field test unit no 001 "model T"

[Davor]

Only Rx will tell you. If noisier with IRF, go for Cree, but don't forget to share your conclusions with us

[moodz]

Ok .... avalanche issue cleaned up ... the front end ( adaptive damping system and amplification on its own with no back end processing ) can detect the larger test targets out to nearly a meter now ... ... this puppy is going to be insane when the backend goes on.


..the usual set up ...

18 inch mono coil
0.4 ohm
100 us TX

BTW ... synchronous demod / sampling is now obsolete ....

[6666]

Which Cree silicon carbide MOSFET did you use ?

[hdphilip]

hello 6666

I used a cree CMF10120D, mosfet, note the package size... i had to mount it next to the pcb with jumper wires,

Philip

[6666]

Hi Philip, they are US$124.00 here, ouch.

[hdphilip]

6666

i got them from digikey here about 16$

what i wonder is with the 1200 volt breakdown would they damage a coil with a higher flyback?

philip

[Davor]

... or a cable, connectors, rig entrails...
How about a snubber instead? If you integrate the curve under the PI peak, you'd hardly find justification for excessive peaks.

[moodz]

6666

i got them from digikey here about 16$

what i wonder is with the 1200 volt breakdown would they damage a coil with a higher flyback?

philip

... on second thoughts dont feed too much voltage into a commercial coil ... once an arc causes carbonisation you could blow your dough. What are the voltage ratings of commercial coils ??

[chudster]

... or a cable, connectors, rig entrails...
How about a snubber instead? If you integrate the curve under the PI peak, you'd hardly find justification for excessive peaks.

Following rationale from the forum we want fast flyback to detect the target signal don't we. Fast flyback means higher voltage and higher breakdown doesn't it? Good question though. Do the aftermarket coil people rate the coils for voltage? I spotted these folks who advertise their flyback as a marketing thing (electromagnetic waves). http://www.metaldetection.net/Englis...raison-Eng.htm Up to 1100V. Obviously they feel its an important thing to tell customers the breakdown peak voltage. Can't say I know much about their detectors or whether that gives an edge.

Chudster

[Davor]

Avalanche limit is a natural way of arc prevention, but is not very nice regarding noise, and it stresses the components. So instead you can have a higher voltage device, but nothing to prevent arcing. Not a very good idea.

In case you have some charging control that limits the charging period when flyback goes critical - fine.

[hdphilip]

well maybe putting a commercial coil on a "megger" would be one way to find out about it's insulation property are.

[Midas]

I wouldn't worry about it too much. We're not talking orders of magnitude more voltage. Any decent PI coil is bound to have more isolation than required because its good for reducing capacitance. So if you blow it up it was probably crap to start with, therefore no big loss.

Midas

[hdphilip]

thanks Midas,

i was thinking the same, the coil is a 11" Minelab mono, ( sorry BW ) it's new and hopefully be here in a day or 2, i'll use my Irf 740 at first, then after the fresh coil smell, wears off, i'll drive it with the Sic mosfet..... but in the meanwhile, i think i'll wind a 11 or 12 inch coil with some 16 awg wire, just to see how durable the Sic's really are.


Philip

[Davor]

In case you get arcing anywhere, you'll be able to see it on a scope. Most probable places for arcing are connectors. Make sure cables are rated for over a kV.

[moodz]

... the field test unit makes it off the bench.

... initial results for the experimental unit are showing a magnitude increase in sensitivity above prior art PI techniques.

... oh ... did I mention ground balance and discrim

Fine tuning the code now the hardware is done.