So here’s a fun one (AD-100 clone woes)

[Mr_bibbles]

So, there is a customer of mine with a boo-teek ad-100 copy. It has never really worked properly.

When he received it, blew fuses, blew tubes, it had been through 3 shops before it arrived here. The builder has assured me it is ‘not a design issue’, and has sent two different transformers to my shop for install, and has now washed his hands of it. (He also seemed upset the customer was placing the amplifier on top of his cabinet...) The physical construction was solid, and his company seems to do well. This was his first amp with 8 kt88’s though.

I’m seeking ideas to make this functional for this poor customer. My only real thought it to run 4 kt88’s instead of 8, or to build a second supply rail for the screens as Doug from eurotubes suggested after only a minute or two or discussion about this.

History:

The first time in the shop, was brought in with a new transformer, the builder had mailed it and said to install, no other real information so I did a careful job and sent it out.

Second time, came back shortly after with a huge carbon arc spot between plate and heater on the TOP (tube side) of the socket. The tube’s 1k screen was heat scorched, the physical metal rails that connected the sockets together had melted their clear heat shrink, and basically it needed major surgery. Cleaned and replaced everything, mounted discreet wires instead of rails...

To fix this, designer sent me another transformer, one with lower plate voltage, that was closer to the screens..

Two tubes had blown out, so I advised the customer to run with 6 for the time
being until we could prove it was stable. Now that the plate was closer to the screen and with 2 less kt88’s worth of draw, we should have an easier time right?

Amp worked for a month before he had crackling issues, but no outright failures. Brought it back, and it seemed like the screens have all taken an intense amount of heat damage for the short period they were out. I decided to up the value to 1.5k and change the 2w Metal Oxides to 5w cements. Customer really wanted 8 tubes worth of juice, so installed a new set of jj’s and biased them as conservatively as I could.

Everything tested fine In the shop, sounded good, thermally tested good for 8 hours without issue.

Went back to customers space, blew within “first minute of playtime”. Everything looks %100, except the first KT88 in line with the HV OT leads went white-topped, and had a small heat mark on the socket, Tube-side, around pin 3.

SO

What do I do? What can I do to help this poor guy get a working amp? I can’t attach the actual schematic, but I can answer any questions you have. Really would like to help this guy, probably not charging him as at this point I really feel for the guy.

Some basic info:

8 jj kt88’s
I.5k, screens 2.2k grids.
Pt500 heyboer
250w OT heyboer?
No tube voltage:
Plate: 496
Screen: 396
Grid: -47

Voltages measure great in sockets 1-8.

[Chuck H]

What is an AD-100? And if it has 8 kt88's why don't they call it an AD-200?

How is the customer using it? Are they really hitting the grids hard? What sort of speakers are plugged in and have there been any reports of sound cutting out when in use?

The last amp I built with big bottles over 450Vp blew through two pair before I landed a pair that would survive. I don't think they make 'em like they use to.

Have you clipped your meter on the grid pins to read bias and slapped the amp around a bit to see if it might be intermittent with vibration?

[Mr_bibbles]

Chuck, thanks for your reply!

Maybe I’m misremembering, it’s possible the designer said Ad200! I know that’s not incredibly helpful...

Customer uses two bass distortion pedals , seems like he is cranking the volume/pre a bit, but not like an insane person would. This is the third set of KT88’s having issues, these are JJs rated well above these plate voltages. I’ll test for vibration but that seems unlikely if it was running for a month at one point, right?

[Chuck H]

Have you noticed any pattern regarding WHICH tube/s are failing? If, for example, it's the same two all the time that could be part of a diagnosis. And though the JJ's may be rated for higher voltages, in my experience it doesn't mean they'll actually take it. There's also the matter of screen voltage and screen circuit impedance. I've had plenty of trouble with new tubes and I incinerated a pair of JJ's in the above mentioned amp. There are some things that can be done but it's better to know the cause before affecting a solution since unnecessary repairs can complicate diagnostics. Do you have the amp on the bench now?

[Enzo]

And what is the nature/mode of the failure?

I maintain it isn't voltage that kills the tubes anyway.

[Mr_bibbles]

Enzo,

Both times (three including the failure before I got it) it’s blown the main fuse, but once it arced from heater to plate on the first tube fed from the filter caps, and once in a separate socket it white-topped a tube and left a burn mark on the tube side of pin 3 plate. This tube was tied directly to the OT lead.

[Mr_bibbles]

Chuck! See my response to Enzo about the different failures. Not the same each time, the only difference between operation was the heater—> plate arc happened with 6 tubes and 1k screens, the white-topper happened with 8 tubes and 1.5k screens.

