Are new production EL34's just this bad or is there a problem in my amp I don't see?

[jeffroman]

I have a 2203 (modded by Lee Jackson in the mid 80s sometime with the extra preamp stage and 5 position FAC selector switch) and recently gave it a bit of an overhaul. Some new pots, new caps, replaced a bad tube socket. Since this work, I have noticed that new quads of tubes bias fine and run fine for about a half hour and then one either shorts or causes a red plate across two sockets. I'd be inclined to blame one of the original sockets but I kept two of the set I ran for a year before the work (Ruby EL34-BSTR) and the amp is totally fine to run continuously with them in either the inner sockets or outer sockets. On the first set (EHX EL34) I identified the shorted tube and ran the amp fine with two others from the set, then I just got a set of Svetlana's and these ran fine for 30 min and then two started redplating. The guts look fine, and as I said properly burned in and known working tubes operate as duets in either position, but I cant seem to get any quads I can trust to run without a failed tube. I've been biasing to 36 mA, and never experienced any problems with JJ's or Ruby's in this amp. Is it likely that I'm just getting poorly tested tubes, or is there an issue i'm not considering? Bias is stable, screen grid resistors are all fine, and the connections in the power section are all solid.

Before the BSTR's I had JJ E34L, and am now considering switching back to the JJ and trying some 6CA7s since I have never gotten a failed tube from Eurotubes but didn't like the E34L that much. The Svetlanas sounded great before the redplate (as it goes...) but I'm starting to worry a lot more about tube suppliers for these EL34s, which has never been a concern for my 6L6 based amps, which seem to do just fine regardless of the tubes i throw in them.

[Enzo]

You got good service from the Ruby's, and if I read you right, they still will work in it? If so, why not get a new set of the Rubys? I have used a lot of them and so far never had one come back on me. They seems to work well and reliably for me.

[J M Fahey]

Agree and add: by using the same kind and brand of models you used since forever, you ara chalking down one more variable.

If new ones work fine, you solved your problem; if problem reappears, now you know (or at least suspect) that it's NOT tube related, so you can search for something else.

Worst is having many *possible* suspects and not enough data to decide on one or another.

[lowell]

Just like to add I just repaired an amp that was eating power tubes. It was a bad OT. Amp worked, but a power tube would arc shortly after installation. Worth checking that the bias is equal on both sides of the OT....and further to check the DC resistance of each primary to CT. Just a thought.

[soundmasterg]

I had a Bassman Ten several years ago that I repaired for a customer and it seemed fine on the bench for hours of play and testing, but as soon as it got to a live show tubes would red plate. Ended up being sockets. Once a tube in a pair or quad red plates then the others are suspect also, and once a socket arcs, especially the old plastic ones, the socket will likely keep arcing. Ceramic sockets are immune to this problem though they are harder to tension.

Greg

[jeffroman]

@Enzo, I have also had great luck with Ruby Tubes, a good bit of this post is my frustration at really ONLY having good luck with them. I like the sound of their 6L6GC-MSTR in the high gain Peaveys I and friends use, and dug the EL34-BSTR in the Marshall, but was hoping to find a new flavor with a new brand. Unfortunately,except for Eurotubes tested JJ's (which I haven't dug the sound of much), it seems the Ruby's are the only brand I've used that are reliable for me. I am mostly curious to see if anyone else has noticed a very high failure rate in EL34 quads from new production Russian tubes. I understand both Svetlana and EHX tubes are made at the Reflektor factory in Saratov and I have some friends who swear by old Reflektor made 6L6 tubes from ebay, so it seems to me that their production should have only improved if they've been continuously operating or at least their burn-in and testing procedures before sorting/matching. I will say that both sets sounded very nice before they ruined my day.

I may swap out the three original tube sockets and slap on some new screen grid resistors for good measure/thoroughness, and will definitely check my OT windings, but the last show I played with the amp it sounded great and ran totally fine. This continued exactly until putting in the new set. I am fairly inclined to blame tube failure here, but I'll stick the quad in a tester before getting the return invoice.

[lowell]

JJ hasnt failed me in 10 yrs. I might have had 1 or 2 bad power tubes ever. And I've used A LOT for repairs over the years. IME they are super reliable and the only brand I use unless something is specifically requested. JME and JMO. On the other hand I tried Ruby's to check em out and had a bad tube. May sound crazy but it's true. I guess tubes are tubes. They're fragile. Even if they leave the factory good who knows how they're treated in transit. Stick with what you're comfortable with. When it comes to tubes I choose reliability, availability, and affordability ONLY. Not tone. Tone can then be tweaked in other ways.

[Enzo]

So buy JJ tubes from Ruby, or whatever brands you like. Ruby tests the tubes before putting their label on them, and they do offer warranty on them.

