Now if the 100 Ohm screen resistor burns open, you still do not know for sure the cause of the problem.
Failure to find the cause will only mean you have to replace a part every now and then.
Failure to find the cause will only mean you have to replace a part every now and then.
Most cascading failures will not have a self-evident cause. It has to be deduced and tested.
As in post #10, I would swap the output tubes and wait for the next failure. Then, If the same channel
still has the problem, I would replace all of the coupling capacitors.
As in post #10, I would swap the output tubes and wait for the next failure. Then, If the same channel
still has the problem, I would replace all of the coupling capacitors.
And, I might find a way to stand those EL34 tubes upright.
A little external vibration to the amplifier, goes a long way to make a hot 400V screen short to the control grid. OUCH!
A little external vibration to the amplifier, goes a long way to make a hot 400V screen short to the control grid. OUCH!
Looks like the amp could be stood up on its back panel somehow, or hung from its front panel,
to get the tubes vertical.
to get the tubes vertical.
Do not increase the power rating of the resistor! It’s acting as a fuse and burning out because something else is amiss. This is a good situation. Tubes and power or output transformers are expensive and the next likely things in the chain if the resistor did not fail.
Like others have pointed out, problem is quite likely coupling capacitors from the driver to the output tubes (just replace them as a precaution) , a bad bias pot wiper (you can add a safety resistor to bias the tubes off if the wiper opens), bad solder or intermittent sockets (clean and inspect) , or a bad tube ( possibly now, but also the least likely to cause this. Test and replace if needed).
Like others have pointed out, problem is quite likely coupling capacitors from the driver to the output tubes (just replace them as a precaution) , a bad bias pot wiper (you can add a safety resistor to bias the tubes off if the wiper opens), bad solder or intermittent sockets (clean and inspect) , or a bad tube ( possibly now, but also the least likely to cause this. Test and replace if needed).
And, I might find a way to stand those EL34 tubes upright.
A little external vibration to the amplifier, goes a long way to make a hot 400V screen short to the control grid. OUCH!
That's the way the amp is made. It's a commercially made amp, and the tubes are placed horizontally to get them all inside the case. There's no real way to get the tubes vertical unless I was to build a new case and cannibalize the components.
Is this an inherent design flaw?
I checked out the coupling caps. They are Multicaps, and are describes as self-healing. If developing a bad spot, then heals up, could that explain the 5-year gap between failures?An arcing tube could cause this problem but it would most likely not be a five year interval between resistor failures! You might also see an associated flash from the tube. Look at the coupling cap for the cause. Many types are so called self healing and burn out the bad spot and work fine until the next bad spot. That bad spot is a short that will send voltage straight through to the output tube. That will disrupt the bias and then bang the resistor pops! Measure the voltage on both sides of the cap to ground. Compare it to the same reading on the other coupling caps. If it has healed it will test normal. Change it for safety sake.
My other thought is that it could be a tube failing after several years, which might explain why there's 5 years between failures. The tubes I have in the amp were replaced the last time I had this issue. Is that plausible?
Some guitar amp LS cabinets have the amp tube circuitry in the same cabinet with minimal ventilation and the resonance the tubes suffer is appalling. That practice of one cabinet for both has been around ever since I´ve known them since the 1950´s, so it remains no excuse for failure !A little external vibration to the amplifier, goes a long way to make a hot 400V screen short to the control grid. OUCH!
rj
benchbaron,
Please post all the exact details of each reliable model of guitar LS cabinet tube cabinets, including schematics, voltages, and currents, and the vintage and manufacturer of the EL34 tubes (vibrations are assumed to be high levels).
Sideways EL34 amplifiers are not all equally as reliable as others, some factors are:
Quality of EL34
Angular rotation of the sideways EL34
Plate Voltage and current
Screen Voltage and current
Plate to Plate primary impedance
Pentode mode or Ultra Linear mode (Ultra Linear mode is friendlier, except that the quiescent screen voltage is higher than the plate voltage).
A pentode mode (constant screen voltage) push pull guitar amplifier that is played so that the amplifier puts out a square wave, will cause the screen current to be at Maximum for 50% of the time.
That integrated screen current will far exceed the tube data sheet's maximum screen current rating.
That will heat the screen like you might have never imagined.
Exceed the screen ratings at your own risk.
"All hotter screens are equal, but some hotter screens are more equal than others".
Again, a little vibration might go a long way.
Perhaps you have never heard of a vacuum tube guitar amplifier that was yarded on so hard, that it failed during a gig.
Please post all the exact details of each reliable model of guitar LS cabinet tube cabinets, including schematics, voltages, and currents, and the vintage and manufacturer of the EL34 tubes (vibrations are assumed to be high levels).
Sideways EL34 amplifiers are not all equally as reliable as others, some factors are:
Quality of EL34
Angular rotation of the sideways EL34
Plate Voltage and current
Screen Voltage and current
Plate to Plate primary impedance
Pentode mode or Ultra Linear mode (Ultra Linear mode is friendlier, except that the quiescent screen voltage is higher than the plate voltage).
