Blown 1 OHM current-sense resistors

[Gaz]

Stan, Enzo, thanks once more for the thoughts. I never bothered to look at curve for fuses, that's really interesting! It's much easier to copy known working values for the most part.

Haha, you all must think I'm an idiot, but I'm swear to you I have never once thought the resistors were the actual problem I know at least that much! My concern was that I underrated them, so that when the fault did occur, they weren't able to handle the current long enough for the HT fuse to pop. Because of my lack of real electronic knowledge I'm having trouble figuring out in theory how large they'd need to be to for their be a likelyhood, for say, my 1.5A Slo-blo fuse to go. If someone were to explain to me that with a short like that even a 10W would go first, then I'm willing to accept that Maybe a better questions is if there is a heater to cathode short in a 100W amp how much current might appear at the cathodes opposed to the AC side of the rectifier?

With that said, of course I'm still interested in the root of the problem, but the whole thing really piqued my interest on the topic of the cathode resistors and 'protective' diode found in some more modern amps.

Thanks again! I'm learning here!

[Steve Conner]

I think the key to this is to consider the energy stored in the B+ filter capacitors.

If you shorted the B+ with a 1 ohm resistor and then wound the line voltage up with a variac, a fuse would surely pop before the resistor burnt up.

But if you charged the B+ up to its full 400V (say) and then dabbed a 1 ohm resistor across it, the charged filter capacitors would briefly deliver something like 400 amps. A 1/4 watt resistor would find itself dissipating 160 kilowatts and would probably explode instantly.

This is essentially what can happen when a tube arcs, although the impedance of the output transformer is in the way and reduces the peak current quite a bit.

The humdinger pot will reduce the available current from a heater-cathode short unless it happens to be turned fully to one side.

[Gaz]

I think the key to this is to consider the energy stored in the B+ filter capacitors.

If you shorted the B+ with a 1 ohm resistor and then wound the line voltage up with a variac, a fuse would surely pop before the resistor burnt up.

But if you charged the B+ up to its full 400V (say) and then dabbed a 1 ohm resistor across it, the charged filter capacitors would briefly deliver something like 400 amps. A 1/4 watt resistor would probably explode instantly.

This is essentially what can happen when a tube arcs, although the impedance of the output transformer is in the way and reduces the peak current quite a bit.

The humdinger pot will reduce the available current from a heater-cathode short unless it happens to be turned fully to one side.

Steve, that's a really helpful example, and I think I can imagine more clearly now. So if a tube shorts, it sees the 1 Ohm resistor as the fast way to ground and tries to dissipate all the power through the resistor, correct? And that's why it blows, correct? And on the B+ side it has to pull the current through OT, and has to do it for enough time for the fuse to actually blow (as Enzo pointed out). Right?

If I've got my head wrapped around that much, do you think it's futile to try and size the cathode resistors large enough and/or add the diode across to keep them from blowing in this kind of fault, in hopes that it gives the B+ fuse enough time to go? Or is that a kind of all bets are off kind of scenario?

[Enzo]

Just an opinion, but I think you are focused laser-like on this one part, and trying to ensure that this one part never fails, no matter what. And really, now many of them have burnt up on you? I fix amps for a living, and I don;t see many burnt up 1 ohm resistors at all.


DIodes are like 2 cents apiece. Add them. We have two things here, we have an academic discussion of how things work, and that is fine. But we also have the practical discussion of making an amp work. In the practical arena, beef it up all you want. In the academic arena, we can tackle all the ins and outs.

For example, you did some math for a 1N5408 instead of 1N4007. And I am not sure you got it. The 1A 1N4007 has about a half volt drop, and so does the 3A 1N5408. SO a beefier diode won't protect the resistor any different, the same voltage will be across it. It protects itself maybe better than a 1A diode. But as Steve pointed out - I think - the 1N4007 has a 30A surge rating anyway, so I don't fear for its life.

[Gaz]

Just an opinion, but I think you are focused laser-like on this one part, and trying to ensure that this one part never fails, no matter what. And really, now many of them have burnt up on you? I fix amps for a living, and I don;t see many burnt up 1 ohm resistors at all.


Diodes are like 2 cents apiece. Add them. We have two things here, we have an academic discussion of how things work, and that is fine. But we also have the practical discussion of making an amp work. In the practical arena, beef it up all you want. In the academic arena, we can tackle all the ins and outs.

For example, you did some math for a 1N5408 instead of 1N4007. And I am not sure you got it. The 1A 1N4007 has about a half volt drop, and so does the 3A 1N5408. SO a beefier diode won't protect the resistor any different, the same voltage will be across it. It protects itself maybe better than a 1A diode. But as Steve pointed out - I think - the 1N4007 has a 30A surge rating anyway, so I don't fear for its life.