[km6xz]

If they are New Sensor "Svetlana's" they are basically Sovtek and not to be confused with real Svetlana tubes with have the Wing-C logo. They went out of production in 2012 but some are still available at ever increasing prices because they are likely the best quality tubes available. If both tubes on one side are red-plating, it could be one failing and pulling the bias down for both by having a positive leakage. Try swapping the tubes side to side, after marking them with numbers to tell them apart. There were some JJ tubes which had too narrow pins that caused problems with grid connection but new stock have the correct pin diameter.
Unfortunately tubes now are all over the map in specs within the same brand so relating a problem to a brand is not very helpful unless you eliminate the big variable of what the transconductance of the specific tube in hand has. The problem might be as simple as the the GM being way different than the tubes on the other side. A run on a curve tracer showing the curve families on modern tubes show a lot of variation between units.
Brands are not makes, but reseller labels, since there really are only 3 companies making 95% of production. Ruby(Magic Parts does have a low production tube plant in China that was making 3 tubes;12AX7, 6V6 and GZ34. The rest of their 12AX7s are Chinese, JJ or the New Sensor made Russian tubes which carry brand names of Svetlana, Mullard, TungSol, Sovtek, Electro-Harmonix and several others. Shugaung in China makes a lot of tubes of various markings for different resellers.
That is pretty much it. JJ have a love'em or hate them reputation and are essentially unusable here with horrible failure rates but in the US some techs think they are great, it all depends on which distributor they get them from. All the companies sell different grades of the same tubes to distributors, which is what causes the problems. Tubes that have to little or too much Gm are still sold but in the old days, those tubes not meeting spec were simply crushed as defective in QC and never sold. The color code, current rating number or type code on the tubes that the rebranders put on their retail version give a clue what they match to but none of them say what those codes mean in terms of real specs. A good Mutual-Conductance tube tester can help but a curve tracer that can sweep realistic parameters and rated plate load is really the only way to see what they really are. The brands do not have curve tracers from what I can tell, I know a lot about how Ruby tests theirs and they do not have one. Apparently none of the companies would bother testing that way anyway, so it is up to you to figure out whether a set of tubes or a batch will work for you. The most confusing element for users is the internet chatter that convinces the technically unsophisticated of how certain brands have a "sound". Often you will read BS about one brand having a brighter top and muddy low mids or a shimmer upper mids and compression or others happy talk. What they are talking about is the tube they tried, due to the different transfer function of unidentified tube characteristics of the exact tube interacting with their exact amp, with their exact settings. Change any of those and all bets are off. The biggest single cause of internet gurus stating a particular brands has a certain sound is the Gm of the exact tube they tested in their amp leaving the settings alone. A tube with more gain will sound different. Brand does not matter. The claims of some brands sound brighter for example ignores the fact that any functioning 12AX7 will have at least DC to 60,000,000 hz flat response, so if they are hearing more brightness, or think they do, it could be that tube is getting into a nonlinear portion of its conductance curve where it starts to compress earlier or later than another tube based solely on gain of the tube. Different gain gettings on the amp would greatly change the perceived tone character. In tests where gain and circuit conditions are normalized between tests, A/B blind listening tests trip up the golden ear gurus every time. Normalize for drive signal and gain and differences disappear.
So find out if the tubes are failing or are fine but just not of the same Gm of the other side, by swapping the two lower tubes with the two upper tubes. If the red plating moves, pull one tube from each side and change you dummy load Z to double and see if the problem is with one tube. When one tube is identified as red-plating in which ever socket it is in, you have to decide whether it is defective or just has different gain or sensitivity than the other side. If still could be fine, but just not compatible with the bias setting you selected for all tubes overall. When buying tubes, how they are tested, rated and the ease of warranty replacement are the most important criteria in selecting a distributor.

[Mick Bailey]

The red plating situation needs to be investigated. Two tubes red plating suggests a bias issue. The chances of getting two tubes that drift so far off they red plate is pretty slim. Set the bias really cold and monitor the current draw and bias voltage.

Because the problem has only started since the overhaul, I would first suspect it's a consequence of the work rather than the tubes. Modern tubes have their problems and early-life failures aren't uncommon (whatever you buy), but even the factory-test Shuguang EL34s hold up pretty well once they're initially matched and burned in. So long as the current draw, plate and screen voltages are acceptable.

[pdf64]

The red plating situation needs to be investigated. Two tubes red plating suggests a bias issue. The chances of getting two tubes that drift so far off they red plate is pretty slim. Set the bias really cold and monitor the current draw and bias voltage....

Also it may be a good idea to monitor the output periodically, at various signal levels, ideally with a real speaker load; it may be that it's flipping into oscillation under certain conditions.

[Mick Bailey]

Ah, I overlooked oscillation. Good point.

[Chuck H]

Excellent commentary. And of course the OP's answers are already here. I'm going to add that my own recent experience mirrors the OP's in some respects. Trying to find new production EL34's was/is a real PITA as soon as you step away from Ruby selected tubes. Some Sovteks suffered a slow heater death on one tube. One of the JJ's shorted plate to heater the moment the amp was clipped. Another pair off JJ's had the bias drifting around and just wouldn't stabilize. The JJ E34L held strong, but sounded like a pair of crows (caw caw) as the OP noted. All this happened while trying to tube one amp, one time! Stay with the Ruby tubes. They were the only EL34's that held fast, biased up as I would expect and sounded good. I actually like the tone and performance of the Ruby selected Shuguang EL34's about the best out of several I've tried. Unfortunately this is a combo amp and of three sets of Ruby Shuguangs I purchased, all three were unacceptably microphonic for use in combo amps. I wouldn't hesitate to use them in a head cabinet though. YMMV

[jbltwin1]

Honestly, I have used and installed a lot of JJ's and don't HAVE these problems. Iuse them in cust stuff and my own. I swear, it seems to be something to do with the shipping. Now I only buy from one place but I work on a bunch of Music Mans and they are tough on tubes and the E34L's are the only thing I will install. Maybe the old saying of "caution fragile-throw underhanded" comes to mind. Mike.

[Chuck H]

Maybe the old saying of "caution fragile-throw underhanded" comes to mind.

Keep trying! You almost got a spit take with that one