I want to prove a single failure mode for sure, and with the knowledge that this amp has NEVER worked correctly, even when brand new, I want to examine the working parameters as if they may be faulty. This is a bit above my pay grade, but it’s on the bench (when I’m back Tuesday) and I’m happy to include any measurements and information!

[g1]

Multiple arcing issues make me immediately suspicious of too high cab impedance or speaker cable going open.
Did the 1st replaement PT fail or was the 2nd replacement just to try with tamer voltages?

[Chuck H]

Multiple arcing issues make me immediately suspicious of too high cab impedance or speaker cable going open.
Did the 1st replaement PT fail or was the 2nd replacement just to try with tamer voltages?

THIS^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

With failures to both the main fuse AND the tubes themselves I'll wager there's shenanigans in this amps daily operations. Maybe an MOV and some protection diodes on the OT primary??? Your customer is probably abusing this amp. Both in being cavalier about plugging everything in properly, possibly the AC wiring where he plays and maybe with how hard he's driving it relative to it's intended function (read that too much grid drive in a wide open circuit will exploit a low G2 circuit impedance every time!).

I think you and your client could benefit from such things as an MOV on the AC primary, protection diodes across the OT, a 220 ohm dummy load resistor across the output (A big one, obviously. Maybe chassis mount) and a significant increase in screen circuit resistance. Gotta keep the tubes alive!!!

[Mick Bailey]

A white-topped tube is perhaps an unrelated failure to the others - it's caused by a vacuum leak will cause the tube to break down and short. I had a problem KT88 amp in a while back. It all measured OK on the bench even when driven very hard. So, the first time a tube had shorted (JJ). Screen resistors were 470R so maybe the amp was designed originally to take 6L6. I changed these to 1K and fixed some issues with wiring dress. It went back out and after a few weeks the owner reported it was glowing brightly and hot so had to turn it off. All 4 tubes had the logo bleached white - a sure sign of overheating. I reworked the bias circuit to fail-safe, replaced the scorched screen resistors and fitted a new set of tubes. Tested it brutally. It went out and came back a few weeks later with the same fault.

Had it on the bench again - no fault. Two hours at maximum output (constant sine input) into a dummy load until I thought it would self-destruct. So my final action was to install a large 220R resistor across the speaker output wiring and the owner supplied me with another set of tubes - this time Genalex rather than his usual JJ.

This is now coming up to 14 months of hard gigging by a full-time pro with no further problems. Another change is the amp is now running off a power conditioner and I supplied the speaker cable. B+ is 580v, screen voltage 320v

[Chuck H]

So my final action was to install a large 220R resistor across the speaker output wiring and...

OOOooh! Good one that I missed. I edited and added it to my post above

[Mr_bibbles]

Hey G1, just to be sure I had him bring the cab in and it measured ok and was the proper impedance. He uses the cab with another head, no issues.

No PT has failed in this amp, I believe it was he designers attempt to alter voltages. This first time bumped up the heater voltages I believe, the second time lowered the B+.

[Mr_bibbles]

Thanks for getting back Chuck! Let me address hear one at a time:

MOV: the amp failed at a show one f the times it went down, so while I agree they’re a good idea to install, I’m not convinced the AC is causing these issues. Same story with the protection diodes!

As for the dummy resistor, there are 2x 10W 4.7k cements hooked
From ground —> speaker terminal positive lug.

Considering all the different failures, in different environments, is there a way to look into the screen supply/draw to find an inherent fault?

[Chuck H]

You need to know what the screen circuit is. Is it choke fed after the first HV node? What is the inductance (so that the impedances can be calculated)? Is it resistor fed? What value? Is it some combination? of these? Are individual screen grid resistors in place on every socket? What value?

What I've found is that if you're really hammering the grids then if the screen circuit impedance is low, peaks and spikes try to find a way through the screens instead of the plates, which are almost always at a higher impedance than the screens. At least that's the perception as I've seen it (corrections accepted). Increasing screen circuit impedance (and sometimes that can just mean resistance) seems to help greatly.

[Mr_bibbles]

Mick! Thank you.

My initial thought after the designer “washed his hands” of the amp was to increase the value of the screens from 1k to 1.5k to cool it down. Right before I did, it was running but you could tell every screen resistor in there was getting hit with a lot of heat. Unfortunately, right after this is when the tube popped.

As I said above, the amplifier has 4.7k 10w resistors from ground to the speaker +. Since you suggests 220, is it possible these dummy loads are so large that they’re causing issues??