KT88 Bias Troubles

0 Members and 1 Guest are viewing this topic. Read 10111 times.

Freo-1

KT88 Bias Troubles
« on: 14 Jun 2008, 10:55 pm »
Over the past year, I've tried several quads of the Gold Lion KT88 reissues.  Although they sound great, they simply do not seem to have good longevity in MY amplifier.  Two of them had shorting out problems internally, while another five of them had issues with bias runaway problems (going cherry red).  The tubes seem to be fine for awhile, then,  out of nowhere, will go cherry.

I am at a loss to figure out the root cause of this issue. I've had feedback suggesting the amp was the cause.  Although I understand the comment, I am not sure that is the case.

The problem with blaming the is as follows:

1) All other 6550/KT88 types used work fine (GE 6550A, Svelenta 6550C, JJ KT88, Svelenta KT88)
2) The above tubes that work all have spec sheets, which define their operating parameters (and can be analyzed). These tubes all work fine in circuit.
3) I've had a tech go over the amp to ensure nothing is wrong (and replace bias resistors when the GL tubes shorted).
4) The GL KT88 Re-Issues have no data sheets available. 

Because of the fact there are no spec sheets available, it is very difficult to figure out what's going on.  While some customers seem to not be having problems, I have seen other posts where people have reported this bias runaway issue.  There seems to be something that appears to behave as excessive grid leakage.


Any ideas?  :scratch:

« Last Edit: 15 Jun 2008, 12:11 am by Freo-1 »

richidoo

Re: KT88 Bias Troubles
« Reply #1 on: 14 Jun 2008, 11:10 pm »
Are you watching the bias closely during break in, and setting it anew when swapping brands? Sounds like too much bias current burning them out prematurely, the behavior is typical of a worn out tube. Typical bias on KT88 is 75ma, how much bias are you putting to them?

You could email Jim McShane, he sells them and might have additional insight. jimmcshane -at- prodigy.net
From Jim's website:
"Genalex Gold Lion KT-88 reissues - good stock on hand! The listening reports are virtually unanimous, regardless of what gear the tubes go in. This tube is clearly the equal of and according to many may even be better than the original GEC. It is absolutely the best on the market now! My impressions from handling and matching the tubes are extremely positive. The tube is very substantial in your hand, they feel heavy and look well made. Pin soldering is excellent! There have been just a small handful of the nearly 800 tubes I've sold that have had any issues at all."

Freo-1

Re: KT88 Bias Troubles
« Reply #2 on: 14 Jun 2008, 11:54 pm »
Are you watching the bias closely during break in, and setting it anew when swapping brands? Sounds like too much bias current burning them out prematurely, the behavior is typical of a worn out tube. Typical bias on KT88 is 75ma, how much bias are you putting to them?

You could email Jim McShane, he sells them and might have additional insight. jimmcshane -at- prodigy.net
From Jim's website:
"Genalex Gold Lion KT-88 reissues - good stock on hand! The listening reports are virtually unanimous, regardless of what gear the tubes go in. This tube is clearly the equal of and according to many may even be better than the original GEC. It is absolutely the best on the market now! My impressions from handling and matching the tubes are extremely positive. The tube is very substantial in your hand, they feel heavy and look well made. Pin soldering is excellent! There have been just a small handful of the nearly 800 tubes I've sold that have had any issues at all."


As a matter of fact, I have been in touch with Jim, who is a great guy. He's at a loss as well.  He reckons its just a combination that does not play well together.  I understand where he is coming from. For example, he pointed out that some of the new production 7591s do not work work well in the old Scott amps.

Without the spec sheet, its pretty much a shot in the dark to get to the root cause.  The fact that all the other ones I listed earlier work fine is what's bugging me. 

richidoo

Re: KT88 Bias Troubles
« Reply #3 on: 15 Jun 2008, 12:33 am »
Yup, makes sense.  But despite the lack of a logical answer to satisfy the intellect, the symptoms suggest over-driving the tubes. So maybe try running them at 50ma or even lower and see what happens. Maybe start at 30ma and add 5ma at a time until you hit the point of dimishing returns, trying to find the lowest bias point at which it still sounds great. You will have to do it over time to allow the brain to adjust, because higher bias will usually sound "better" in A/B. Verify the resistors for setting bias are correct, etc.  Maybe you have inacurate bias gaging, and the other brands just tolerate overdriving better.
Good luck freo