A pentode mode (constant screen voltage) push pull guitar amplifier that is played so that the amplifier puts out a square wave, will cause the screen current to be at Maximum for 50% of the time.
That integrated screen current will far exceed the tube data sheet's maximum screen current rating.
That will heat the screen like you might have never imagined.
Exceed the screen ratings at your own risk.
"All hotter screens are equal, but some hotter screens are more equal than others".
Again, a little vibration might go a long way.
Perhaps you have never heard of a vacuum tube guitar amplifier that was yarded on so hard, that it failed during a gig.
Last edited:
You´re checking my age.. You didn´t mention or research Combo tube/valve amps with speakers in the same box (put it simply).....Peavey, Marshall, Selmer, are just a few examples of tubes used in savage environments with minimal ventilation. In the 1960´s, it wasn´t uncommon after an hour or two of hard playing to hear the effects of amp heat warping a speaker cone to rub on the magnet, however those times the Combo box was very popular....loads of others still out there.
LS material sciences have improved but no point delving into this. I would add that many tube data sheets which we all use DO mention tubes can be used in any position, even upside down with jock-strap retainers. It´s all in the KT88 data and other sheets, i.e 5881 which could be electrically "hammered". Many of today´s made tubes probably aren´t as mechanically robust as their predecessors that the manufacturers now suffice with electrical tests and less shock tests that the internals may not be as "robust". Expense and opinions of use differ.
Problem solved. You are quite correct over screen ratings.
LS material sciences have improved but no point delving into this. I would add that many tube data sheets which we all use DO mention tubes can be used in any position, even upside down with jock-strap retainers. It´s all in the KT88 data and other sheets, i.e 5881 which could be electrically "hammered". Many of today´s made tubes probably aren´t as mechanically robust as their predecessors that the manufacturers now suffice with electrical tests and less shock tests that the internals may not be as "robust". Expense and opinions of use differ.
Problem solved. You are quite correct over screen ratings.
The resistor only dissipates 16 mW with 0.4 V across it. Even a resistor rated for 125 mW would be fine with this. So the fact that the resistor blows indicates that it sees overload when the amp is not operating normally. I would look for:
Overload during startup and shutdown.
Overload due to a faulty or failing tube.
If the amp works fine once the resistor is replaced, it's probably not a faulty tube. But I'm not an expert on tube failures so maybe there is a failure mode that just shows up periodically.
By overload I mean that the resistor dissipates more power than it's rated for. The pulse rating of resistors is usually only around 2-5x the continuous rating of the resistor. I suppose over-voltage is possible, but most resistors can handle 250 V or so. It seems unlikely that the resistor would see that in this application, but that's something that could be verified with a scope. I'd look during startup/shutdown.
Tom
Overload during startup and shutdown.
Overload due to a faulty or failing tube.
If the amp works fine once the resistor is replaced, it's probably not a faulty tube. But I'm not an expert on tube failures so maybe there is a failure mode that just shows up periodically.
By overload I mean that the resistor dissipates more power than it's rated for. The pulse rating of resistors is usually only around 2-5x the continuous rating of the resistor. I suppose over-voltage is possible, but most resistors can handle 250 V or so. It seems unlikely that the resistor would see that in this application, but that's something that could be verified with a scope. I'd look during startup/shutdown.
Tom
Could the weakness be in the bias supply? Do you have the schematic for that?
If it is slow to stabilise at the negative voltage so the tubes are hot before bias is correct, then that would stress the tubes each time the amp is switched on, which could be indicated by the failed resistor. Are the PSU caps old? They will be pretty toasty in there by the looks of it.
If it is slow to stabilise at the negative voltage so the tubes are hot before bias is correct, then that would stress the tubes each time the amp is switched on, which could be indicated by the failed resistor. Are the PSU caps old? They will be pretty toasty in there by the looks of it.
Let’s think about the screen resistor as fuse. 1 watt is dissipated when that screen resistor is conducting 100 mA continuously. Will a 1 watt resistor burn out when dissipating 1 watt? Probably not. Is it save to to have the EL 34 screen conduction 100+ mA until it burns up the resistor?
Last edited:
Please lets be helpful. Without solid new data, this thread is continually circulating speculation. The original poster needs to do the following logical steps that will correct almost all of the most likely causes of his problem. These are not expensive or difficult to implement. Specific questions regarding any of these actions members of this board will happily help with.
Replace coupling caps with high quality (not boutique) poly caps rated at 600 volts.
Implement bias pot open wiper protection resistors.
Clean tube sockets.
Inspect and correct bad solder joints.
Replace output tubes if they test bad.
Call back in 5 years if problem still exists. Have fun! Its just a 25 cent resistor after all.
Replace coupling caps with high quality (not boutique) poly caps rated at 600 volts.
