# Which Diode?

Discussion in 'Electronic Basics' started by Mike, Nov 9, 2004.

1. ### MikeGuest

I'm making a very simple coilgun. This basically works by charging up a
capacitor and discharging it through a coil. I understand that I should
place a diode across the coil to stop any current going back into the
capacitor. Which properties of a diode should I be looking at when I make
my selection? In this example let's say that the capacitor has a voltage
rating of 100V and a capacitance of 200uF. When I look at the available
diodes the properties of each is meaningless to me.

TIA

2. ### John PopelishGuest

Reverse voltage capability is cheap, so there is no good reason to tey
to cur this close. A 200 to 400 volt unit will cost almost the same
as a 50 volt one. The biggest concern will be peak forward current
capability (surge current rating), since this is what you are pushing
through the coils. 3 amp continuous rated lead mounted diodes tend to
have oversized junctions to let them get rid of the normal heat
through their leads. This tends to give them high surge current
ratings. Even more so for 6 amp units.

For instance:
1N5404, 400 volt, 3 amp, has a surge current rating of 200 amps.
http://www.diodes.com/datasheets/ds28007.pdf
6A4-T, 400 volt, 6 amp, has a surge current rating of 400 amps.
http://www.diodes.com/datasheets/ds28008.pdf

3. ### MikeGuest

Okay, that helps a bit, thanks. How do I calculate the peak forward
current? Do I calculate the resistance of the coil (plus the ESR of the
capacitor) and then just apply Ohm's law (using the maximum voltage of the
capacitor)?

Thanks again.

4. ### John PopelishGuest

You would do better simulating the circuit with spice. The inductance
of the coil should be a big part of the current pulse shaping, with
the wire resistance and the capacitor internal inductance and
resistance minor players. If not, there will be a lot of energy
dumped into places you do not want it to go. Of course, you will also
have to get hold of the formulas that allow you to predict the
inductance of various coil forms. E.G.

http://www.linear.com/designtools/softwareRegistration.jsp

5. ### rayjkingGuest

Mike,

You can determine which diode to use by multiplying the cap size times the
voltage stored, this gives you Jules. Jules = C*V = I * T so if 100 volts *
200uf = 0.02jules ( V*C ) then a diode with a rating of 30 amps surge for
60hz ( really only 8.33ms sees current ) ( I * T ) = 0.25 Jules is more than
enough. You should not go beyond 25% of the surge rating for continuous
applications. The surge ratings are for fuse considerations and these surges
are for less than 10 times for the life of the diode The one amp diode with
a 30 amp surge rating should not be used beyond 7 amps for 8.33ms or 0.058
Jules.

Ray

6. ### Neil PrestonGuest

Before answering the question, I would like to know how he intends to
connect the diode.....

My first impression was that he was told to put a quench diode in parallel
with the coil to clamp any oscillatory transients. In this case, I believe
the diode need not be very heavily rated.

Most of the responses I've seen, though, appear to assume that the diode
will be in series with the coil, in which case the coil current will rise
slowly according to the time constant formula to a maximum of approximately
Ec/Rcoil (depending on the relative values of C and L.)

Shouldn't be much more complicated than that, should it?

Neil

7. ### Fritz SchlunderGuest

Errr... Better double check that energy storage formula. The energy stored
in a capacitor is E=0.5*C*(V^2), where E is energy is Joules, C is
capacitance in Farads, and V is the voltage on the capacitor in Volts.

So for a 200uF capacitor fully charged to 100V, it should store one Joule of
energy.

I apologize for disagreeing with you, but one must nip misinformation in the
bud before it spreads and gets out of hand.

I'm not sure I agree with these ideas either. Limiting yourself to a peak
current of 7A (for 8.33ms) on a typical 1A rated diode such as the 1N400X
series devices is probably more conservative than really necessary in many
cases (emphasis on the words "many cases", not "all cases"). Do you have
some references that you can point us to that support your ideas and
methods?