Failure of JJ KT88

for Dynaco Mark II/III/IV and DIY PP monoblocks

Failure of JJ KT88

Postby kt88pppamp » Fri Jun 17, 2011 2:49 pm

In my pair of MKIIIs I today experienced a tube failure due to arcing. When I switched on my amps in the morning, I heard a SNAP SNAP SNAP SNAP ... POP POP POP POP sound in my speaker. Looking inside one of the KT88's, there was a blue flash in synch to the popping.

I once had a brand new pair of JJs that did that. The ones that that crapped out today are around 2 years old with daily use.

The tubes in my other amp are around the same age. Which brand do you recommend? I have a bunch of Electro Harmonix ones on hand. Are they any good?
User avatar
kt88pppamp
KT88
 
Posts: 530
Joined: Tue Sep 26, 2006 12:11 am
Location: Rochester, NY

Postby Geek » Fri Jun 17, 2011 4:04 pm

Uh-huh.... I've replaced a few sparky KT-88's that were JJ. I refuse to use them anymore.

Sovtek or Winged =C= Svetlana I've used no problems.

Cheers!
-= Gregg =-
Fine wine comes in glass bottles, not plastic sacks. Therefore the finer electrons are also found in glass bottles.
User avatar
Geek
KT88
 
Posts: 3585
Joined: Sun Oct 21, 2007 3:01 am
Location: Chilliwack, British Columbia

Postby Ty_Bower » Fri Jun 17, 2011 5:02 pm

I bought a set of New Sensor "Gold Lion" reissues when I rebuilt my Mark III. They are expensive, but I've been nothing but happy with their performance. I do admit that I've got too many amps, and my Mk3 do not see continuous use. I'll also note the reissues do seem to show rapid getter wear, or at least more rapid than I might have otherwise expected. So far, it hasn't affected their sound.

When it comes time to purchase another set of tubes, if I didn't want to pay the premium prices I'd probably go for the ElectroHarmonix. I've heard they perform well and the cost is reasonable considering the alternatives. If I could land a set of JJ for cheap, I might be tempted to spring for those as well. Many people seem to bash JJ (especially their octal tubes), but almost as many support them. I have a pair of 7027A from JJ and rather quite like them a lot.
"It's a different experience; the noise occlusion, crisp, clear sound, and defined powerful bass. Strong bass does not corrupt the higher frequencies, giving a very different overall feel of the sound, one that is, in my opinion, quite unique."
User avatar
Ty_Bower
KT88
 
Posts: 1494
Joined: Wed Mar 21, 2007 2:50 pm
Location: Newark, DE

Postby kt88pppamp » Fri Jun 17, 2011 5:21 pm

The sparking took out the rectifier tube and fuse. How will sparking tubes affect a regulated power supply? Would it kill the regulator IC's and diodes? I am using a maida circuit with a FET as the pass element.
User avatar
kt88pppamp
KT88
 
Posts: 530
Joined: Tue Sep 26, 2006 12:11 am
Location: Rochester, NY

Postby dcgillespie » Fri Jun 17, 2011 5:26 pm

If your amplifiers do not have screen stability resistors in place, that should likely be the first order of business to resolving your arcing issues. If your amplifiers operate the output stage with fixed bias, in Ultra Linear mode (or operate the screens very close to the plate voltage), and otherwise operate the tubes at more than about 450 v B+, these resistors are an absolute must. If they are missing, simply installing small 1/2 watt resistors of around 100 ohms or so in series with each screen lead will almost certainly eliminate the problem if the above conditions exist. Keep the resistors physically close to the screen pins for maximum effectiveness. Good luck!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby kt88pppamp » Fri Jun 17, 2011 5:47 pm

Your talking about the MKIII's right? Screen stability is needed?
User avatar
kt88pppamp
KT88
 
Posts: 530
Joined: Tue Sep 26, 2006 12:11 am
Location: Rochester, NY

Postby dcgillespie » Fri Jun 17, 2011 6:41 pm

I'm absolutely talking about the MK IIIs. However, I was somewhat general in my comments because some of these amps are modified so much that the only thing resembling the original amplifier is the chassis. I did not know if yours are stock, or highly modified amps, so I tailored my comments accordingly.

One area that is almost guaranteed to encourage output arcing in an otherwise normal MK III output stage is the use of increased power supply capacitance at the point supplying the output stage. It is a very common mod to increase this value well over the value of the stock component. When that is done, arcing events are likely, so that screen resistors become mandatory when this modification has been performed.

If you would like to read more about this, you can do so in an article I published for Audio X Press some years ago, wherein I identified the problem, and first offered the solution of installing screen stability resistors to resolve the issue. You can find the article here:

http://www.audioxpress.com/magsdirx/ax/ ... ie2544.pdf

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby kt88pppamp » Fri Jun 17, 2011 7:19 pm

I'll drop in some stoppers in my MKIII's because the capacitance is higher than stock.

My 120 watt KT88 PPP amp that I am working on has 47 ohm 2 watt stoppers (both screen and plate) and a regulated solid state B+ power supply. I just am concerned about arcing here because I saw how violently the GZ34 failed. Should I put in some safeguard in that maida regulator or will the stoppers bring everything under control?

I am wondering too, is it really instability causing the arcing or defective tubes?
User avatar
kt88pppamp
KT88
 
Posts: 530
Joined: Tue Sep 26, 2006 12:11 am
Location: Rochester, NY

Postby dcgillespie » Fri Jun 17, 2011 8:05 pm

If the regulator and/or rectifier failed from the output tubes arcing, then solving that issue will eliminate likely the most major cause of regulator failure.

In my experience, cheap power tubes may be somewhat more prone to catastrophic failure, but in all my experiments, the tubes that were arcing were high quality GE manufactured 6550s -- hardly a poorly manufactured tube. Since I began installing stability resistors nearly some twenty years ago now, I have not had a single arc event from any tube type of any manufacturer.

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA

Postby Shannon Parks » Sat Jun 18, 2011 5:31 am

Thanks for the article link, Dave!

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Postby kt88pppamp » Sat Jun 18, 2011 7:59 am

Great article. I will now kick on my driver stage 30 seconds after power up, and the power B+ 10 seconds after.

A microcontroller will control two reed relays, rated at 1KV each. On power up, it will start a 30 second timer, which will kick on DB+ (the driver stage B+). A second reed relay, in the path of the output stage B+ will kick on 10 seconds later.

Lets say the power goes out. The relays and micro are powered by the 6.3V winding so the relays will open as soon as power is lost. To take care of the charge (I have been shocked with several times) on the capacitors, bleeders will be installed across them.
User avatar
kt88pppamp
KT88
 
Posts: 530
Joined: Tue Sep 26, 2006 12:11 am
Location: Rochester, NY

Postby dcgillespie » Sat Jun 18, 2011 9:45 am

I hope the information will help save your tubes!

The micro-controller is a great idea. At the time I was working with such circuits, I could not get the relays to consistently and absolutely fall out during very short power cycling events -- of which I was experiencing a lot of at the time. Therefore, I went the series AC timer/relay approach, that would immediately fall out, and then restart through the full time cycle, even with very short power cycles. The concept has not failed me yet in over 20 years, but I have no doubt that you can make your approach equally good, if not better! Good luck with it!

Dave
dcgillespie
KT88
 
Posts: 399
Joined: Sun Mar 08, 2009 5:34 am
Location: Ball Ground, GA


Return to poseidon

Who is online

Users browsing this forum: No registered users and 20 guests