PTC voltage ratings - fusing B+?

[NateS]

I know that in previous discussions about the lack (or lack of selection) of glass fuses w/ >250v ratings that we all use 250V fuses for HT and call it a day (for those that chuse to fuse).

But what about PTC fuses from LittleFuse/Polyfuse/etc... ? That's a device which continues to conduct a tiny bit, and it seems that voltage rating would in fact matter very much here. Manufacturer's aren't going to say, because the rating is the rating, but my understanding, plus the fact that the prevalant ratings run on the low side - suggest I can't interrupt B+ with a PTC unless I were to use it to trip a FET or some other interrupter to keep the PTC from being exposed 460-500v in the fault condition.

Anyone have comment about PTC's in this application?

[Enzo]

We have been using simple glass fuses now for a lifetime in high voltage circuits. What would this more complex and expensive circuit do better?


Saying that it keeps conducting a bit is not desirable when you want the fuse to open.

[trobbins]

Polyfuse is certainly low voltage rated per device - so it doesn't look at all valid for high voltage applications as you infer. A quick comparison of I-t curves seems to indicate they also have a much worse (longer) time response for a given overcurrent %.

Can you think of any practical benefit for a resettable over-current device in an amp - I can't (especially for gigging).

[NateS]

Because 1) a $0.40 PTC is cheaper than many of fuse holders for 5x20 fuses, 2) because it would save a ton of board space compared to a 5x20 fuse, and I'm finding that I can often save money overall on the project by using stuff that lets me shrink the board, 3) my HT fuse(s) are going to be internal, self resetting is not bad for this. 4) keeps conducting is a relative thing, it just means that a few microamps are still passing holding through, the device has still tripped and is still preventing the transformer from cooking. 5) except for panel mount fuse holders, many of them have exposed metal would be at high voltages.

Having said all that, I don't think it's going to be viable here - a FET capable of holding off B+, plus zener, and gate pullup are going to nullify any of the potential advantages. That's all in the $1-2 range for onesey/twosies. So my original question was still - is the voltage limit for a PTC a real limit or simply 'that's as far as we tested'. (Not that its not cheap, just that its not cheaper than what it was supposed to replace or worth messing with)

But still if nobody ever asks such questions, we end up with hot wires in glass bottles in 2012. Ironic aint it.

[R.G.]

It's a real limit.

As an aside, if someone were ever to be harmed in any way, right down to hot coffee in their lap, by the few microamps, and it was proven that a part was used outside its specifications, that could easily be portrayed as "reckless endangerment" by a good opposing counsel.

I'm guessing that by saying "A FET capable of holding off B+" you're referring to something very much like Return of the More MOSFET Follies or something like the tube save I've described here, with current monitoring in the B+ or cathode return lines tripping a latch and turning off the MOSFET.

As a second aside, it's in general a bad idea to make tube amp PCBs as small as possible, for a number of reasons. Paraphrasing Albert Einstein, everything should be as small as possible - but no smaller.

[NateS]

Yes - I was thinking about something like your example. I figured that this was in fact a real limit, and significant over voltage would l likely cause it to revert to a traditional fuse, since it does have some amount of hold open current. I'll also note that this is for transformer fusing, and not mains - that's a glass fuse internal to the IECPEM. (Hopefully that will ease some of your lawyer fears)

This is not going on the main board, it's an aux board stuck over in the empty space above the power transformer, to handle a couple of power supply related added functions - fusing, ICL, bias, and heater elevation. BatchPCB works in units of 5cm.

[g1]

Can you think of any practical benefit for a resettable over-current device in an amp - I can't (especially for gigging).

Sorry if this is a tad off topic as it is not for B+, but here is an example of using PTC resettable and polyswitches for heaters and low voltage supplies (cub head). I hadn't seen them in tube amps before either.
I believe if the PTC's are strategically placed they could also function as overheat shutdown's ?

[Steve Conner]

There's a "600R" series of polyfuses, the data table says these are rated at "60/600V" whatever that means. I think exceeding the voltage rating would let out copious amounts of magic smoke, as would attempts to use them in series for higher voltage rating.

To be pedantic, the glass fuses we use in our HT circuits aren't rated to break high voltage DC either.

[NateS]

To be pedantic, the glass fuses we use in our HT circuits aren't rated to break high voltage DC either.

Hence my original question.

The polyfuse 60/600R series are rated for 60VDC operating, 600VAC interrupt. The UL sheet only lists the 60v figure.

I'm not sure about fusing each leg separately either in a FWR design. Bridged, of course - but the bias network is heavily current limited by the dropping resistors - and fusing CT only me down to Heat and HT fuses.

[trobbins]

Performance wise, the polyfuse is way off the mark for I-t. For example a 160mA 600R hits 2 seconds at 2A (nominal 10x overcurrent), whereas a 5x15 250mA 350VAC rated interupts in less than 0.2 seconds (eg. Littlefuse 209 series) and they can come with leads (to compare apples with apples, such as leaded polyfuses). If you have a pcb then you will always need a touch barrier (if there is a safety requirement), no matter whether parts on the pcb are metal or not, as part encapsulation is not normally rated as an insulation barrier.

