discounted SED factory seconds

[imaradiostar]

tubedepot.com is selling winged c SED tubes that are listed as "factory seconds." They say they pass their testing so I ordered some KT88's and some EL34's. I'll post when I get them. If they're seconds I don't have to feel bad about leaning on them a bit!

I don't know if they have a lot of them or not but I thought I'd mention it!

jamie

[antieatingactivist]

I just got a quad for my soldano. They sound great!. One even has a blue groove tubes print on it.

[Gtr_tech]

Strange......as dodgy as new mfgr tubes are these days I'd be a little interested in exactly *why* they are deemed a "second". Not so sure I'd wanna be a crash test dummy......

[Enzo]

Maybe they forgot to put the Surgeon General's warning on them. Or possibly they contain peanuts without a notice on the nutrition label.

Peanut Butter - WARNING, contains peanuts.

[antieatingactivist]

so far, they are exactly as tube depot described. 3 of them have tiny scratches on the base, and one has a groove tubes screen print on it for whatever reason. They sound noticeably better than the sovtek's that I had in there before. Hopefully they will outlast them too.

[Bruce / Mission Amps]

A couple years ago, I bought 300 (three hundred) JJ ECC83s 12AX7s in a sale like that.
I had exactly 17 that were completely unusable, ... bad/no good right out of the boxes... and 71 others with the two triodes being about +- 35% different in terms of transconductance between the two side, as measured on my Hickok 539. All the other tubes were +-10%.
It looked bad but those 71 tubes still seemed to work fine in every guitar amp put them in anyhow... ?? Go figure
I guess they randomly checked a handful from each box and thought they were all going to be bad and offered them up as seconds and or unchecked and uninked.

[imaradiostar]

Haven't put them in an amp yet but both the EL34's and the KT88's test very well on my BK fet transconductance tube tester thinger...and they appear to be nicely matched. Wish I'd bought more!

jamie

[drewl]

Thanks, I ordered some tubes I needed.
What the heck are those Black Sable tubes? Who makes those?
Dayum they're expensive.
Makes me feel good to know my stash of Mullard and Telefunkens are going for $150 a pop

[Diablo]

Thanks for the tip. I just blew another $120 on EL34 and 6L6 tubes.....
I always wanted to try the SED brand.

[imaradiostar]

Made up a test rig to test these tubes and various output and power transformers. The SED KT88's did great- about 65 clean watts output at 510 plate volts, 350 screen volts and 4k output z. Looks like they'll be good for un upcoming high-gain project.

The EL34's seem to always have a lot of crossover distortion, even when biased hotter than safe levels. Would this be indicative of mismatched tubes?

jamie

[imaradiostar]

OK, some updated info on these things. I spent some time with a test rig trying different currents and voltages and different output impedances.

One of the two EL34's tends to redplate at idle. It doesn't draw what I would consider to be excessive current- just gets a little cherry color, even when biased relatively cold. It goes away once it's passing signal but it does make me a little wary. It would appear the matched pair isn't matched very well- when going for max clean output one side always clips before the other and it seems to follow the tube. I haven't abused them other than the occasional blast to check max output and current draw under clipping.

I managed to kill one of the kt88's- really it's totally my fault. I had it cranked up for square wave clipped output all in the output tubes- 530 volts on the plates, 350 on the screens, idle bias was around 40ma/tube. Screen grid stoppers were 300 ohm wirewounds. I wasn't paying attention and all the sudden it arced and glowed blue and ran away. Even after cooling the cathode glows blue when I pass signal and it tends to get hot and run away. I guess I'll keep the other one for a single ended amp.

I'm probably going to order some more though- I intended to beat on these and I'm OK with them dying for only $20 a tube!

Here's what I found in testing- all tests were into an 8 ohm non-inductive load.

The El34's made a pretty clean 66 watts into a 4k plate load. They were biased at about 25ma each with 350 on the screens and 550 on the plates. At max clean output they used 271ma at 341 screen volts and 516 plate volts. There was a little bit of crossover distortion and phase shift that I chalk up to the output transformer and mismatched tubes, but overall not bad. When I connected the output TF as 8k the primary inductance was an issue at lower frequencies but it still made 55 watts.

The KT88's were capable of a little bit more until I blew them up- I was getting about 70 watts which seems in line with what they're supposed to do. I'm frustrated that I never got to crank up the voltage! The whole point of this exercise is to find out what it takes to get high real-world wattage from a pair of tubes. I was hoping for 100 watts a pair from these at over 600 volts. I'll have to order another pair and try again.

jamie

[Steve Conner]

Maybe your KT88 blew up because the grids weren't aligned properly? That's the kind of thing that would make it a "factory second".

Under heavy drive, the misaligned part of the screen grid would draw too much current and melt, and it sounds like that's what happened to you. The tube is now gassy.

The EL34 that redplates, does it do it in one small spot only, or just on one side of the plate? Could be misaligned grids again, this time the control grid. If it's unevenly spaced from the cathode, the current flow at idle will be uneven too.

[imaradiostar]

The EL34 gets red in the middle of one of the bent sections of the plate on one side only- about a 1cm by 2-3cm area. At first I got a pretty clean waveform. If I let it get a bit hotter (and redder) it starts to effect the performance of the tube and the scope trace starts to suffer; power drops off a little bit and one end of the sine clips before the other.

It sounds like you're right about the KT88 too- I was surprised how little abuse it took to cook one. I know the screens can be fragile on them but my screen voltage was only 350 and I was using 300 ohm screen resistors per side. I have a feeling the upward mismatch could be a contributing factor- I was pushing them pretty hard into a 6600 ohm load. They never complained when pushed into 3k, 4k or 5k with other transformers. I also was easier on them at those loads.

I intend to purchase some non seconds for some more testing. It's fun to play with the numbers and see what you get. From my observations thus far I'd say that either the data sheets lie or the modern tubes I'm using aren't up to the task like the old stuff...or both!

I barely got 30 watts out of a pair of old 7868's at 550 plate volts and 350 screen volts into a 6k6 load. From looking at the datasheets I would have expected closer to 35 or 40, even without the extra 50 screen volts required to produce the 45 watts shown on the datasheet.

On the other hand a pair of modern production EH 7591a's produced exactly the datasheet power output shown for a cathode biased ultralinear output- so who knows. I did match all the operating values shown on the datasheet exactly though.

jamie

[imaradiostar]

I managed to kill one of the EL34's today as I upped the screen voltage. I kept swapping back and forth from some old Chinese El34b's to the SED tubes. The SED made consistently less power and the screens would glow much sooner under load. I was using undersized screen resistors and torturing them pretty badly so I'm not surprised that one quit. Now it does nothing- just sits their while its neighbor still tries to pass signal.

I did notice that when the screens started to glow on the SED EL34, the one that went bad tended to glow unevenly- parts of the wire were brighter than others. I suppose a part either melted itself to another element or disconnected the screens altogether. Whatever the case- it doesn't work anymore!

On the other hand, I was surprised at the abuse my old Chinese El34's could take. When I risked the screens glowing they put out a pretty clean 88 watts with a little over 500 volts (under load) at the plates and screens into a 4k:8 load.

I want to understand what makes different amps do what they do so this has been an interesting experiment!

jamie

[Diablo]

I put in a pair of EL34s in my Tweed Twin Clone, biased them at 36mA and 432 volts, the pair are almost a perfect current match within a few tenths of a mA. So far so good, but I'm not doing any torture testing.