Freo-1

Re: KT88 Bias Troubles
« Reply #4 on: 15 Jun 2008, 12:56 am »
Yup, makes sense.  But despite the lack of a logical answer to satisfy the intellect, the symptoms suggest over-driving the tubes. So maybe try running them at 50ma or even lower and see what happens. Maybe start at 30ma and add 5ma at a time until you hit the point of dimishing returns, trying to find the lowest bias point at which it still sounds great. You will have to do it over time to allow the brain to adjust, because higher bias will usually sound "better" in A/B. Verify the resistors for setting bias are correct, etc.  Maybe you have inacurate bias gaging, and the other brands just tolerate overdriving better.
Good luck freo


Thanks for the thoughts, but no, the bias circuit is correct.  The meter is an NOS Simpson ammeter (very accurate), and each tube has a MIL-SPEC Dale pot in line.  The bias circuit is not the issue.

The 6550C, JJ, and GE are are pretty rugged tubes.  I suppose it's possible that the GL is not as rugged, but it is pitched that it can handle high voltage/current (but WHO knows for sure, no spec sheet available)

I normally run the tubes at 60MA. One of the times I had trouble with the GL's, they were biased at 50MA, so that is not the issue either.

I will be curious to monitor peoples progress with the GL longevity with the variety of amps it is used in.  I hope it works well, as it does sound great.  My fear is that it will not be a tube know for long life, but time will tell

GBB

Re: KT88 Bias Troubles
« Reply #5 on: 15 Jun 2008, 01:02 am »
Can you provide more details on your amps?  What brand are they?  Do you have a copy of the schematic?

The behavior you're seeing would be consistent with the Gold Lions having a bit more grid current than the other 6550 tubes.  If your amps use a large value grid resistor then they could start running away.  Dropping the value of the output tube grid resistors and perhaps increasing the coupling caps to keep the same low frequency roll off would be the cure.  But that would require some soldering to see if it helps.

---Gary

Freo-1

Re: KT88 Bias Troubles
« Reply #6 on: 15 Jun 2008, 01:19 am »
Can you provide more details on your amps?  What brand are they?  Do you have a copy of the schematic?

The behavior you're seeing would be consistent with the Gold Lions having a bit more grid current than the other 6550 tubes.  If your amps use a large value grid resistor then they could start running away.  Dropping the value of the output tube grid resistors and perhaps increasing the coupling caps to keep the same low frequency roll off would be the cure.  But that would require some soldering to see if it helps.

---Gary

I can get a copy of the schematic from the tech.  The amp is a restored/modified Citation II.  The front end has been converted to triode, using a 5687 for a driver, and a 5687 as the phase splitter.  The power supply has significantly more reserve, and the bias circuit was re-done to make it more stable.  The ammeter is a NOS Simpson (which is accurate), and each bias pot (one per tube) is a NOS Dale MIL-SPEC. 

The thing that bugs me is that ALL the other 6550/KT88 tubes I've used all work fine.  So, does that mean that the GL Re-issue is not a true 6550/KT88, but something else that is close?  The problem is, who knows?  There is no spec sheet on the tube.

I think you are correct about dropping the resistance and  changing the coupling cap may help.   However, it seems that the GE6550A and GL KT88 should be interchangeable.

GBB

Re: KT88 Bias Troubles
« Reply #7 on: 15 Jun 2008, 02:03 am »
I know the Citation II very well - I have one.  I totally rebuilt it with new diodes and power supply caps from Jim McShane, redid the bias supply, and replaced most of the caps and resistors plus a few other mods.  I've got the output tubes wired in triode mode but haven't changed the input or driver tubes.

If your tech replaced the bias resistors, it would be good to check that he didn't change the value.  If he used a smaller resistor then you might be biasing the tubes hotter than you think.

It's also possible that when your output tubes started running away, they took out or least made the grid resistors go out of spec a bit.  The grid resistors are supposed to be 100K (R24 and R31) so you should check to see if they're still in spec.

Good luck.

---Gary

Freo-1

Re: KT88 Bias Troubles
« Reply #8 on: 15 Jun 2008, 12:38 pm »
I know the Citation II very well - I have one.  I totally rebuilt it with new diodes and power supply caps from Jim McShane, redid the bias supply, and replaced most of the caps and resistors plus a few other mods.  I've got the output tubes wired in triode mode but haven't changed the input or driver tubes.

