• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

KT88SE -burnt cathode resistor

Hi all,

Hoping someone can provide some advice.

For 2yrs my KT88SE amp has been running for around 12hrs a day with no problems.

Today, a distinct burning smell for the first time and having removed the bottom plate, one of the cathode resistors had all the gold colouring removed, and was extremely hot to the touch. (See photo, left hand side). After about half an hour, still extremely hot to the touch. I wanted to mount these on heatsinks but I was a month or two away from doing this.

The amp, other than smelling, was still sounding great, I was baking a loaf of bread (lol) and thought it was that so, after an hour of smelling burning, I realised it wasn't the bread but the amp!

Ambient temp is hot at the moment but would anyone know why this has happened, only on one channel, the other seams fine.

Circuit is the Abdella Blue Glow version of the KT88SE, cathode resistor is 500ohm (10w in circuit, 25w in my amp) parrelled with a black gate, which is quite old now and has been subject to high heat.

Thanks in advance.
 

Attachments

  • 20230626_220629.jpg
    20230626_220629.jpg
    381 KB · Views: 133
Moderator
Joined 2011
If the tube is ok, then the resistor is probably under-rated.
These resistors need a substantial heat sink to meet their spec.
Otherwise you get maybe 20% of spec.
You may have to use the larger size of that package, plus sink.
 
Last edited:
  • Like
Reactions: 1 user
Today, a distinct burning smell for the first time and having removed the bottom plate, one of the cathode resistors had all the gold colouring removed, and was extremely hot to the touch. (See photo, left hand side). After about half an hour, still extremely hot to the touch. I wanted to mount these on heatsinks but I was a month or two away from doing this.
You realise wood isn't a very good heat sink. ;) I don't see any vent holes either.

jeff
 
  • Like
Reactions: 1 users
Thanks for the replies,

I'll order 50w versions, mount them on heatsinks, vented through the top panel. I've made heatsinks, as mentioned, I was about to the use them, have new coupling and bypass caps. Heatsinks are 3nr pieces of 30 x 80 x 3mm thick per resistor which will vent through 35mm diameter openings in the top plate (hope that will be enough)

I'm assuming the valves are OK but wouldn't both resistors be overheating at the same time? Or maybe thats a 'not necessarily'

I shall remember that wood is not the best conductor of heat in future!
 

Attachments

  • 20230626_225328.jpg
    20230626_225328.jpg
    260.5 KB · Views: 52
Do check the tubes separately.
Also use heat sink compound between the resistor and the sink.

Is there a simple way of checking tubes? Having no valve tester. They are SOVTEK tubes that, we're, far from matched when purchased but sounded really good nevertheless.
 
Check the coupling cap too, if it is leaking the bias will be way off and that could cause too much current to flow. That would hurt that resistor and possibly the tube too. Put the tube in the good side and carefully try it to see if it’s not hurt. If it sounds bad or red plates shut down quickly and replace it.
 
  • Like
Reactions: 1 user
Check the coupling cap too, if it is leaking the bias will be way off and that could cause too much current to flow. That would hurt that resistor and possibly the tube too

Thanks for this suggestion, this could be the issue as I believe it is more than just overheating.

Today, the resistor smells terrible, burnt, scorched, hest has stripped the gold colour to silver. The other channel is fine.

To check the caps, I'll remove the bad channel Power Valve (KT88), connect (-) lead to chassis ground and the (+) lead to the grid pin of the valve base pin 5 (or maybe even just after the coupling cap leading toward the 3k series resistor which then leads to pin 5). Switch on.

Looking through my build notes and test measures from 2yrs ago, I had a reading of -40.98v at pin 5 (fixed bias) so if there is considerable movement (i.e. from -40.98 to -15v or so) then this would indicate a leaky cap.

Would this method be acceptable test, please accept my apologies I know to most this will be very basic but I would like to be detailed in trying to find the fault for future readers who, like me, are not very well educated.

In the interim, I've ordered new cathode resistors, will place them on heatsinks, form vent openings in the top plate. Coupling caps and bypass caps I already have and I will get some new KT88's too as my current SOVTEKs have between 4000 - 6000hrs on them.
 
