Re: High Voltage but Low Current fuses...

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Malcolm Watts by way of Terry Fritz <teslalist-at-qwest-dot-net>" <m.j.watts-at-massey.ac.nz>

Hi Terry,
           I would be surprised indeed if you could design a fuse that
would protect against anything let alone events of the nature you are
talking about. An evaporated fuse wire just creates a nice plasma
path for the current to follow. The electronics is always the first
to go, the fuse acting as an visible alarm to its demise. I have to
agree with Dan. I think a robust front end is the way to go, i.e. a
string of resistors with a suitable voltage rating. It's woth bearing
in mind that the carnage mentioned below happened to probes etc.
which were not designed for the purposes they were used.

Malcolm

On 17 Jul 2003, at 22:22, Tesla list wrote:

 > Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>
 >
 > Hi Dan,
 >
 > At 11:12 PM 7/17/2003 -0400, you wrote:
 > >I'm still very confused to why you are worrying about providing high 
voltage
 > >fuses for the differential probes.  The first stage of your differential
 > >probe should be
 > >a very high impedance divider on the order of 1Meg up to 100Meg (assuming
 > >pure resistance now) depending on the voltage being used.  You should 
not be
 > >providing
 > >any type of fusing or transient suppression until after this divider.
 >
 > Unlike "normal" uses for diff probes, users "here" might connect them to
 > NSTs, pole pigs, MOT circuits held together with duct tape...  We may be
 > measuring a 20 volt gate signal in a 20kV IGBT array...  Or, we may be
 > monitoring IGBT gate voltage while running the IGBT at 1000% of it's Ic
 > rating ;-))  In other words, We ain't normal folks :o))
 >
 > If I hook my Tek diff probe to a MOT, it will make a very pretty and "very"
 > expensive fire.  The sudden application of a 20 kV impulse cap to the input
 > will tear right though the divider crap of a normal diff probe.  All I am
 > trying to do is harden the probe.  To increase the chance of survival (of
 > the scope) if instead of it seeing a 20V gate signal riding on 500V,
 > suppose it has to suck down a 20000V 20 Joule hit!!  Or eat the output of a
 > 60,000V 10,000A hit from my Lichtenberg project!!
 >
 > >Therefore, there is no need for a high voltage fuse etc...  If you look at
 > >any run-of-the-mill differential probes made
 > >by Tektronix or similar, they are all built this way.  Remember, you could
 > >have a 1000:1 divider, so a 10,000V transient spike would only look 
like 10V
 > >at the output of this stage.
 >
 > No, it will arc right through the dividers, fry them like ants...  Tek
 > probes would then require a "replacement module"...  I have seen a fair
 > number of very nice Tek, HP, and Agilent devices fried like bacon...  In
 > one case, the only recognizable part the was left just showed the number of
 > the lab we borrowed it from ;-))  I have seen HP34401 multimeters "float"
 > around the lab fueled by flame spewing out the bottom...  Fluke data
 > loggers give off a very distinctive odor when fried...  I once handed an
 > intern an HP 34970a data logger exposed to 1200 volts and asked him to
 > salvage the usable parts.  He returned the line cord, handle, and the two
 > rubber parts on the ends.  He reported that the insides were "cremated to
 > the max!!"...   Many times I have loaned things out to people only to have
 > them return saying "Terry!  %^#%, I sure am sorry!!  I am filling out a PO
 > for a new one for you right now..."  I too have gone to the lab with PO in
 > hand just writing the part numbers down after an "event"...
 >
 > So I am trying to "increase" the chance of scope survival (mine cost "me"
 > $3800) in the case of "the probe is in a ball of plasma" event...
 >
 > Our little probe idea costs 1/40 of two Tek probes that does the same
 > thing.  They may be put together by "anyone" without and CE certified
 > testing....  We just have to try and make them much more bullet proof than
 > the average probe...
 >
 > >If say 5V was
 > >the limit going into your differential amplifiers etc..., then a clamp /
 > >suppression network at 5V would suffice here.
 >
 > Got any 5 volt clamps that can take the 20000V,  25amp shorted pole pig
 > "input" ;o))
 >
 >
 > We can't protect against "everything", but we can "try" to protect against
 > the "expected"...
 >
 > Cheers,
 >
 >          Terry
 >
 >
 >
 >