Thoughts on cathode current sensors

[RWood]

I've built a 1987 Plexi in which I installed 1R from pins 1/8 to ground to read current for biasing. I originally used a 1/2w but changed to 2w after reading a few articles on amps toasting 1/2 resistors during gigs.
After more study on the tendency for EL34s to short between cathode and heaters, it seems that a 1/2w acting as a fuse would be a better idea.
Seems to me that the resistors cooking causing failure during a gig would indicate problems that would have caused catastrofic failure which would take out other components anyway.
I'd love to get thoughts/experiences from the TAG gurus on the subject.

[matt h]

(deleted)

[Leo_Gnardo]

No guru here. But I've been using Dale 3W wirewound 1% 1.00 ohm resistors for this function. "Way overkill" some say, but I have accuracy when measuring and so far no failures. Score one tiny point for mental health: one less thing to worry about. I get 'em at Mouser, 100 at a time for the quantity discount. Note Ampeg uses the same resistor at 10 ohms for cathode current-sensing in their SVTs. On occasion I've found those roasted by shorted output tubes.

Here's your gurus, pick the one that makes the most sense to you:

[rp]

Sexy Sadie what have you done
You made a fool of everyone
You made a fool of everyone
Sexy Sadie ooh what have you done

Sexy Sadie you broke the rules
You layed it down for all to see
You layed it down for all to see
Sexy Sadie oooh you broke the rules

One sunny day the world was waiting for a lover
She came along to turn on everyone
Sexy Sadie the greatest of them all

Sexy Sadie how did you know
The world was waiting just for you
The world was waiting just for you
Sexy Sadie oooh how did you know

Sexy Sadie you'll get yours yet
However big you think you are
However big you think you are
Sexy Sadie oooh you'll get yours yet

We gave her everything we owned just to sit at her table
Just a smile would lighten everything
Sexy Sadie she's the latest and the greatest of them all

She made a fool of everyone
Sexy Sadie

However big you think you are
Sexy Sadie

[pdf64]

Yes, I think it's beneficial to have 1/2 watt resistors in the cathode return, in order that they're likely to fuse in the event of fault current.
In view of that function, best for them to be MF / MO with a flame retardant coating.
Particularly beneficial on 4 tube amps with a lot of B+ capability / large B+ fuse, for which the overcurrent of one tube entering meltdown may take a while for a secondary or primary PT fuse to blow.

Whereas (on the same topic) I think that it's beneficial for heater circuit ground reference resistors to much more robust than the 1/2 watt normally fitted, eg >3 watt with a 500V element.
Because a power tube short can pull the whole heater line towards VB+ when the heater ground reference is lost, thereby stressing the cathode-heater insulation of every tube.
So robust resistors hopefully can maintain the fault current until a secondary or primary PT fuse blows.

[trobbins]

I've come across a few old amps where a cathode sense has clearly 'cooked', and the PT and OT were still alive. I've also come across old amps with failed OT, which had no cathode sense resistors.

Typically those amps just relied on a mains AC fuse only.

Its not uncommon for a high current fault to develop in one or more output tubes - that current often can't get to a sufficiently high level to blow a primary or even a secondary side CT fuse, depending on OT resistance and the failure mode.

So I guess I've been pleased to get a dead amp with a blown cathode sense resistor, as it usually means the OT and PT are still ok.

For sure, if you can fuse the cathode(s) for extra protection then great, but not everyone wants to go that far.

[Leo_Gnardo]

Sexy Sadie what have you done
You made a fool of everyone

Yep, that's the one. Such a bogart too, never did pass de dutchie. No wonder he was always giggling.

Here's the part he didn't introduce to John Paul George nor Ringo. And totally appropriate for today's subject.

[trobbins]

I agree with pdf64 for 'poor mans fusing' on cathodes.

Faults to heater are interesting. There are certainly instances of pin 2 to 3 flashover, and some examples that have 'chewed' away for a while at the valve base. The conduction may have started from anode peaks, and possibly continued even under idle conditions due to carbon tracking.

That would be likely on just one valve base (compared to say a fault that caused all power tubes to conduct), and would bypass any inherent valve resistance, and so the fault current may end up depending on OT winding DCR and the power supply.