If your tech replaced the bias resistors, it would be good to check that he didn't change the value.  If he used a smaller resistor then you might be biasing the tubes hotter than you think.

It's also possible that when your output tubes started running away, they took out or least made the grid resistors go out of spec a bit.  The grid resistors are supposed to be 100K (R24 and R31) so you should check to see if they're still in spec.

Good luck.

---Gary


Sounds like the topologies of your amp and mine are mirror images.  (Pentode/Triode vs Triode/Pentode)

I have checked the bias resistors, and they are fine (thanks for the feedback, though).  The values are same, and in spec.


Just curious, are you using the GL Re-Issues?


Cheers.

GBB

Re: KT88 Bias Troubles
« Reply #9 on: 15 Jun 2008, 05:50 pm »

Just curious, are you using the GL Re-Issues?


Nope - I'm running some of the Ei KT-90 tubes.  Unfortunately these are no longer available, but I've got a back up NOS quad for whenever my current set bites the dust.  According to Jim McShane, who I trust, the GL KT-88 reissues are the best tubes available for the Citation II right now.  So I believe that there is some quirk in your amp that you need to track down.  Or just enjoy the GE 6550s which don't give you any problems.  They're supposed to be good tubes.

---Gary

Freo-1

Re: KT88 Bias Troubles
« Reply #10 on: 15 Jun 2008, 08:23 pm »

Just curious, are you using the GL Re-Issues?


Nope - I'm running some of the Ei KT-90 tubes.  Unfortunately these are no longer available, but I've got a back up NOS quad for whenever my current set bites the dust.  According to Jim McShane, who I trust, the GL KT-88 reissues are the best tubes available for the Citation II right now.  So I believe that there is some quirk in your amp that you need to track down.  Or just enjoy the GE 6550s which don't give you any problems.  They're supposed to be good tubes.

---Gary


I also think highly of Jim, but to show you reasonable folks may not always agree, Kevin Deal (who I also trust) would not sell me KT-90s when he still had them for the Cit II, as he said that "they will blow up your amp".

Also, one of Jim's preferred mods is to run the KT88's in triode. I reckon that if I was running them in triode, they would not have this bias runaway issue anywhere near as much (if at all).

I keep going back to the lack of operating spec sheets for the GL.   Perhaps something would pop out if one was available. The fact that JJ KT88, Svet KT-88, Svet 6550C, and GE6550A all work (and you CAN look up their specs) makes pointing the finger at the amp suspect.  Unless or until New Sensor provides specs on the tube, it's a crap shoot trying to get to the bottom of these types of problems. 

GBB

Re: KT88 Bias Troubles
« Reply #11 on: 16 Jun 2008, 04:04 am »
. . . Kevin Deal (who I also trust) would not sell me KT-90s when he still had them for the Cit II, as he said that "they will blow up your amp".

Well there are KT-90s and then there are KT-90s.  I know that McShane also stopped recommending the Ei KT-90s the last year or two.  The quality of the Ei factory went down hill after all the troubles in that part of the world and you couldn't count on them any more.  But about ten years back the KT-90s were really good and that's the time period when I bought mine.

Regarding the impact of triode mode outputs vs. ultralinear on the grid current, I'm not really sure I know.  It is true that you don't get as much power out of them when you're running in triode mode but the watts you do get sure sound nice.

All the best.

---Gary

JoshK

Re: KT88 Bias Troubles
« Reply #12 on: 16 Jun 2008, 02:17 pm »
I am assuming this is a fixed bias circuit?  Can you provide a schematic as modified?  Do you know how to measure voltage on the tube and bias?  We can then work out the op point to ensure it is where it is suppose to be.

The other thing to try is put the older KT88s in, say your Svet (or JJs) KT88s and see if they work out ok. 

I don't profess to know how to solve your issue.  It could be any number of things that all look like the same symptoms.  Gassy tubes could do that, which results in excessive grid current (Morgan Jones).  But it is going to be damn near impossible to trouble shoot with you without a schematic as it has been modified and without measurements to show that the op points are where they are intended to be. 

Were the GL KT88s you had tried all bought together at the same time (thus implying they might have been from same lot)?  Or did you buy some at one point and some later? 

