Jump-Start = Too Cool

[Anonymous]

Well that wasn't so bad. Got the Jump-Start put together and working. It does work, and as advertised, unlike some commercial detectors.
Poor little 2N2907A I'm using for the PNP pass transistor isn't too happy right now but hasn't died. The current waveform through the coil is just too cool. It takes off at a steep angle and then breaks over into the last of the tailing off of the coil time constant, when the 12 volts takes back over. Like I thought it would but is neat to actually see it on a scope.
I thought of this circuit some many months ago to get current moving in the coil, and not so much to save power. And guess what, you always learn something after you get everything together, and one thing I had not thought of is this high voltage is also getting put across the damping resistor, wasting more power in it. But not too bad cause it doesn't stay high voltage for very long. But if your goal is to save power then you will get the honor of disconnecting the damping resistor.
JC

 

[Anonymous]

My setup right now for this test is the following.
12 volt supply
770 uh 9 inch 28 turn litz wire < 1 ohm coil
8 ohm 20 watt current limit resistor
IRF740 mosfet.
1N4007UF capacitance blocking diode as learned from Eric with 5 kohm around mosfet.
Used two more 1N4007UF diodes for jump-start.
This gives a bit more than 1 amp though the coil and a 400 volt (barely clamped) flyback without jump-start circuit. Had to turn down the current to experiment with low voltage PNP (60 volt) I have on hand. At first used 0.47uf polypropalene capacitor for the flyback storage cap. Used a 2N2222A to turn on the PNP xsistor.
Didn't spend any time calculating anything just shoved in parts and it worked, so none of the following values is probably perfect. You may choose your own.
Between the emitter and base of pnp = 1k
between the base of pnp and collect of npn = 6.8k
between the base and emit of npn = 2.2k
into base of npn from ttl level signal= 2.2k
with the 0.47 uf cap the integrated flyback voltage is about 22 volts (not too exciting)
put two 0.47 uf in series voltage is 40 volts more exciting.
I used the same ttl signal which through a 2N2907A drives the mosfet to drive the jump-start without advance or delay (may have lucked out here but anxious to try).
I stopped at this point to post this.
Anyway, if you have interest in this, you will not be completely wasting your time on something which does not work at all. Have Fun.
The jump start time happens in about 15 usec in this setup.
JC

 

[Anonymous]

Zetex bipolar super-e-line PNP # ZTX776 is rated 200 volts and will probably handle an ampere pulse gracefully. In current Digikey catalog #T021, page 387. Data books shown upper right same page. Zetex might have data sheets posted on web.
--Dave J.

 

[Anonymous]

1N4007's are pretty slow. A high-speed switching rectifier would probably be a better bet; there are quite a few different types in the Digikey catalog. Schottkies have a lot of capacitance, which could increase damping time, and their advantage of lower forward voltage doesn't amount to anything in this application.
--Dave J.

 

[Anonymous]

These diodes are 1N4007UF and NOT 1N4007, they are much faster, the UF is for Ultra Fast Recovery and the reason I put it in the part number. Yes they are alot faster than plane jane 1N4007.
JC

 

[Anonymous]

I use ZTX458 and 558 at work, they are 400V. Guess I could swipe some. <IMG SRC="/forums/images/smile.gif" BORDER=0 ALT=""> but they are only 0.3 amp, 0.5 amp peak. ZTX957 looks better 300 V and 2 amp peak. or ZTX956 , 200 V, 5 amp but there are alot of transistors which would work.
JC