Certainly a PT secondary side fuse would be the easiest protection to hopefully protect against a heater type fault, and would need fusing set at not too 'robust' a value. But protection is problematic when just a single heater winding is available, and a humdinger pot is used to minimise input stage noise, in which case the pot is likely to end up being the 'protection' element.

[Phil_S]

It seems to me a resistor is a resistor and a fuse is a fuse. Use a stout resistor with a proper size fuse? Don't know how that affects the sense function.

I am thinking if you fuse the HT CT, you have good protection on the secondary.

[matt h]

(deleted)

[pdf64]

a resistor is a resistor and a fuse is a fuse

An alternative view is that a fuse is a resistor with a specified fusing characteristic!
The great thing about 1 ohm 1/2 watt MO/MF cathode resistors is that they seem to blow quickly in response to a fault, faster than a 'proper' B+ fuse, which often need to be a time delay type.

Ideally yes, there would be fuses on all the power tube plate/cathode, screen grid and heaters but it's not really feasible, whereas screen grid and cathode sensing resistors are there anyway, so it seems reasonable to use a type that will quickly fuse in response to fault current.

Also the advantage that they're device specific, so the amp may be able to carry operating (at a reduced level) following an output tube failure, especially with a 4 tube output.

[RWood]

Looks like we go both ways.
The amp was built in the chassis of a Carvin TS100 power amp that burned the traces out of the circuit board when the heater and plate shorted in one of the EL34s. The heater winding only produced 4.8 volts after that so I gutted the whole thing saving only the 2 OTs. Replaced it all with PTP 1987 plexi circuit.
The cathode ground was also burned which is what got me to researching this topic.
Thanks for the replies...and for what it is worth, I do consider most of you my gurus because you always lead me to the answers I'm seeking!

[rp]

Just an aside, anyone seen, or ordered accidentally like me, those 1ohm 1W 1% that are as small as 1/4 watters? Good to to use or should I toss the dozen I have? They sure don't look good to use. Anyone have a Mouser or Farnell part number for a 1W that at least looks right?

You guys are my gurus too but half the Bhagavans always say a resistor is not a fuse and the other say it's even better than a fuse. Are you testing me, my masters? Today I follow Master PDF64.

[Leo_Gnardo]

Just an aside, anyone seen, or ordered accidentally like me, those 1ohm 1W 1% that are as small as 1/4 watters? Good to to use or should I toss the dozen I have? They sure don't look good to use. Anyone have a Mouser or Farnell part number for a 1W that at least looks right?

Those poxy little R's sound exactly like the ones that started a long long thread at MEF some months back. All the same argy-bargy back and forth. "theoretically they can't blow." "Ah but they do." Experience trumps theory and that's the way the cookie bounces. Over a hundred replies; the time wasted typing far outweighed the cost of resistors whether you choose tough ones or fuses-in-disguise.

If you intend to use them as fuses, go ahead, use 'em. If not, put in something that won't break. Mouser # 71-RS2B-1.0 or 71-RS2B-1/R 93 cents each, 75 cents in 100 quantity, -65C to +250C (!), 50 ppm drift, that's a dam' good resistor. Costs more but won't break the bank. You may have to forego a slice of tiramisu at most to fit an amp with these.

Next, photons. They're a wave! They're a particle! Rinse lather repeat discuss. Makes a nice sandwich.

Who needs a guru when you have The Worlds' Foremost Authority:

[pdf64]

fusing the CT only dinks with the DC, not AC voltage-- so you can still have high nasty AC doing crazy stuff

Yes, I agree that it's not the best solution, and Merlin's view is that each winding leg / phase should be fused.
A 1960 Wem Westminster I worked on had that arrangement, but CT fuses seem a lot more common.
The problem is that is fairly easy to find room to fit a CT fuse somewhere in the amp, and it protects the winding from maybe 95% of likely failures.
What it misses are overcurrents across the phases, eg shorted rectifier, which isn't unknown with tube rectifiers.
However, to manage that, silicon diodes in series with the tube plates can be used.
So, I think a CT fuse isn't ideal but on balance, if we wish to limit the number of back panel fuses to 2, then it may be that the B+ CT is a better option than a B+ fuse after the rectifier (the other fuse being in the line circuit).