Injectors driver - Sergey89

[Sergey89]

For high-impedance injectors.



Pre-driver with ignition outputs.

[AndreyB]

Is the same driver suitable for other solenoids, like idle air valve solenoid?

My experience (especially with valve solenoid - maybe because it has higher frequency, like 200Hz) was that you need to isolate - otherwise, you are getting all kinds of noise fed back into the micro controller.

[Sergey89]

Is the same driver suitable for other solenoids, like idle air valve solenoid?

Yes. But we must choose between TVS or diode clamping circuits.

My experience (especially with valve solenoid - maybe because it has higher frequency, like 200Hz) was that you need to isolate - otherwise, you are getting all kinds of noise fed back into the micro controller.

I have not seen optocouplers in OEM control units. It is necessary to further explore this question.

[kb1gtt]

OEM's have lots of control over how things are manufactured, so they can predict and control things very well. Most notably is the low side impedance when under full injector loads. As an example, if you have 8 cyl, with small injectors, such that you run 99% duty, you'll have a min of 8 amps flowing, with ripples in current when the 1% part of the duty cycle comes around, as well spark pulses will creates several ripples in the low side currents. At 8A with a 12 AWG you'll have a .21V drop across this low side wire. If you snap the throttle shut and go to 0% duty, you'll suddenly change that voltage drop to nearly 0V drop. This ripple in voltage on the common wire can cause errors on your ADC readings. If you are an OEM you can create virtual ground references, or because you know the issue, you can warp your signals. However with an aftermarket device, we don't know the exact length of wire or the quality of the crimps and electrical connections and such, so we don't have as many options to compensate for this kind of noise issue. Also remember if you go staged 4 cyl, you'll have the same problem. Even if you bump up to a massive 2AWG wire, you'll still have .021V ripple, which is several clicks in ADC land.

So I suggest galvanic isolation for after market devices power hungry circuits. Doing so will allow for many more variations in the field MFG capabilities, and makes things much easier to deal with in terms of field installation issues. I'm sure we all know of some other ECU's that have gained a reputation of needing to sacrifice a purple chicken next to the car on the second Tuesday of a Klingon leap year. If you control what you can control, you can make this a much more enjoyable experience for the installer.

I can also talk about how .1 ohm crimps create tank circuits, these tank circuits can cause abnormally large voltage spikes. Of course the galvanic isolation will prevent such voltage spikes from making it to the ECU.

[erich]

National used to have 4 channel peak/hold injector drivers but I don't see them on the TI website.
They still have the LM1949 but it's kind of bulky. They have lots of high current automotive drivers as well.