Implement bias pot open wiper protection resistors.
Clean tube sockets.
Inspect and correct bad solder joints.
Replace output tubes if they test bad.
Call back in 5 years if problem still exists. Have fun! Its just a 25 cent resistor after all.
Thanks for the summary. That's very helpful to have all the information here distilled down like this.Please lets be helpful. Without solid new data, this thread is continually circulating speculation. The original poster needs to do the following logical steps that will correct almost all of the most likely causes of his problem. These are not expensive or difficult to implement. Specific questions regarding any of these actions members of this board will happily help with.
Replace coupling caps with high quality (not boutique) poly caps rated at 600 volts.
Implement bias pot open wiper protection resistors.
Clean tube sockets.
Inspect and correct bad solder joints.
Replace output tubes if they test bad.
One last question. It's been mentioned a few times that the resistors seem "too small" to be 1W resistors. My replacement resistors came today. They are made by Ohmite, and one reason that I selected them is that they are rated for 275ºC, higher than the other options Mouser listed.
They also seem "small" by the standards that the previous comments suggest. But so do the other options that Mouser has, and I assume that they have a good variety of resistors to offer.
Here are the results of filtering for 10 ohm 1W wirewound resistors: https://mou.sr/3qWKGLe None of them seem particularly large.
Is there anyone that thinks that the size of the resistor is not important here, assuming that I can believe the spec sheet?
That's the repair man's perspective. 😉 I'd expect more than five years of service life from an amp between repairs.Call back in 5 years if problem still exists. Have fun! Its just a 25 cent resistor after all.
I agree that there's lots of speculation here. But the resistor blows for a reason. And, more importantly, it blows even though it should not. If it was dissipating 1 W in a 1 W rated part, I'd say it's a marginal design, suggest installing a 3-5 W part, and boxing it up. But that's not the case here. 16 mW in a 1 W part should make that resistor last longer than any of our lifetimes. I can't speak for you but I certainly plan to live longer than five years from now.
I think it would be worthwhile to look for potential design flaws for a little bit before just swapping the resistor and calling it 'fixed'. The startup/shutdown issues I mentioned could easily be measured with a storage oscilloscope (or any digital scope) for example.
Tom
I would believe a data sheet from a reputable manufacturer. So if you can find the resistors on Mouser and their data sheet says they'll dissipate 1 W, then I'd accept that as fact.They also seem "small" by the standards that the previous comments suggest. But so do the other options that Mouser has, and I assume that they have a good variety of resistors to offer.
Here are the results of filtering for 10 ohm 1W wirewound resistors: https://mou.sr/3qWKGLe None of them seem particularly large.
Is there anyone that thinks that the size of the resistor is not important here, assuming that I can believe the spec sheet?
Do keep in mind that the dissipation limit is usually given assuming that the resistor sits in 25 ºC air with plenty of breathing room for convection. A tube amp will be pretty warm inside and might not offer much air flow. It may be at 40-50 ºC with hotspots here and there. If you look at the power dissipation derating curve in the data sheet you'll likely find that the resistor can only dissipate, maybe, 0.5-0.75 W under those conditions.
As you've found out some of the newer resistors are pretty small and are rated for rather high temperatures (250-300 ºC). I've made the same observation. A newer 1 W resistor is the size of an older 250 mW type. A 3 W type is smaller than a 1 W used to be. I think that's due to newer, higher temperature, materials.
Note that this also means that at 1 W dissipated in a 1 W resistor rated for 300 ºC will result in a resistor that's 300 ºC with 25 ºC ambient air! That' almost as hot as my soldering iron!! This is why I typically derate the power rating by 3-5x. So if math says the resistor should dissipate 1 W, I'll choose a 3-5 W type for that location. Then the resistor typically "only" gets to about 100 ºC.
16 mW * 5 = 80 mW. So still low enough that an itty-bitty 3.2 mm long 125 mW resistor type would be plenty. So if it's frying there must be something else going on.
Tom
As do I!That's the repair man's perspective. 😉 I'd expect more than five years of service life from an amp between repairs.
I hope you agree that my recommendations do address likely root causes. It would be foolish to just replace resistor and hope. If there are still problems, and I don't think that's likely, digging into less likely causes will be necessary. Also fun for me as I greatly enjoy a tough problem to solve.
I agree with most of da gurus here that if it is the same resistor that blows and not the others, there is a 'fault' that needs to be found. But my $0.02 is as follows.
- This is a Class B amp. So the resistor will see a LOT more current under use, especially a guitar amp which might be often overloaded.
- Modern resistors can have high power ratings in a small package but they do this by running HOT. This is in immutable law of nature. So choose the LARGEST PHYSICAL SIZE resistor that will fit and make sure you mount it properly ... off the PCB, perhaps on standoffs. Think of emitter resistors on Class B transistor amps.
- Home
- Amplifiers
- Tubes / Valves
- Resistor used for measuring bias keeps failing -- what can I do?