12AX7EH CF Failure

[GainFreak]

Hi,

I don't know whether I hit a bad batch of these or something but recently I had 4 12AX7EH fail in CF position /driving tone stack/. One of them worked several hours, one worked for a month, couple of them failed shortly after turning the HV on.
Anode voltages were 350-390V, Voltage drop at 100k cathode resistor before failure was 160-190V. Tubes were manufactured between 01.01.07 and 01.01.08.
I read somewhere in a forum that other people were experiencing similar problems with 12AX7EH when used as CF. I have no problems when using them as regular triode gain stage. Actually they sound very good and have plenty of gain.
I was wondering if anyone here had similar problems?

[Enzo]

Look up the spec sheet on the tube and see what the max allowable heater to cathode voltage might be. I be suspicious you were exceeding it.

Try elevating your heaters.

[GainFreak]

Look up the spec sheet on the tube and see what the max allowable heater to cathode voltage might be. I be suspicious you were exceeding it.

Try elevating your heaters.

I'm using 12AX7EH from a long time but I started having these issues only recently. When CF tube is replaced with other brands they work OK. Since it's a 12AX7 shouldn't they all have the same ratings?
I'm running regulated DC heaters feeding opamps and other stuff so the elevation of heaters is not an option here.

[Wakculloch]

Is the 12AX7EH of russian origin, cos if so they may be the same as 12AX7WB, which look to have the same structure as the 6N2P, but with different heater wiring.
This could be important as although the 12AX7 has a max h-k of 180V, the 6N2P is only 100V.
Hopefully someone will know if this is just supposition.

[Enzo]

Since it's a 12AX7 shouldn't they all have the same ratings?

NO.

A 12AX7 should have a certain amount of gain, and certain operating characteristics, but that doesn;t mean ALL criteria are the same tube to tube.

[R.G.]

Originally Posted by GainFreak

I'm using 12AX7EH from a long time but I started having these issues only recently. When CF tube is replaced with other brands they work OK. Since it's a 12AX7 shouldn't they all have the same ratings?
I'm running regulated DC heaters feeding opamps and other stuff so the elevation of heaters is not an option here.

Hmmm. Lessee here.

A guy walks into an amp tech's (doctor's) shop (office) and says, "Hey, tech (Doc) it burns this tube out (hurts) when I put this kind brand of tube in (do this.) The tech (doctor) looks up and says "Well, don't do that."

[soundmasterg]

The Sovtek 12AX7LPS is known to fail in the CF position, so it doesn't surprise me that the EH 12AX7 does it too. I'd use different tubes in that spot myself, especially after 4 of them fail for you. Save the EH for different locations in the amp and they'll work fine.

Greg

[GainFreak]

The Sovtek 12AX7LPS is known to fail in the CF position, so it doesn't surprise me that the EH 12AX7 does it too. I'd use different tubes in that spot myself, especially after 4 of them fail for you. Save the EH for different locations in the amp and they'll work fine.

This problem with EH has been reported in other forums as well. In a russian forum for example somebody claims to have relatives working in the factory and they told him they're having technological issues.
Also the comments there come from people using these tubes in hundreds so we're not talking about couple of accidents only.
My tubes in question come from a group buy of 100 pcs 12AX7EH directly from New Sensor. I think if we have a 4% failure so far this looks like a problem to me .
Now I'm using 12AX7EH in non-CF positions and they sound very well with plenty of gain. I'd hate to move to other brand/s but if this issue is not solved I'll have to.

[Steve Conner]

The CF position in guitar amps probably runs the heater/cathode insulation beyond any spec that any 12AX7 ever had. Even if it doesn't exceed the spec on DC conditions, it gets violated hugely on signal peaks. I'm more surprised by the fact that other tubes withstand it, than by the troubles with these particular tubes.

Heater/cathode insulation was always a black art of the tube era. It's no easy matter to find a substance that keeps its electrical insulating properties year after year under orange heat and a high DC bias. Most glasses and ceramics that I know of start to conduct once they get red hot, and once this happens the DC drives ions through them and destroys them.