Reports I've read said that the GL's were really reliable tubes.  But sometimes you get a bad lot, it can happen.  However, it seems like you might be rushing to conclusion. 

One doesn't really need to see a spec sheet on a tube, but they are provided by some manufacturers.  These just provide a suggested operating point, usually the one most commonly noted in tube manuals, etc.  But a KT88 is a KT88 and by definition is intended to be operated at the original manufacturer's specified conditions.  This usually coincides with the standard operating point.  If it fails at this point, then it hasn't met the parameters for being a KT88.   

I don't think this is the case here, as I've read some who have experimented with the tube and suggest it is quite robust and can be pushed a bit beyond spec.   You either got bad tubes or something in your bias circuit is playing poorly with the tube. Fixed bias and gassy tubes don't mix well, if that is the condition you have, which usually results in a run away tube and the red glow of death.

P.S.  just read Gary's response.  Seems he is on the same thought process.  I just got done reading the chapter on biasing in Morgan Jones that spoke to the conditions of grid stoppers, gassy tubes and runaway conditions.  If I am correct that this is fixed bias, or even if it isn't, then lowering the grid leak (edited: stopper to leak) can help with preventing runaway tubes. 



« Last Edit: 16 Jun 2008, 09:20 pm by JoshK »

richidoo

Re: KT88 Bias Troubles
« Reply #13 on: 16 Jun 2008, 06:13 pm »
I wrote to newsensor looking for the datasheet for you. I was happy to receive a response today.

"The Genalex Gold Lion KT88 reissue is a copy of the original.  Due to use of a superior plate material, the plate dissipation has been uprated to 42 watts."

He also sent a pdf of the original tube's 1959 datasheet. If you want it PM me, I can't upload it to the gallery as pdf.

Freo-1

Re: KT88 Bias Troubles
« Reply #14 on: 16 Jun 2008, 09:08 pm »
Quote
The other thing to try is put the older KT88s in, say your Svet (or JJs) KT88s and see if they work out ok. 

I've done that, and both JJ and Svet work fine, as does the GE 6550A and SED 6550C.


Quote
Were the GL KT88s you had tried all bought together at the same time (thus implying they might have been from same lot)?  Or did you buy some at one point and some later?

The bias setup is a fixed bias setup, with each tube having it's own pot, and the meter is a NOS Simpson ammeter (which is pretty accurate).  The ones that shorted out were from the same lot (and appeared to be from a potentially bad batch).  The set that had bias runaway issues were made somewhat later.  The tubes checked out fine after the incidents (except for the fact it was prematurely worn from going cherry).


Quote
Reports I've read said that the GL's were really reliable tubes.  But sometimes you get a bad lot, it can happen.  However, it seems like you might be rushing to conclusion.

One doesn't really need to see a spec sheet on a tube, but they are provided by some manufacturers.  These just provide a suggested operating point, usually the one most commonly noted in tube manuals, etc.  But a KT88 is a KT88 and by definition is intended to be o


I've also seen good reports overall on the tube.  That's why I went back and tried it a couple more times.  While I had no more shorts, I did have bias runaway (on a couple of tubes after operating for a period of time, from a few to over 100 hours).  The only reasonable conclusions I can reach only applies to my amplifier (especially since all other tube types I've tried do not behave badly). 

As far as the comments about the spec sheet, I would in general tend to agree. However, this situation is unique, and since all the other tubes I've tried all work in circuit, I wanted to just see if something would jump out. Maybe a KT88 is a KT88 except....? 


I am appreciative of people's feedback on this.  I'll see if I can get a copy of the schematic from the tech that did the work. I'm sure it's fine, as he has done many amplifiers over the years.  If I  can't get it in short order, then I'll draw it out and post it. 

Lastly, I wonder if triode vs pentode output plays into this.  A lot of the folks who rave about the tube are running with triode output.  My gut instinct tends to think if I was running triode, would not be any problems. 


JoshK

Re: KT88 Bias Troubles
« Reply #15 on: 16 Jun 2008, 09:15 pm »
I don't think running it in triode would help any, but I could be wrong.  If your tubes were gassy for some reason and having a fixed bias, that could lead to runaway and going cherry or self-destruction.  A smaller grid resistor will help with gassy tubes, the trade-off is a bit of gain loss (usually small and tolerable) from the front end.