[R.G.]

I figured that this was in fact a real limit, and significant over voltage would l likely cause it to revert to a traditional fuse, since it does have some amount of hold open current.

That gets tricky. The internals of a varistor is something like a sintered powder/granules. If the voltage gets too high, it may simply never be able to interrupt the series of small arcs between particles. I don't really know. More research is needed here. I tend to think that unless I really, really know more than the manufacturer does about his products, the absolute maximums ought to be observed unless I'm standing there with a fire extinguisher every time it's turned on. But that's just me.

I'll also note that this is for transformer fusing, and not mains - that's a glass fuse internal to the IECPEM. (Hopefully that will ease some of your lawyer fears)

Not my lawyer fears. I'm pretty sure I won't be a party to any eventual case that may arise out of this.

I fully believe in everyone's inalienable right to do whatever they want, given only that they get a chance to get the info to make an informed decision. You have the ultimate responsibility there. Just trying to help in case there were things you hadn't thought of yet. Good luck with whatever you decide.

This is not going on the main board, it's an aux board stuck over in the empty space above the power transformer, to handle a couple of power supply related added functions - fusing, ICL, bias, and heater elevation. BatchPCB works in units of 5cm.

Again, I'm fine with whatever you do, for whatever reason - see above. I tend to insist my PCBs be how I want them, not how the manufacturer wants them, and go find a different manufacturer where this is a problem. But... well, see above. As the EPA notes on every car sold in the USA, your mileage may vary.

@g-one: Yes, PTCs are thermally sensitive, and can be used for thermal shutdown. The actual cutout temp is not all that clear in the datasheets, and it would also be current sensitive like it normally is, so it would take some clever design and testing to get any accuracy. These days you can get bimetal thermal switches for a buck or two, so I would tend to pick one where I could just pick a temperature, but there's a lot of room for creativity.

I've done enough microcontroller programming that I have the setup to write, compile and program a controller in a short time, so I tend to go for solutions where I can simply say to the machine:
"IF (thus and such condition happens) THEN
DO (this to cope with the emergency)
ELSE (get on with the rest of what I told you to do)"

So I would stick something like a thermistor or amplified-diode where I wanted to sense temps and then let the controller shut off the B+, or whatever was appropriate.

I also like the ability to handle many conditions this way, perhaps putting times on how long the over-red-line condition has been going on, and whether other red-lines are being exceeded, as well as keeping a slow-witted user from resetting it too many times too fast.

@OP again:

The polyfuse 60/600R series are rated for 60VDC operating, 600VAC interrupt. The UL sheet only lists the 60v figure.

Interrupting AC is inherently easier than interrupting DC. Ask a welder. AC goes through zero volts and zero amps (although not necessarily at the same time) twice per AC cycle. DC never does. UL may have either not tested, or have tested only the 60V figure. Reading datasheets is an acquired skill. You have to notice what's NOT there particularly and wonder why.

I'm not sure about fusing each leg separately either in a FWR design. Bridged, of course - but the bias network is heavily current limited by the dropping resistors - and fusing CT only me down to Heat and HT fuses.

Actually, I am sure about this one. To protect a winding, you need to fuse one side of any current loop to interrupt it (given that your fuse really can interrupt it, etc.), so a single, non-tapped winding only needs a single fuse. A FWB on a single untapped winding needs only one fuse. A FWCT rectifier setup has two winding sections, so it needs two fuses, one in each side. A single CT fuse can open and might protect in most instances, but leaves you open to an end to end overload. Hence, one fuse per side. Or just count on being lucky.

[loudthud]

I read a story about a guy who designed a switching power supply for military use. An opto-coupler was need so one was chosen rated at 3000VAC. Should be ok at 250VDC right? Wrong. They failed after about 200 hours. Turns out there was a metal migration problem that caused failure at DC but was not a problem with AC because the metal would migrate the other direction every half cycle. Don't try to second guess the data sheet.

[NateS]

Don't try to second guess the data sheet.

Thanks RG for the as always enlightening discussion, and being willing to engage and educate.

Meanwhile you'd better change your screen name to 'silentthud' until you can locate some 500v glass fuses that fit your amps fuse holder.

[trobbins]

Littlefuse had some neat little 450VDC encapsulated fuses (808 series) - they only started from 1A, but looked great for amps if you really wanted a B+ DC fuse - 5mm pcb spacing, but would be neat to solder in free-style.

[Steve Conner]

GE "Red Spot" fuses are rated up to 600V DC, however they are considerably bigger than a standard fuse holder, and I think the smallest one they make is 5A. I've used them in power electronic projects where a short across the DC bus would have caused a sizeable explosion. I'd probably use them in a giant tube bass amp too.