Fuses, 2 in series to comply with needed voltage rating

Yes, I will post back heren when they respond (probably with a no can do)

Cheers

Klaus

uses

he power module then It may do the trick

You are right, I am stubborn. A property that sometimes is bad, sometimes g ood. I'm not arrogant, don't have a PHD or anything fancy like that

e or what sort of fault current is possible.

The one I was looking at just for fun was:

Has very inferior breaking capacity, 20 times less than the one we are usin g currently.

hat answer MUST suit YOU.

I am no expert oon fuses and that is the reason for posting here. The reaso n I ask follow up questions, probably pushing it a bit to far (sorry if tha t got tiresome), is that I have previously been told that something could n ot be done by collegaues etc and then I gut stubbord and went and did it in the lab

Thanks for posting back

Klaus

There is no requirement for a fuse in this product. It's only for protection for internal components (one higher power variant has no fuse, but has other components that don't need protection)

Cheers

Klaus

I didn't know that, thanks

Just reading a bit about arcing and how to prevent it, I saw some interesting methods like arc blowouts, magnetic arc deflection ect.

Way to advanced for our product, so it seems we have to stick with the current fuse and find ways to reduce thermal stresses (one idea is cutouts in the via supporting the leads, haven't seen it done anywhere before though)

Thanks

Klaus

However, it probably behooves you to not add a component to the circuit which may create an unquenched arc inside of your device. Your cure may work out to be worse than the disease.

Do I hear the sound of back pedaling?

Have you asked an applications engineer at your fuse supplier? They may have good ideas, either on how to make your current fuse work, or how to use a different (single) fuse to do what you want.

Matt Roberds

Would they not just say here is our 500v fuse, and yes it does cost more than 0.1c

NT

The first link clearly says the voltage rating of each individual fuse should be greater than the total suppply voltage. The second also gives that as a general rule but then becomes confusing - I suspect there may be a typo and the word *not* is missing, as in : .... inside some converters the fuses do [not] melt at the same time and a 30% imbalance factor is introduced.

All this seems to me to say that one cannot use low rated fuses in series to achieve higher voltage breaking.

There are now plenty of HV DC rated fuses made expressly for solar PV applications, they are characterized by being physically long!

piglet

Once saw a MeanWell supply that got some moisture ingress. This was one of those metal frame, DIN rail mount, industrial parts, made for

240-480VAC input. I believe they use a ceramic body pigtail fuse inside these power supplies... I couldn't tell based on the wreckage, though -- just two fuse caps sticking out at odd angles from the board. The casing (whether it was ceramic or glass) was pulverized, nowhere to be found.

There was evidence of arc flash near the breakdown site (likely, residual moisture between the switching nodes and the edge of the plastic insulating sheet they put over the chassis), but not by the fuse. Seems it died, rather spectacularly, but did its job to clear the fault and extinguish the arc.

Forget if it was a 5x20 or AGC size fuse.

Arc flash is interesting. Against a PCB, exposed pins, pads and vias are eroded (in proportion to their current capacity and proximity to the hottest part of the arc cloud), but coated surfaces erode very little if at all. Likely, soldermask is ablative, eroding from the intense heat, but only generally so, not offering electric field nor current flow to the plasma. Afterwards, the combined metal vapor and carbon soot condense on nearby surfaces, giving that characteristic charred appearance.

Semiconductor fuses (so-called because they're fast enough to maybe protect the most robust types of semiconductors -- diodes and SCRs) may not be a very practical choice over conventional cartridge fuses, if you already expect all the semiconductors to fail (or can't get low enough I^2*t to do so), but they will limit the propulsion of schrapnel.

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

Somewhere in this thread, Klaus said he has an appropriately rated single fuse now, but it's failing due to thermal issues. The supplier might be able to help with a suggested PC board layout, or something like "ah, you're using model 123, you might try model 345 which has lower internal resistance" or "we have an extended temperature range model 123-T, with different solder inside, that might help".

I agree that they're not going to tell him how to use two 250 V fuses in series to do what he wants. :)

Matt Roberds

Yes... years ago we used open wire links for fuses. I wonder if the OP could make one compliant somehow, eg with fibreglass sleeve and a cutout in the pcb.

NT

Make sure to coat them in Napalm, just in case they don't all catch on fire.

--
Anyone wanting to run for any political office in the US should have to 
have a DD214, and a honorable discharge.

We have been in contact with the supplier. Sadly their approcach to fuses a re trial and error, so not much help there

In the meantime I have been drawing up some ideas to relieve the mechanical stresses on the fuse. The current fuse is leaded, and the assumption is th at the stress comes from the very thick leads. So one idea is to mill some of the arc geometry, so that the PCB suppport in the pad is angled to the d irection of the fuse (that way forces are transferred to the PCB instead of the fuse itself)

I just hope that does not produce another unknown failure mode

Cheers

Klaus

are trial and error, so not much help there

al stresses on the fuse. The current fuse is leaded, and the assumption is that the stress comes from the very thick leads. So one idea is to mill som e of the arc geometry, so that the PCB suppport in the pad is angled to the direction of the fuse (that way forces are transferred to the PCB instead of the fuse itself)

leave the pins long with a "curl" on them?

mount it standing instead of laying?

-Lasse

are trial and error, so not much help there

al stresses on the fuse. The current fuse is leaded, and the assumption is that the stress comes from the very thick leads. So one idea is to mill som e of the arc geometry, so that the PCB suppport in the pad is angled to the direction of the fuse (that way forces are transferred to the PCB instead of the fuse itself)

A standard approach in such cases is to put a loop in the component leads. But I thought your fuses were sm not th.

NT

:

e:

r

es are trial and error, so not much help there

ical stresses on the fuse. The current fuse is leaded, and the assumption i s that the stress comes from the very thick leads. So one idea is to mill s ome of the arc geometry, so that the PCB suppport in the pad is angled to t he direction of the fuse (that way forces are transferred to the PCB instea d of the fuse itself)

We have allready tried with a "curl". The pin diameter is large, so the cur l is not enough

Cheers

Klaus

:

te:

ote:

st

ier

g e

es

uses are trial and error, so not much help there

anical stresses on the fuse. The current fuse is leaded, and the assumption is that the stress comes from the very thick leads. So one idea is to mill some of the arc geometry, so that the PCB suppport in the pad is angled to the direction of the fuse (that way forces are transferred to the PCB inst ead of the fuse itself)

url is not enough

2 curls :) Does it not suggest something is wrong mechanically with the fus e design?

NT

What if the arcing reduces the current, so the second fuse never blows but the first continues to arc? I'd not do this.