[GainFreak]

The CF position in guitar amps probably runs the heater/cathode insulation beyond any spec that any 12AX7 ever had. Even if it doesn't exceed the spec on DC conditions, it gets violated hugely on signal peaks. I'm more surprised by the fact that other tubes withstand it, than by the troubles with these particular tubes.

Heater/cathode insulation was always a black art of the tube era. It's no easy matter to find a substance that keeps its electrical insulating properties year after year under orange heat and a high DC bias. Most glasses and ceramics that I know of start to conduct once they get red hot, and once this happens the DC drives ions through them and destroys them.

I'm not familiar with tube technology but the fact is most of 12AX7s can withstand the conditions in question. I don't mind replacing them from time to time but not every month or so /or every couple of minutes.../.

[soundmasterg]

What I was trying to tell you is that the Sovtek 12AX7LPS and 12AX7LP does not work in that position in amps. Every vendor I know of tells you not to use those tubes in that position. Given that EH is made at the same factory, I would suggest that it is certainly feasible that it will have the same issue when used in that position. All of these tubes work great in other positions in the amp. So use the EH somewhere else and use some other brand in the CF position such as JJ as they work fine there. Problem solved.

Greg

[Enzo]

Yeah, I don't see the mystery either. This particular tube doesn't like being used as a CF, it is fine elsewhere. Most other tubes are OK with it. So don't use this brand 12AX7 as a CF, use one of the most other brands there.

[GainFreak]

Every vendor I know of tells you not to use those tubes in that position.

I didn't know that. Certainly no such statement on New Sensor's site.
Thanks to all for the feedback.

[soundmasterg]

New Sensor doesn't say anything on their site except advertising their products. You also have a hard time getting accurate info and specs on some of their products if you try to get it from them. The non-use of the EH 12AX7, Sovtek 12AX7LPS and Sovtek 12AX&LP as a CF has been all over the amp forums for at least 2 years. Now you know I guess...

Greg

[Diablo]

What's "CF" mean? Center-fold? Center-field? Crummy-Fender? What?

[Daver]

CF=cathode follower.

[Steve A.]

Originally Posted by soundmasterg

The Sovtek 12AX7LPS is known to fail in the CF position, so it doesn't surprise me that the EH 12AX7 does it too. I'd use different tubes in that spot myself, especially after 4 of them fail for you. Save the EH for different locations in the amp and they'll work fine.

Greg

Ah, c'mon- let's sacrifice 4 more tubes to the Vacuum Gods!

Thanks for the heads up on this! Perhaps we can start a database here on problems and quirks of specific tubes...

Steve Ahola

[GainFreak]

I asked two European online shops about this issue. One of them confirmed about 12AX7EH adding to the list Tung-Sol, Sovtek LPS and LP. The other wasn't aware of the problem. Then I asked a major distributor and they also confirmed the 12AX7EH issue adding to the list any other spiral filament type tubes like 12AX7LPS, JJ ECC83s, Tung-Sol 12AX7, and Mullard 12AX7.

[myfoot]

This was posted over to the Metroamp forum. Sounds like it works fine anywhere but the cf stage

http://forum.metroamp.com/viewtopic.php?f=13&t=20150

[GCA]

Originally Posted by GainFreak

I asked two European online shops about this issue. One of them confirmed about 12AX7EH adding to the list Tung-Sol, Sovtek LPS and LP. The other wasn't aware of the problem. Then I asked a major distributor and they also confirmed the 12AX7EH issue adding to the list any other spiral filament type tubes like 12AX7LPS, JJ ECC83s, Tung-Sol 12AX7, and Mullard 12AX7.

I'm not seeing that same problem with the JJs. The 12AX7EH, though... I'm wondering if there's an issue with that particular heater design.. poor insulation (material, thickness, etc). The one EH I have doesn't glow like the rest of similar construction, leading me to believe I may have received a slightly defective tube. Not all CF designs elevate the heater, so I'm wondering if the "less than happy heater" that I'm seeing is related to the CF death of those tubes due to insulation failure. Morgan Jones mentioned this on page 266 of Valve Amplifiers (3rd edition):

"Increasing voltage across the heater/cathode insulation increases leakage currents. Although Vh-k(max.) is specified on data sheets as being anywhere from 90V to 150V (except in some ruggedized and 'P' series TV valves), this is a very 'soft' limit, since it is usually given at an arbitrary leakage current, nevertheless, a suffieciently high voltage will [his emphasis, not mine] punch through the insulation to rupture the heater. Heater failure due to heather/cathode insulation breakdown is uncommon, but is most likely in cathode followers with high signal voltages or output stages with distributed loads (such as the McIntosh design)."

