Driving Ignition Coils w/ MicroSquirt
Posted: Tue Sep 11, 2007 4:54 pm
A couple of things to note for those installs that are direct-driving ignition coils with MicroSquirt (i.e. not using an ignition driver module but the VB921 drivers in the MicroSquirt):
1) Make sure to set the Spark Output to 'going high (inverted)'. This sets the correct polarity for the output.
// By the way, for those just curious on why this particular nomenclature, the answer is that the driver IGBT in the MicroSquirt inverts the signal in its configuration. The processor output on the MicroSquirt (you do not have access to this wire, it is hard-wired to the IGBT) goes high in order to turn on the IGBT driver, which then pulls the output low (i.e. inverts). //
2) You need to have a handle on your dwell setting for the coil you are using. All the dwell number represents is the amount of time (in milliseconds) it takes to charge up the coil. Yes, because of of a parameter known as inductance it takes time to bring the coil up to a desired energy.
The dwell number is driven by the primary inductance and resistance. Chances are you do not know the inductance or resistance. If you do not then you can do one of two things:
a) You can measure the inductance and resistance yourself. You will need an ohmmeter, which you probably already have. The other item is an inductance meter, which you may not have. Go find one of your geeky friends and ask to borrow their inductance meter (betcha they have one..). You can also become a geek yourself and purchase an inductance meter, you can get one on Ebay often for little money. With the inductance and ohmmeter, grab the following PC application:
http://www.megamanual.com/files/softwar ... nalyze.zip
Follow the instructions included with the application and you will quickly determine the ignition coil dwell time. One item to note is that the IGBT driver used in MicroSquirt saturates at 7.5 amps, so use this number as the charge current inthe application.
b) Guess on the dwell number! You can start with a number, like 3 milliseconds and work from there. In fact 3 milliseconds should be enough to get the engine running and is an all-around good starting number.
c) Obtain a 0.1 ohm resistor at 10 watts or higher (something to look for at the next Hamfest or other electronics gathering) and put inline with the +12V going to the coil, and use a oscilloscope to measure the voltage across the resistor. Using Ohm's law, the current is equal to the voltage divided by the resistance. What you will see is the scope trace ramp up as the coil charges. If you keep on increasing the dwell number in the software, you will get to a point where the ramp levels off - at this point is the target dwell number to use.
3) Realize that driving an ignition coil generates a lot of noise. This noise can get back into the other MicroSquirt circuitry and cause false tach signals, processor resets, sensor noise, etc. On MicroSquirt, everything is packed tightly together, so it has more chance for noise coupling. A few things to help with this:
a) Keep the ignition drive wires (and injector) away from the other wires on the MicroSquirt connector. It does not take much to couple noise back, especially thru the serial link.
b) Make sure the grounds are correct - see the sticky on grounds.
c) One thing that you can do to significantly reduce noise is to use a snubber capacitor. Remember in the old days with points (Kettering) ignition? With the points was the condenser. And everyone knows that the condenser was there to keep the points from arcing and burning out.
Here is why the condenser (a.k.a. capacitor) reduced the sparkingL A capacitor blocks direct current (DC) but will pass AC current. When the points open there is a fast transient in voltage and current. What the condenser would do is to briefly shunt the current around the points and complete the path. But it would only do this for a brief moment, but enough to reduce the arcing.
The same thing can be done with solid state ignition drivers - in fact systems like the Ford EDIS and GM HEI use capacitors located at the ignition coil. These capacitors will briefly shunt the flyback current, bypassing the driver, for a brief moment. What this does is reduce the stress on the driver AND reduce the radiated noise emission (EMI). Its a significant benefit. The capacitor does change the discharge pattern a bit (it sets up a L-R-C circuit) but not enough to notice.
You can easily do the same thing. On the IGN signal from the MicroSquirt going to the ignition coil, connect a 0.01 microfarad capacitor rated at 630 volts or so, connect the other terminal to ground. Connect this up right at the ignition coil (EDIS does this) such that the shunt path is contained right at the ignition coil. You need to use a Polypropylene (PP) capacitor rated for pulse discharge operation. My favorite is the Wima MKP-10 series:
http://www.wima.com/EN/mkp10.htm
You can also use an equivalent, like Epcos B32621A6103J (you can get this at http://www.digi-key.com).
