fastest switch

[technosquirrel101]

hi!

i'm working on a project that needs quite fast switching. does anyone know the time it takes a transistor until it has reached max conductivity (or actually the time until electrons flow at usual rate)?

also, i'm thinking 'bout using FETs. i've read that VMOS-FETs are faster than normal FETs and SIPMOS-FETs are even faster than that. is that true and are there parts that are even faster than those?

in case i'm taking FETs i thought 'bout letting the FET conduct a small voltage when driven as logic 0 (and blocking the stream with a Z-diode) and letting go full voltage when switched to logic 1. any comments?

thanks very much,
technosquirrel101.

 

[kinjalgp]

The switching time of a transitor depends upon its physical characteristics i.e. the doping levels of base, emitter and collector. Higher the doping levels, longer it will take to restore back to original state after it was switched off from ON state. For high switching speed, you should use transistors from "high speed switching" category.

Ofcourse MOSFET's are much faster switching devices than transistors.

 

[technosquirrel101]

thanks,

back to FETs: what about SIPMOS-FETs, are they really the fastest common FETs or are there others?

technosquirrel101

 

[kinjalgp]

It seems you are looking for some GHz switching transistor. What is your application?

Well, there are many high speed FET, but I've never compared thier speeds. Philips has something called a "TrenchMOS". Check it out if its faster than SIPMOS @ **broken link removed**

 

[technosquirrel101]

i'm building a high-efficiency coil gun. the idea is to redirect the current through a bypass-resistor into the second capacitor when the projectile crosses a light barrier at the end of the first coil and so on. thus i got two advantages in one:

first, i have a very short impulse-time with max discharge and second, the 'suck-back' is reduced to min. (plus much of the energy is sent into the next capacitor and doesn't need to be taken from the main source)

the FETs are triggered by the light barrier to switch coil/resistor and are supposed to (in best case) switch off the electromagnetic field the instant the projectile reaches the end of the coil (the light barrier is a little further into the coil to give the em-field time to break down). i hope to get as close as possible...

technosquirrel101

 

[Klaus]

Interesting, I toyed with that coil gun idea many years ago but never got to build one. No fancy FET's in those days, the idea was to use the projectile (metal) itself as the switch. Two carbon brushes placed opposite each other in the barrel at a strategic places would complete the switch circuit as the passing projectile makes contact.
About as simple as it gets, the speed of the projectile itself determines the switch speed.
Maybe easier to work out the other parameters with that without having to worry about switch speed and current handling of FET's. You could fit many extra pairs of brushes near each coil to experiment the best location for each.

Whadda are you going to do with this thing?

Klaus

 

[technosquirrel101]

hi!

thanks for the idea, but the bad thing about brushes is that they need to have direct contact with the projectile which means that there's increased friction at that point. (not to mention that i'm using talcum powder which would more or less get between projectile and brushes and might at times cause problems)

also, the projectile takes up eddy currents from the electromagnetic field which result in very high spikes when the brushes get in contact with the projectile.

in the meantime i found some SIPMOS-FETs with switching speeds around 3ns at 24V.

never the less, thanks,
technosquirrel101