Would this work?
An externally hosted image should be here but it was not working when we last tested it.
Catching fire?
Amplifiers have many different failure modes.
It is a good idea to not have fires in any amp, but to protect against some of the causes:
All my tube amps have a fast blow fuse wired in series with a slow blow fuse (on the primary circuit of the power transformer).
The fast blow amp rating is larger than the slow blow amp rating.
The fast blow is the smallest rating that does not blow during power-up inrush current.
The current can vary depending on exactly when in the power line cycle you close the mechanical on switch (with a dead cold amp).
The slow blow is the smallest rating that does not blow during warmed-up continuous current, and loudest music played.
Of course, the amp needs to be designed so that no component is pushed beyond its rating during any of the above.
Amplifiers have many different failure modes.
It is a good idea to not have fires in any amp, but to protect against some of the causes:
All my tube amps have a fast blow fuse wired in series with a slow blow fuse (on the primary circuit of the power transformer).
The fast blow amp rating is larger than the slow blow amp rating.
The fast blow is the smallest rating that does not blow during power-up inrush current.
The current can vary depending on exactly when in the power line cycle you close the mechanical on switch (with a dead cold amp).
The slow blow is the smallest rating that does not blow during warmed-up continuous current, and loudest music played.
Of course, the amp needs to be designed so that no component is pushed beyond its rating during any of the above.
Last edited:
This thread started as a tube output protection question.
But since it has now included the amp catching fire, I thought a little
protection of the amp from a power transformer fire might be in order
(and the fusing method I listed may also protect against some output tubes or output transformers that short, or even catch fire).
What secondary are you wanting to fuse?
B+ 1, Center tap, B+ 2, Filament?
Output transformer Secondary?
But since it has now included the amp catching fire, I thought a little
protection of the amp from a power transformer fire might be in order
(and the fusing method I listed may also protect against some output tubes or output transformers that short, or even catch fire).
What secondary are you wanting to fuse?
B+ 1, Center tap, B+ 2, Filament?
Output transformer Secondary?
Power transformer primaries short, power transformers secondaries short, output tubes short, electrolytics short, fixed bias supplies go dead (or the adjustment pot wiper opens) output transformers short (I once had an output transformer that had a short from primary to the laminations and end bells . . . to ground).
Of course you could fuse the secondary. But the problems would be multiple.
1. You could fuse the two B+ taps
2. You could fuse the B+ center tap
Caution:
Both of these would require a very high voltage fuse, not the usual 250V fuse.
And high voltage fuses generally do not come in as many current values as the 250V fuses. And it may be harder to find the type you want (fast blow, slow blow).
3. If the power transformer shorted (any of the transformer windings) fusing the secondary or secondaries would not protect the transformer.
That is why I do it the way I do, I fuse the primary.
Of course you could fuse the secondary. But the problems would be multiple.
1. You could fuse the two B+ taps
2. You could fuse the B+ center tap
Caution:
Both of these would require a very high voltage fuse, not the usual 250V fuse.
And high voltage fuses generally do not come in as many current values as the 250V fuses. And it may be harder to find the type you want (fast blow, slow blow).
3. If the power transformer shorted (any of the transformer windings) fusing the secondary or secondaries would not protect the transformer.
That is why I do it the way I do, I fuse the primary.
Overcurrent protection does little to protect devices if the root cause is overvoltage. They only prevent the fire after the fact - the output transformer is still ruined if you burn thru the insulation. Even if the damage is localized and doesn't bring the fire department or set off the smoke alarm.
And I'll bet good money that the flyback spikes produced by an under-loaded OPT sound better (as in more desirable harmonic structure) compared to just clipping them off with proper free-wheeling diodes. That kind of distortion in general usually sounds "buzzy" or "transistor-like".
And I'll bet good money that the flyback spikes produced by an under-loaded OPT sound better (as in more desirable harmonic structure) compared to just clipping them off with proper free-wheeling diodes. That kind of distortion in general usually sounds "buzzy" or "transistor-like".
Let's see if this link works
An externally hosted image should be here but it was not working when we last tested it.
I always try to protect the output transformer, not the tubes.
I have seen dramatic OT failure (Marantz 8b), with primary to chassis short.
Took down the choke and damaged the PT, primary fuse went off, too late.
A fuse on the B+ would have saved the choke and the PT.
I have seen a bad driver tube on a McIntosh MC240 with a beautiful purple tube arc. The internal fuse on the HV supply blew. No damage to anything.
I have seen 2 runaway EL84s with loss of bias. Smoked the cathode resistor, B+ fuse did not blew as it was sized to protect the 2 channels: mistake. No damage as the user was quick to shut off the amp.
Since then I always add a fuse either on the HV center tap if any, or before the voltage doubler or on the B+ supply rail before the rectifiers. Here I use a slow blow 1 or 2 amp.