You're totally right, I am just focusing on this one part. It was the subject of the thread after all, and I hope I'm trying to show that I'm trying my best to wrap my head around the concepts with limited electronics knowledge. I'm not trying to make sure the part "'never fails", I'm still just trying to understand that if that's even a worthwhile effort considering the types of failures that would cause them go in the first place. I really don't want to go over-sizing parts for no reason, and would like to understand why I do so.

I understand the diodes will turn on at the same time, however I thought maybe the 1N5408 may be a better choice because of its higher amperage, so that if there is a really high fault current it won't die itself. From your comment, it appears as though I clearly don't understand the surge ratings, or still what's exactly happening when there's a fault. But see, now I want to know what exactly qualifies as a "surge", etc.

More often than not, I leave the Ampage with more questions than answers, and I think that good in general. There's a high level knowledge and experience here that goes over my head. I actually intentionally post in the maintenance section in hopes that I would get some more basic answers, as discussion in the Theory section quickly escalate over my head. Trust me, I am very aware of my naivete and apparent density on this very subject, but I still swear that I don't think the resistors caused the problem or that making them larger will solve the conflict in Syria

But seriously, thanks again for the heroic effort to educate the masses!

[g1]

Just thought I'd mention that some amps use fuses for the power tube cathodes. An example being JCM900 Marshall.
It also has a LED to show you the fuse is blown .
jcm900-mkiii-poweramp.pdf

[R.G.]

I understand the diodes will turn on at the same time, however I thought maybe the 1N5408 may be a better choice because of its higher amperage, so that if there is a really high fault current it won't die itself. From your comment, it appears as though I clearly don't understand the surge ratings, or still what's exactly happening when there's a fault. But see, now I want to know what exactly qualifies as a "surge", etc.

It's kind of tied up in that I²t thing I mentioned. If a diode is rated for 1A or 6A or whatever, that says "you can put this much current through it semi-forever and it won't burn up".

But in fact, "burning up" takes some time.

This is just like a pan of water. How long does it take to boil a pan of water? Depends on how much water is in the pan, how big the pan is and fast you can get heat into the pan. A nominal pan of water may NEVER boil if you put a flame the size of a lit match under it. The pan and water surface can get rid of that amount of heat without the water temperature ever getting to boiling. So from zero flame up to some larger size flame, the water never boils. For larger flames, the water boils and the time to boil depends on how much bigger the flame is. You can zap the water with a megawatt laser for a microsecond and it won't boil the pot.

If I put 10A through a 1A diode, will it burn up? Probably, if I force it to eat 10A constantly. The current makes it heat up, and the temperature ramps up from room temperature to burnout (which is generally about 125C to 150C for silicon junctions). But the ramp up to burnout temperature takes some time. So if you put 10A through a 1A diode for 1 microsecond, the heat generated doesn't have time to heat up the silicon, leads, body, etc. to a high enough temperature to burn it out. 10A in a 1A rated diode generates heat at some rate. It takes that heat some time to heat the junction to damage. So any time less than the burnout time is within the surge rating of the diode for a 10A pulse.

A "surge" is "a great big pulse, bigger than the thing could normally stand". A "surge rating" is "any surge less than this amount of energy probably won't hurt it".

Diodes are typically surge rated for a half-sine wave of some peak value. So a diode with a "100A" surge rating probably has weasel words on the datasheet saying that it will take a half-sine wave of 100A peak for 8.33mS (that is, half a 60Hz sine wave). Or if they're rated for pulse duty, they're generally rated in terms of I-squared-T.

More often than not, I leave the Ampage with more questions than answers, and I think that good in general.

We're all over our heads about something. Your approach is a very good one.

[Gaz]

It's kind of tied up in that I²t thing I mentioned. If a diode is rated for 1A or 6A or whatever, that says "you can put this much current through it semi-forever and it won't burn up".

But in fact, "burning up" takes some time.

This is just like a pan of water. How long does it take to boil a pan of water? Depends on how much water is in the pan, how big the pan is and fast you can get heat into the pan. A nominal pan of water may NEVER boil if you put a flame the size of a lit match under it. The pan and water surface can get rid of that amount of heat without the water temperature ever getting to boiling. So from zero flame up to some larger size flame, the water never boils. For larger flames, the water boils and the time to boil depends on how much bigger the flame is. You can zap the water with a megawatt laser for a microsecond and it won't boil the pot.

If I put 10A through a 1A diode, will it burn up? Probably, if I force it to eat 10A constantly. The current makes it heat up, and the temperature ramps up from room temperature to burnout (which is generally about 125C to 150C for silicon junctions). But the ramp up to burnout temperature takes some time. So if you put 10A through a 1A diode for 1 microsecond, the heat generated doesn't have time to heat up the silicon, leads, body, etc. to a high enough temperature to burn it out. 10A in a 1A rated diode generates heat at some rate. It takes that heat some time to heat the junction to damage. So any time less than the burnout time is within the surge rating of the diode for a 10A pulse.

