Electronics > Projects, Designs, and Technical Stuff
issues with capacitor discharge circuit causing system resets -- help wanted
MagicSmoker:
--- Quote from: NiHaoMike on February 23, 2019, 10:18:08 pm ---A half bridge for a capacitive discharge ignition seems like an overcomplicated design - a simple SCR circuit is a more usual solution.
--- End quote ---
One advantage to the half-bridge approach is that it disconnects the HV supply from the capacitor during the firing of the spark plug, so a modest savings in energy use. That said, I otherwise agree with you that it isn't worth the bother.
--- Quote from: NiHaoMike on February 23, 2019, 10:18:08 pm ---
--- Quote from: MagicSmoker on February 23, 2019, 09:36:58 pm ---CD ignition systems produce a lot of radio frequency noise since dumping the charge of a capacitor directly into an inductive load - the ignition coil primary - forms a resonant tank circuit. I did some recent work on just this thing and the noise was so bad it caused two Fluke meters and a very expensive isolated, battery-powered scope to malfunction - stuff which really shouldn't misbehave like that.
--- End quote ---
I'd imagine the spark plug itself would also act as a spark gap transmitter, only more pronounced with the higher spark energy. Joe Smith built a rig out of actual ignition parts specifically to test automotive multimeters for EMI immunity.
--- End quote ---
The spark plug does emit quite a bit of E- and H-field noise, but during my testing of different ignition systems (with the end goal of somehow getting 20J into each spark discharge, vs. the 2mJ that is more typical), only the CDI type (both an off-the-shelf unit from MSD and one I designed) wreaked havoc with my test gear; two different small engine magneto systems (Honda and Briggs) and a much higher energy aftermarket "coil on plug" system intended for late model cars had no effect.
Interesting side note that relates to this issue: the ignition transformer operates in flyback mode in magneto, Kettering (aka "points-style") and even the modern coil-on-plug system I bought, but it operates in forward mode (ie - as a true transformer) in the CDI system. However, the peak current is so high in the CDI system (>100A) that it is guaranteed to saturate the primary up until the plug fires (at which point normal flux cancellation from secondary load current can occur).
silverback:
Sorry did not get back to the bench to test. Changed the two 33k resistors feeding the output drive IC power to 27k and 24k. Intermittent false triggering of output stopped. Confirmed that the output IC supply was dipping two low and causing the IC to power down triggering the output FET's.
If you want to stop the EMI from the ignition system go fully FFR military or do Diesel.
jmelson:
When Q4 turns on, C24 pulls the BOOT pin on down toward zero. D13 will turn on and pull Vs and all associated circuits down to zero, also.
I think what you need is a small resistor in series with D13 to limit current. Also, C24 is vastly too large, I've never used more than 1 uF for bootstrap capacitors.
Also, R44 and R45 seem awfully large to keep all that circuitry powered, they can only deliver about 3 mA. Also, I see nothing to limit voltage on C20.
R44-R45 is delivering 3 mA from a 225 V source. If the circuit ever draws less than 3 mA, the voltage on C20 and U7-Vs will go above 16 V.
Jon
silverback:
The operational frequency of this circuit is low, therefore the boot strap capacitor is larger.
Not using the oscillator at high frequency, used at RPM + multiple spark produced by the additional components around the oscillator pins.
Secondly U7 has a shunt regulator which can limit the voltage to~14.4 to 16.6Volts @5mA, 15mA max.
Turn on is 8.3 to 9.7 volts and turn off 7.3 to 8.7 Volts, so voltage must be above 8.7 volts under all events.
silverback:
One last comment, check the bottom FET for gate breakdown. Gate drive not reaching ~15Volts.
Navigation
[0] Message Index
[*] Previous page
Go to full version