I would check that your output transformer with an ohm meter, and the wiring, socket, and blocking cap haven't caused the problem before swapping tubes around to be sure you're not going to destroy another tube. Yes, check the voltage on the grid pin to confirm the bias setting. I don't understand what you're saying about 10w / 25w for the cathode resistors. I guesstimate a little over 3w dissipation in the cathode resistor (when operating properly), so a 5 or 10 w resistor should be plenty of margin. But I wouldn't mount it directly on the wood.
 
I don't understand what you're saying about 10w / 25w for the cathode resistors. I guesstimate a little over 3w dissipation in the cathode resistor (when operating properly), so a 5 or 10 w resistor should be plenty of margin. But I wouldn't mount it directly on the wood.
The 10W resistor is the rating Blueglow used in his build, I used 25W. I was just confirming that the rating is well above what is required so we could rule out that being the cause of the burning. They will be on Heatsinks when I've confirmed what the fault is and undertaken the necessary mods.

I think both tubes are fine, as stated, the amp was working, at least sounding, fine. It was the smell of burning that made me switch off and then realised it was the cathode resistor that was causing the smell. Other than the smell, there was no indication at all, by sound, that there was a problem. No hissing, pops, snaps, crackles. As always, sounding great.
 
Had a similar issue with my KT88SE after the power went out and quickly came back on. It blew the fuse and damaged the rectifier. Put new rectifier in and amp seemed to work fine then after a few minutes the left channel quit and the magic smoke left the cathode resistor for the channel.
Popped a new resistor in and checked voltages, which were fine(43 v drop across cathode resistor, -43 at grid). Sounded great then after a time pow went the resistor again. New resist and this time put a meter on both cathodes to compare between channels. Both channels were the same but after about 15 minutes the left channel voltage would begin to rise. Long story short, swapped tubes and fault followed tube.
Sure sounds like the same issue. New tubes and amp is back to normal, so monitor cathode voltage for a while, see what you get.
 
  • Like
Reactions: 1 user
Long story short, swapped tubes and fault followed tube.
Sure sounds like the same issue. New tubes and amp is back to normal, so monitor cathode voltage for a while, see what you get.
Great, thank you, another possible solution, so tests are as follows:

1) Take the faulty channel KT88 out and test the coupling cap for any leakage, as described in #11.

2) If that is OK, or even if it is not, test the good channel coupling cap to confirm it is not leaking.

3) if all is well with the caps, take the good channel KT88 out and put the 'faulty' channel KT88 in and watch for voltage rise.

As the faulty channel cathode resistor is burnt, I won't put the KT88 back in, leaving it out. Will the amp be OK like this for 15 minutes or so whilst I keep an eye on the voltage rise of the other channel, with the valve in?
 
Forgot to add:

4) Check values of the Cathode Resistors.

Well, this is the first thing I checked, having taken it out. Should be 500R:

Cathode Resistor good channel - 499R

Cathode Resistor bad channel - 392R

So, first check there is a fault with the resistor value. Maybe this is the cause? Or maybe, a fault elsewhere has caused the resistor to become faulty...
 
Moderator
Joined 2011
The resistor was damaged by the overheating by excessive cathode current,
due to a bad tube or a leaky coupling capacitor.

That type of resistor must be severely derated if used without a proper heat sink.
A 10W rating would become only a few watts, especially in such a confined environment.
A small piece of metal is not really enough.
 
  • Like
Reactions: 1 user
Yeah, your bad channel resistor is out of tolerance, needs replacement. In my troubleshooting, I left both tubes in original position and verified fault....rising voltage. Monitor both channels to compare. Voltage drop should remain constant across both resists. You will notice when the voltage begins to change, so once it starts to fail, you shut down the amp to swap tubes.
Since your one resist is out of spec, you can put it in the other socket. I would replace resistor in bad channel, then test so you can compare voltages. Either way, as long as you monitor voltages and shut down before voltage gets out of control , cathode current won't get too high and cause damage.
Remember, cathode current will remain very stable with normal operation and no input. Any change in voltage is cause for concern,
 
The resistor was damaged by the overheating by excessive cathode current,
due to a bad tube or a leaky coupling capacitor.
Again thank you Rayma, I really do appreciate your and all the others helpful comments in tracking down this Gremlin.

I'll carry out the tests in Post #16 tomorrow (unless anyone has any reservations with them), it is a little late here, would like to be focused and alert.

I do hope you and the others will keep an eye for updates tomorrow...