A "surge" is "a great big pulse, bigger than the thing could normally stand". A "surge rating" is "any surge less than this amount of energy probably won't hurt it".

Diodes are typically surge rated for a half-sine wave of some peak value. So a diode with a "100A" surge rating probably has weasel words on the datasheet saying that it will take a half-sine wave of 100A peak for 8.33mS (that is, half a 60Hz sine wave). Or if they're rated for pulse duty, they're generally rated in terms of I-squared-T.


We're all over our heads about something. Your approach is a very good one.

Thanks, RG, that's a really helpful explanation. It was confusing to me to see some Fenders that had both 1W/1R resistors (seemingly good for 1W dissipation), and 1A rated IN4007 diodes. It makes more sense now because the diodes can handle a much bigger surge.

Another thing that's confusing to a novice like myself is trying to understand what might actually qualify as a surge. I know it depends on the fault, of course, but say a tube short - that could last for a few seconds.

[Gaz]

Thinking more about this whole thing... Would it be possible to add fuses inline with the resistors? What I'm wondering is if the resistor would affect the operation of the fuse. I think the only requirement would be that it would be sufficiently larger, so that it did not blow at the same time as the fuse, but I'm not certain if there's some interplay between the two.

Thanks.

[Jazz P Bass]

Sure you can fuse it.
The question to me would be "what is the benefit?"
You will never be able to protect every component in a tube amp against every type of failure.
In that case, you have to accept certain failure modes taking out components.

[Gaz]

Sure you can fuse it.
The question to me would be "what is the benefit?"
You will never be able to protect every component in a tube amp against every type of failure.
In that case, you have to accept certain failure modes taking out components.

I asked a legitimate technical question, and you said 'sure'. I know I can put a fuse there, but that wasn't my question. Y

You are totally straw man-ing me. I never said I was trying to protect every component, I'm only try to figure out the smartest way to prevent a serious and expensive failure when a tube shorts without relying on resistors as fuses. Do you think that I should take all the fuse out of the amp then? What's the point right?

If you're going to give a 'old wise amp builder' answer, then please keep it to yourself, because I know you have more valuable and less snarky ways of putting it.

[g1]

I think you are taking JazzP's response the wrong way. It seemed to me a technical answer with reasons behind it.
It didn't seem to me there was any malice intended.

[Leo_Gnardo]

At the risk of being called "another snarky old amp fixer" (worries me none) your 1 ohm resistors seem to be acting perfectly well as fuses already. Stock up on some of those 3W Dale 1% wirewounds and never worry about it again, why not?

FWIW the Scandinavian electric safety authority (SEMKO?) has been demanding fuses on every dam' thing for many years now. What's the point of fusing filament lines & bias supplies anyway? They say do it, so manufacturers who send gear there do it, although it doesn't make any sense.

Now let's all make up & be friendly. It's not worth raising the blood pressure over it.

[Gaz]

I think you are taking JazzP's response the wrong way. It seemed to me a technical answer with reasons behind it.
It didn't seem to me there was any malice intended.

To be frank, I just feel like it's a common attitude in any hobby like this, and I see it all over amp message boards. It's a euphemistic way of saying: 'Sure do what you want, but I know more than you, and I'm gonna tell you to stop freaking out, but not actually answer your question.' I didn't call anyone an "old-man amp fixer", I just said, that's the tone he took when I asked a specific question. I respect Jazz P Bass a lot for all the free advice he gives out on this forum, and I understand he was perhaps trying to be helpful, but I was simply annoyed by his response because he went out of his way to join the 'stop worrying about it, dude' bandwagon.

I've mentioned in this thread many times I acknowledge that I'm a neophyte in many regards when it comes to electronics, but I try to be specific how I word my questions. I just wanna try something new and learn something. I get that over-fusing can be bad and it's not everyone here's cup of tea, but that doesn't mean it's not worthwhile, or that my question should be ignored. Maybe this should have been moved to the theory section long ago, so people stop thinking I'm just trying to fix something that broke.

My green question still remains: If I were to, hypothetically, add fuses between the current-sense resistors and ground, should I anticipate any interplay between them when there is a surge condition? Do the resistor and fuse share the current somehow? What I mean, is will the 1R resistor limit the current before it 'gets' to the fuse?

Thanks.

[frus]

No, the same current flows through all elements that are in series simultaneously.
Btw, it seems to me that fusing the cathodes is just another way of fusing B+, but only from power tubes faults. Does your amp have the B+ fuse?