"High signal voltages" sounds a lot like guitar / bass amps to me, where gain is god!

Is speculation, no more, and I can't speak for the rest of those in the list. Perhaps raising the heater to the cathode level in CF might save a few tubes? Worth a shot, but probably not.

It's my understanding that the EH is a Tung-Sol reissue. Again, could be wrong.

[GainFreak]

The one EH I have doesn't glow like the rest of similar construction, leading me to believe I may have received a slightly defective tube.

This is normal for them. I've used 12AX7EHs alot - you can barely see their glow.

Perhaps raising the heater to the cathode level in CF might save a few tubes?

I've tried it and it works /heaters elevated at 80-85V/ but I don't use more than 360V on CF now /220k plate resistor/ because with higher anode voltage and/or lower value resistor you'll exceed the heater to cathode voltage again. The other option is to elevate only the heater of that triode section in question as much as necessary because often voltages on cathode are well above 200V. This makes me think about some big amp companies using mainly chinese tubes which don't fail even at higher voltages.

[GCA]

[quote=GainFreak;110284]

I've tried it and it works /heaters elevated at 80-85V/ but I don't use more than 360V on CF now /220k plate resistor/ because with higher anode voltage and/or lower value resistor you'll exceed the heater to cathode voltage again. The other option is to elevate only the heater of that triode section in question as much as necessary because often voltages on cathode are well above 200V. This makes me think about some big amp companies using mainly chinese tubes which don't fail even at higher voltages.

Mhm. I'm mainly using a 320 B+. Thinking of giving a voltage divider off of that a shot instead of trying to elevate the whole heater supply for now since I have plenty of mA to spare in the pre. See if the EH lives thru it. If they live with the elevation then I suppose we can assume it's weak/thin insulation on the filament and be on our way.

You know, if that works nicely I could probably just simplify the supply on my pre to do that for all the heaters... then just pop a relay like a turkey timer when they're hot and the resistance climbs Save a regulator in the process.

Oh, well... maybe in a few days.

[GainFreak]

See if the EH lives thru it.

After elevating the heaters all my EHs are still OK more than a year now.

[GCA]

Originally Posted by GainFreak

After elevating the heaters all my EHs are still OK more than a year now.

Nice. I think we can wrap up the mystery, then... smaller than normal Vh-k. I wonder what the real rating should be?

[novosibir]

Originally Posted by myfoot

This was posted over to the Metroamp forum. Sounds like it works fine anywhere but the cf stage

MetroAmp.com Forum • View topic - Tung Sol Reissue 12AX7 not for CF stage !!!

Yep, this was me, who started this topic there, because I was so pissed about failing 3 tubes one after the other

But I've tried meanwhile Tung Sol reissues from 2009 production again in V2 of Marshall amps, after the vendor told me, that they've changed and upgraded the insulation issue of their preamp tubes - and... no more failure yet

Yes, I like this tube very much soundwise and it's IMO the best alternative, if you don't go NOS. I've put this 2009 production tube in 6 or 7 Marshalls (w/o elevated heaters), amps of people, who're playing 30-50 gigs a year - and no claim until now! Hopefully no claim will be coming in future

Larry

[GCA]

Nice. This didn't go out as a email notice to those with dealer accounts... they should be at least emailing about changes to their product lines. They don't, so I wonder how your vendor found out they changed insulation?

Other than that, yay!!

--GC

[novosibir]

Originally Posted by GCA

... so I wonder how your vendor found out they changed insulation?

I'll ask him tomorrow, today it's already too late here in Germany, business time is over.

He's one of the biggest tube sellers here in Europe

BTB Elektronik-Vertriebs GmbH - Preislisten

and a few times every year he's in Russia, Saint Petersburg & Saratov. So I guess, that they've told it to him there in the factory?

Larry

[GCA]

That would definitely explain it!