I have a few resistors that burnt out on my cheapie tube amp
three 820Ω
one 560Ω
three 820Ω
one 560Ω
Obviously I want to replace them and not have to worry about it later on. Coincidentally the KT88 that was connected to those burnt out, but I don't know which occurred first. I ordered replacement KT88s and when that didn't work that's when I knew I had to dig deeper.
My questions:
1. What wattage would be appropriate? They are about ½" long (~13mm) long each. (the 560Ω is just a little bit shorter). I want to replace them just this once and not have to worry about them ever again.
2. What type of resistor would be an "upgrade" ? I believe these are standard metal-film resistors?
Thanks for your help
It seems to be 2W units. As the values are low, they may be in a cathode or screen circuits.
In any case, when new resistors and tubes ar at hand, prior to fire the amp, make use of a current limited power supply for it. You can do it inserting a filament lamp (not fluorescent nor led) in series to the live pole of the amp. At turning amp on light must start as a flash as heaters are cold and psu caps discharged. As tubes take temperature and start to drain current, the lamp will again star to bright im more stabe way. If something goes wrong, quickly lamp brights and then limits the possible of risk to perts. Lamp must be in the order of 1 to 2 times greater than the normal power drained by amp and the no load of the amp is suggested. Except power pentode output stage with no nfb in which case the secondary must be shorted, or still better, loaded with a low voltage lamp like headlights of a car. No use the speaker load when finding the problem.
In any case, when new resistors and tubes ar at hand, prior to fire the amp, make use of a current limited power supply for it. You can do it inserting a filament lamp (not fluorescent nor led) in series to the live pole of the amp. At turning amp on light must start as a flash as heaters are cold and psu caps discharged. As tubes take temperature and start to drain current, the lamp will again star to bright im more stabe way. If something goes wrong, quickly lamp brights and then limits the possible of risk to perts. Lamp must be in the order of 1 to 2 times greater than the normal power drained by amp and the no load of the amp is suggested. Except power pentode output stage with no nfb in which case the secondary must be shorted, or still better, loaded with a low voltage lamp like headlights of a car. No use the speaker load when finding the problem.
To the replacement tubes? Nothing. I simply noted that the channel where the old KT88 burnt out still didn't work, so I turned it off.
So one channel works OK?
Are all those resistors from just one channel?
Perhaps one or more output tubes went into thermal run-away.
Or perhaps a hot screen melted and shorted.
Or . . .
Is it a design that pushes the tubes to their limits, then one or more output tubes gives up?
Can you post a complete schematic, you might get better guesses from us.
Are all those resistors from just one channel?
Perhaps one or more output tubes went into thermal run-away.
Or perhaps a hot screen melted and shorted.
Or . . .
Is it a design that pushes the tubes to their limits, then one or more output tubes gives up?
Can you post a complete schematic, you might get better guesses from us.
Last edited:
For the nex time, NEVER put a new tube in circuit where the previous one has been destroyed. Although you have a large quantity of them and can spent them.
It uould be intetsting to know the burnt parts, in which part of the circuit were wired. As the burnt parts "speak", learn to "listen" them is important way to know what happent.
It uould be intetsting to know the burnt parts, in which part of the circuit were wired. As the burnt parts "speak", learn to "listen" them is important way to know what happent.
I'll post pics up of the circuit board. No schematic. I don't know if it is being pushed to its limits, only 15 WPC. Oldchen K1
Thanks. So at least 3W for the 820 resistors. What about the smaller 560 resistors? Also, what type should I be looking for in the replacements?
Looking at that schematic Vk is 24v and there's three 840r's in parallel, Rk is given as 270r which means Ik is 88mA. So each R sees 29mA @ 24v therefore that's 0.696w, so a 3w resistor should be more than adequate. Why they used 3 840r resistors is a mystery, one 5w 270r would suffice.
Why did they burn out? Three 3w resistors all burning out indicates a gross fault. No idea what those other resistors are, couldn't find a 560r on the schematic. Try to buy good quality parts, Vishay do some nice green 3w resistors which are pretty bomb proof.
It's a good idea to power up an amp through a lamp limiter after any work has been done, even better use a variac and LL.
Why did they burn out? Three 3w resistors all burning out indicates a gross fault. No idea what those other resistors are, couldn't find a 560r on the schematic. Try to buy good quality parts, Vishay do some nice green 3w resistors which are pretty bomb proof.
It's a good idea to power up an amp through a lamp limiter after any work has been done, even better use a variac and LL.
I suggest you replace like for like. Metal film 3W for the 820 and 2W for the 560.
Almost certainly the KT88 has failed first and destroyed the resistors as a result.
Fitting bigger rated resistors will just move the damage of another KT88 failure to the next weak link, the output transformer... That would be harder to fix.
Infact check the output transformer continuity before going any further, just in case.
Also if there is a cathode bypass capacitor fitted across the 3 x 820 ohm resistors it will be toast too. You should replace it regardless.
Best way to prevent future problems is not to beef up the resistors, rather fit a fuse in the B+ (HT+) supply.
Thanks for the explanation. I have no idea why they burned out. The resistance I mentioned was what I measured after I removed the good ones from the other channel, so perhaps it's just a tolerance thing(?)
I'll see if I can find a single 5W 270.
I would up the power rating a bit on the replacements.
Power resistors are indeed designed to run hot (250 ºC is common), but running them hot is both a fire hazard and a reliability issue. I usually derate the resistors by a factor of 3-4x, so if 1 W is dissipated I choose a 3-4 W resistor type.
Tom
Power resistors are indeed designed to run hot (250 ºC is common), but running them hot is both a fire hazard and a reliability issue. I usually derate the resistors by a factor of 3-4x, so if 1 W is dissipated I choose a 3-4 W resistor type.
Tom
Only 1 channel experienced the problem, so is it safe to say the tube went out first and then took out the resistors?
OscarS,
On the working channel, do those resistors look OK?
If the other channel is in good shape, then for the bad channel . . .
One or more KT88 tubes drew excess current on the blown channel.
Possible causes:
KT88 gross failure; thermal run-away, etc.
Bad solder connection on the KT88 socket and circuit parts; intermittent/open, such as a floating control grid.
Shorted connection in the KT88 socket like an uninsulated wire touching, or on circuitry parts.
Sometimes an unclipped through hole wire on a PCB is very near to the chassis, and stress on the board or vibration on the amp can cause the excess wire to contact the chassis.
On the working channel, do those resistors look OK?
If the other channel is in good shape, then for the bad channel . . .
One or more KT88 tubes drew excess current on the blown channel.
Possible causes:
KT88 gross failure; thermal run-away, etc.
Bad solder connection on the KT88 socket and circuit parts; intermittent/open, such as a floating control grid.
Shorted connection in the KT88 socket like an uninsulated wire touching, or on circuitry parts.
Sometimes an unclipped through hole wire on a PCB is very near to the chassis, and stress on the board or vibration on the amp can cause the excess wire to contact the chassis.
In the picture I posted, you can see the resistors from both channels. The burnt ones belong to Left, and the decent looking ones belong to the Right channel. I removed them all so I could measure the resistance on the good ones from the Right channel, since those weren't burnt up.
From what I could tell, the the solder connections looked good.
So unless otherwise suggested, I'll be ordering some metal film replacements to see if it indeed fixes it all.