It does help but not if you only have 1 runaway tube melting away the OT primary winding it is serving, unfortunately.
For that reason, I also like the cathode fuses option, to protect each side of the primary OT by choosing the right fuse.
I found that 3x the idle current fast blow will hold. Less and you might blow the fuse for no real reason at full blast volume. I also usually add serial to the fuse a 1R sensing resistor for bias adjustment and balancing.
I am aware of the non linearity of fuse with temperature, and it's resistance is far from zero ohm, but I can not honestly say it does affect the sound.
There is no silver bullet. Even the new MC275 with active monitoring sometimes has failure.
It's all about balancing the odd and know what you are protecting and for what condition, without affecting the sound.
I have seen dramatic OT failure (Marantz 8b), with primary to chassis short.
Took down the choke and damaged the PT, primary fuse went off, too late.
A fuse on the B+ would have saved the choke and the PT.
I have seen a bad driver tube on a McIntosh MC240 with a beautiful purple tube arc. The internal fuse on the HV supply blew. No damage to anything.
I have seen 2 runaway EL84s with loss of bias. Smoked the cathode resistor, B+ fuse did not blew as it was sized to protect the 2 channels: mistake. No damage as the user was quick to shut off the amp.
Since then I always add a fuse either on the HV center tap if any, or before the voltage doubler or on the B+ supply rail before the rectifiers. Here I use a slow blow 1 or 2 amp.
It does help but not if you only have 1 runaway tube melting away the OT primary winding it is serving, unfortunately.
For that reason, I also like the cathode fuses option, to protect each side of the primary OT by choosing the right fuse.
I found that 3x the idle current fast blow will hold. Less and you might blow the fuse for no real reason at full blast volume. I also usually add serial to the fuse a 1R sensing resistor for bias adjustment and balancing.
I am aware of the non linearity of fuse with temperature, and it's resistance is far from zero ohm, but I can not honestly say it does affect the sound.
There is no silver bullet. Even the new MC275 with active monitoring sometimes has failure.
It's all about balancing the odd and know what you are protecting and for what condition, without affecting the sound.
A simple resistor across the output will damp the output transformer if there is no load. It doesn't have to be 8 ohms even a 100 ohms will do.
The requirement for a fuse with a voltage rating higher than 250VAC is not the advise provided by fuse manufacturers for secondary side protection.
A 250VAC rating applies to coping adequately with a high prospective fault current that can occur on the primary side. Fuse manufacturers advise that higher secondary winding voltages can be managed by a 250VAC fuse as the prospective fault current is not as high due to winding resistances and impedances. I'd suggest that unless your amp is seriously high B+, then standard 250VAC fusing is fine.
If you're not limited to only buying typical north American UL248-14 style fuses, then the IEC 60127-2 rated fuses provide much better definition of performance, which can be better matched to the particular amp you have if you are keen to apply an appropriately sized and style of fuse.
I had hoped that all who read my post # 86 would read All of the post.
I did not recommend doing secondary fusing, but said you Could do that, and leaving the research up to an educated person who was willing to do the research on what fuses might possibly work safely. And then I mentioned there were other safety concerns (such as a power transformer short).
Suppose you fuse the secondary center tap, and the (sometimes twisted) wiring to the rectifier plates shorts, or the rectifier socket shorts, or the rectifier shorts, the center tap fuse will protect nothing.
And fusing all the secondary leads will not protect all contingencies.
That is why I closed the post with the one method I always use:
Fuse the primary with a fast blow and slow blow in series.
I have 120VAC power, but always use 250V fuses.
I use all the components in the amplifier conservatively versus their maximum ratings.
power transformers, chokes, output transformers, capacitors, resistors, tubes, etc.
I have worked with UL on safety issues (and there are many that you might not suspect).
I have even made recommendations to UL on proper measurement setups for them to get accurate and repeatable test results.
My power source is Hot, Neutral, and Ground.
The amplifier power wiring is IEC socket Hot to fast blow fuse, slow blow fuse, on-off switch, transformer primary lead 1, transformer primary lead 2, IEC socket Neutral.
IEC ground to amp chassis (ground).
Simple, but the shortest path is to the first fuse. (no pun intended)
I did not recommend doing secondary fusing, but said you Could do that, and leaving the research up to an educated person who was willing to do the research on what fuses might possibly work safely. And then I mentioned there were other safety concerns (such as a power transformer short).
Suppose you fuse the secondary center tap, and the (sometimes twisted) wiring to the rectifier plates shorts, or the rectifier socket shorts, or the rectifier shorts, the center tap fuse will protect nothing.
And fusing all the secondary leads will not protect all contingencies.
That is why I closed the post with the one method I always use:
Fuse the primary with a fast blow and slow blow in series.
I have 120VAC power, but always use 250V fuses.
