...
My question is aimed at Gyro who states that the old design of double fuse (one glass and one big one in series) was preferable to him because of the fact that it would not blow the big one most of the times, so saving money and the hassle of replacing it.
That's why I asked what I asked. Other than cost related (that still as I said I don't see as a minus since I blown one fuse in 15 years) I don't understand why such design in his idea is preferable to the current one.
Firstly, please note that nowhere did I mention "glass" fuses. The use of a low breaking capacity physically fragile glass fuse would be patently ridiculous in a serious meter safety setting.
My point was simply that Fluke used the double fuse approach to avoid blowing the large and expensive 10kA breaking capacity mA fuse, designed to clear faults in CAT III / IV situations, due to a silly mistake (eg. forgetting a large capacitor inrush, or forgetting the jacks) in a Cat I environment, you might almost say wasteful. It's not something I make a habit of (I haven't blown one for years, but I'm retired now) but people tend to be a little less cautious on the bench than situations where they may receive an arc flash in the face! Having blown such a fuse, you are
duty bound to replace it with a suitable replacement to maintain the integrity of the meter. Unfortunately this leads to delay (unless in a lab with spares stock), frustration , and of course cost - hence the flourishing market in fake fuses . Even in a large development lab, I remember the gentle ridicule and teasing that people suffered when going for a replacement meter fuse (one step down from having to tell the calibration dept that you had wrapped the needle on your AVO or blown the 50R terminator on your scope!). At least one member on here has admitted to bridging the fuse with a thin strand of wire, and you too frequently hear of examples of test equipment where this has happened - my 25 was probably one step away from that fate when I got it, with both 10kA fuses already blown.
At some point, somebody at Fluke said 'let's not do that any more' and we will never know why. As I mentioned previously they did it on their most rugged DMM, with all sort of certifications, which was marketed for many years. They must have done extensive testing of the safety of the dual fuse approach. Apart from the corner case that bdunham7 mentioned (which I suspect would actually be sucessfully quenched by the 20mm fuse), there are many internal parts that rely on the 10kA fuse to safely clear a fault in CAT II+ environments
including PCB traces, clamping diodes, switch contacts etc. This would all have been tested.
I would love to see Joe Smith put a Fluke 25 / 27 through it's paces, I think the closest he has come is the Fluke 77, but that used spark gaps rather than 5 MOVs. Even then, he does not have the capability to test high current fault clearing. It would probably be one for Photoinduction!
Comprehensive testing would require a large pile of expensive fuses anyway.
As I mentioned, I also rarely use meter current ranges (none of my bench meters even have them), but I felt duty bound to replace the fuses with genuine items anyway (again, to maintain the integrity of the meter). I have a bunch of shunts, ranging from 1k (1mA) up to 150A. These of course are only suitable for CAT 1 lab bench environment (maybe occasionally CAT II in a UK fused plug environment). Current clamp probes are a godsend for AC in higher CAT environments, much less satisfactory for measuring low DC leakage currents in the same situation.