amp that is not properly designed,or rapid cycling of the power switch can cause problems.But as a general rule,the failure rate back then was highly in favor of the tube failing more often than the amp causing the problem.An old out of spec amp is a different story.I still find tube failure to be more common than an amp problem,decades later.Here is the link to their statement.>>[http://www.emissionlabs.com/html/articles/white-spark.htm]