Automotive Electronics - How to cut fuel injectors, for TC Project

[ks0010938]

Hey everyone Im new here! I was wondering if you could help me solve a problem. I am developing a Traction Control System for my Classic Car. A 1974 mk1 Ford Escort.

The car currently has a Digital Dash and is running a CAN network for engine RPM/Wheel speeds etc.....

PICTURE OF DASH:

For the traction controller I am using a PIC18F1320 to read wheel speeds from the CAN network, calculate slip and adjust the firing sequence of the injectors accordingly.

My problem lies with actually cutting the injector pulses so the injectors do not deliver any fuel.

When I was developing the system I designed it with the assumption that the injector is permanently grounded and the Engine ECU supplies the injector with 12v when an injector is required to deliver fuel.

The Traction Control system would sit in-between the Engine ECU and the injector. When no fuel cut was required the Traction Control would simply mimic thevoutput from the Engine ECU, and when a fuel cut was required It would simply not drive the injector:

However, after reading further literature on websites and texts books; it is evident that in most cases the injector is permanently connected to 12v and the Engine ECU grounds the injector when it wishes to deliver fuel. This is where I am stuck; I am unable to come up with any satisfactory way to interrupt the signal.

After researching other TC Systems I beleive an alternate method is this one:

However from this I am unable to figure our how to reliably interupt the injector pulse one pulse at a time. Does anyone have any suggestions?
regards,
Lee

 

[WGJ]

So, don't you just move 'your' traction control box to the earth side of the injector and run a command wire from the ECU to the TCU?

By the way, how long have you been working on this? CAN-bussing a MK1 Escort???


Bill

 

[ks0010938]

Hi, Im spent the last 3 years rebuilding the car and the project started in Summer. The digital dash was done the previous year:

I think I have found the solution to my problem:

 

[SIFLIGINIUS]

Hi ks0010938 !

I'm working on fuel injection cylinder deactivation.

I want to deactivate the ECU pulses with a simpe relay by the moment, later with a transistor.

But... if the 2.4 amperes of consume of 3 injector coils
(I want deactivate 3 cylinders)dissapear ...will not fried the driver chip ecu?

BTW , do you know anyone has work on my similar project?

Thankfully and your escort is v6 or 4l?

 

[EfiOz]

I think you'd be much better off trying to interupt the ignition commands anyway. TC runing on a fuel cut is not ideal as the torque reduction is not immediate.

 

[samplesecond]

Above and beyond the actual 'how-to' cut the injectors..you should closely analyze what is actually going to be happening.

As efioz mentioned, you would be better/safer off if you do an ignition cut (esp under force induction)--(Although a full 'injector cut' WILL result in immediate power decrease..)

The problem with this configuration is that you run a high risk of a lean condition in several instances..

The first is the phasing discrepency between engine speed/cycle, and frequency of wheel speed sampling-and hence, injector interruption...You will be safe as long as you implement something within your coding to elimate the risk of a 'partial pulsewidth'-for example, once an injector signal is interrupted, it cannot recieve another pulswidth signal until at least one whole engine cycle later..I can see at the threshold of the traction control interruption, you can very concievably have a whole host of 'partial pulswidths' as the trac system tries to do its thing..

The second scenario that will cause problems is if you have throttlebody injection, or if the injector is placed upstream in the runner far enough to where the volume of the runner between injector and intake valve is greater than the volume within the combustion chamber. Again, running the risk of a lean condition...

Good Luck!

-OS-

P.S.-got more details on the project?