- Bruce
1) Make sure to set the Spark Output to 'going high (inverted)'. This sets the correct polarity for the output.
// By the way, for those just curious on why this particular nomenclature, the answer is that the driver IGBT in the MicroSquirt inverts the signal in its configuration. The processor output on the MicroSquirt (you do not have access to this wire, it is hard-wired to the IGBT) goes high in order to turn on the IGBT driver, which then pulls the output low (i.e. inverts). //
2) You need to have a handle on your dwell setting for the coil you are using. All the dwell number represents is the amount of time (in milliseconds) it takes to charge up the coil. Yes, because of of a parameter known as inductance it takes time to bring the coil up to a desired energy.
The dwell number is driven by the primary inductance and resistance. Chances are you do not know the inductance or resistance. If you do not then you can do one of two things:
a) You can measure the inductance and resistance yourself. You will need an ohmmeter, which you probably already have. The other item is an inductance meter, which you may not have. Go find one of your geeky friends and ask to borrow their inductance meter (betcha they have one..). You can also become a geek yourself and purchase an inductance meter, you can get one on Ebay often for little money. With the inductance and ohmmeter, grab the following PC application:
http://www.megamanual.com/files/softwar ... nalyze.zip
Follow the instructions included with the application and you will quickly determine the ignition coil dwell time. One item to note is that the IGBT driver used in MicroSquirt saturates at 7.5 amps, so use this number as the charge current inthe application.
b) Guess on the dwell number! You can start with a number, like 3 milliseconds and work from there. In fact 3 milliseconds should be enough to get the engine running and is an all-around good starting number.
c) Obtain a 0.1 ohm resistor at 10 watts or higher (something to look for at the next Hamfest or other electronics gathering) and put inline with the +12V going to the coil, and use a oscilloscope to measure the voltage across the resistor. Using Ohm's law, the current is equal to the voltage divided by the resistance. What you will see is the scope trace ramp up as the coil charges. If you keep on increasing the dwell number in the software, you will get to a point where the ramp levels off - at this point is the target dwell number to use.
3) Realize that driving an ignition coil generates a lot of noise. This noise can get back into the other MicroSquirt circuitry and cause false tach signals, processor resets, sensor noise, etc. On MicroSquirt, everything is packed tightly together, so it has more chance for noise coupling. A few things to help with this:
a) Keep the ignition drive wires (and injector) away from the other wires on the MicroSquirt connector. It does not take much to couple noise back, especially thru the serial link.
b) Make sure the grounds are correct - see the sticky on grounds.
c) One thing that you can do to significantly reduce noise is to use a snubber capacitor. Remember in the old days with points (Kettering) ignition? With the points was the condenser. And everyone knows that the condenser was there to keep the points from arcing and burning out.
Here is why the condenser (a.k.a. capacitor) reduced the sparkingL A capacitor blocks direct current (DC) but will pass AC current. When the points open there is a fast transient in voltage and current. What the condenser would do is to briefly shunt the current around the points and complete the path. But it would only do this for a brief moment, but enough to reduce the arcing.
The same thing can be done with solid state ignition drivers - in fact systems like the Ford EDIS and GM HEI use capacitors located at the ignition coil. These capacitors will briefly shunt the flyback current, bypassing the driver, for a brief moment. What this does is reduce the stress on the driver AND reduce the radiated noise emission (EMI). Its a significant benefit. The capacitor does change the discharge pattern a bit (it sets up a L-R-C circuit) but not enough to notice.
You can easily do the same thing. On the IGN signal from the MicroSquirt going to the ignition coil, connect a 0.01 microfarad capacitor rated at 630 volts or so, connect the other terminal to ground. Connect this up right at the ignition coil (EDIS does this) such that the shunt path is contained right at the ignition coil. You need to use a Polypropylene (PP) capacitor rated for pulse discharge operation. My favorite is the Wima MKP-10 series:
http://www.wima.com/EN/mkp10.htm
You can also use an equivalent, like Epcos B32621A6103J (you can get this at http://www.digi-key.com).
- Bruce