I use all the components in the amplifier conservatively versus their maximum ratings.
power transformers, chokes, output transformers, capacitors, resistors, tubes, etc.
I have worked with UL on safety issues (and there are many that you might not suspect).
I have even made recommendations to UL on proper measurement setups for them to get accurate and repeatable test results.
My power source is Hot, Neutral, and Ground.
The amplifier power wiring is IEC socket Hot to fast blow fuse, slow blow fuse, on-off switch, transformer primary lead 1, transformer primary lead 2, IEC socket Neutral.
IEC ground to amp chassis (ground).
Simple, but the shortest path is to the first fuse. (no pun intended)
No probs - I was just clarifying the voltage rating of a fuse if it were to be used on the secondary side, as that is a common enough practise by many and the issue often arises as to how to fuse a winding that has a higher VAC than the fuse VAC rating.
Fusing beyond just the primary can certainly provide better discrimination for faults within the circuitry of amplifiers, depending on where the fault is. I would certainly recommend diyers to carefully consider if secondary side fusing can be advantageous, over and above appropriate primary side fusing. The linked article goes some way to helping diyers go though the technical hoops needed to make a good choice of fuse size and type.
https://www.dalmura.com.au/static/Valve%20amp%20fusing.pdf
Fusing beyond just the primary can certainly provide better discrimination for faults within the circuitry of amplifiers, depending on where the fault is. I would certainly recommend diyers to carefully consider if secondary side fusing can be advantageous, over and above appropriate primary side fusing. The linked article goes some way to helping diyers go though the technical hoops needed to make a good choice of fuse size and type.
https://www.dalmura.com.au/static/Valve%20amp%20fusing.pdf
I have fused output tube plates, and it can be done OK.
But then you have to consider the implications:
You should not use a common self bias resistor (not when you fuse plates).
(and for good matching over time, should use separate self bias anyway).
With separate self bias, and one plate fused open, and the consequential B+ rise, you have to watch where the other tube will self bias, versus its bypass capacitor voltage rating.
You have to consider once the output tube plate fuse blows, or multi plates fuses blows, will the B+ rise to a high value that will be beyond the B+ electrolytics voltage ratings, not just the first one, but all the way down the CLC R C R C B+ chain.
You can fuse the output transformer primary center tap (but again, you have to watch the B+ voltage rise, and electrolytic voltage ratings).
You can fuse the output tubes cathodes, but then you have to watch the filament to cathode voltage ratings; and depending where you connect the fuse, you have to watch the self bias capacitor voltage ratings, etc.
I remember a Lead Acid Gel Cell in a metal pack, that connected via a wire to a voltage invertor in another metal pack. It was CSA approved.
But there was no fuse leading from the Gel Cell to its output connector. The invertor had a shorted RF bypass capacitor at its input connector. When the two were connected, the wire connecting the two packs smoked.
The product offering was quickly modified to include a fuse in the gel cell pack.
We all make mistakes of one kind or another.
etc.
But then you have to consider the implications:
You should not use a common self bias resistor (not when you fuse plates).
(and for good matching over time, should use separate self bias anyway).
With separate self bias, and one plate fused open, and the consequential B+ rise, you have to watch where the other tube will self bias, versus its bypass capacitor voltage rating.
You have to consider once the output tube plate fuse blows, or multi plates fuses blows, will the B+ rise to a high value that will be beyond the B+ electrolytics voltage ratings, not just the first one, but all the way down the CLC R C R C B+ chain.
You can fuse the output transformer primary center tap (but again, you have to watch the B+ voltage rise, and electrolytic voltage ratings).
You can fuse the output tubes cathodes, but then you have to watch the filament to cathode voltage ratings; and depending where you connect the fuse, you have to watch the self bias capacitor voltage ratings, etc.
I remember a Lead Acid Gel Cell in a metal pack, that connected via a wire to a voltage invertor in another metal pack. It was CSA approved.
But there was no fuse leading from the Gel Cell to its output connector. The invertor had a shorted RF bypass capacitor at its input connector. When the two were connected, the wire connecting the two packs smoked.
The product offering was quickly modified to include a fuse in the gel cell pack.
We all make mistakes of one kind or another.
etc.
A spark gap is the least intrusive and simplest OPT protection against kV spikes.
In the old days a 100W Pa amplifier would have such a simple device soldered directly between the output transformer plate to plate connections.
Now days those spark gaps are unobtainium and I make them myself if needed.
Those without the necessay know-how or not so into DIY should use a reasonable selected MOV instead.
In the old days a 100W Pa amplifier would have such a simple device soldered directly between the output transformer plate to plate connections.
Now days those spark gaps are unobtainium and I make them myself if needed.
Those without the necessay know-how or not so into DIY should use a reasonable selected MOV instead.
Last edited: