S5 omp

[tweiss3]

So, my OMP just took a shit, and limp mode sucks.

I was thinking, is there any way to trick the ECU that the OMP is still there? I know when you do the blockoff, you have to leave it plugged in to stay out of limp mode. What exactly is the ECU looking at? Could it be jerry rigged into a harness?

[RETed]

It's s 4-pole stepper motor with feedback.
You'll need to fool the stock ECU into thinking that this is still the case.
You can easily fool the 4-pole stepper motor part with a bunch of matched resistors.
The hard part is to fool the feedback signal part.

With the prices of stand-alone EMS units coming down (under $1,000 in some cases easily), why bother messing with this?


-Ted

[Pete_89T2]

Are you certain it's the OMP is shot? I say that because the OMP is usually pretty reliable and because old/worn out OMP wiring harness issues are more likely and will also dump you into limp mode. The FSM has a few procedures to test the OMP itself and troubleshoot the system's wiring. Assuming you've done all that and verified that the OMP is indeed dead, your options are:

1. Pick up a used OMP & take your chances with it
2. Send out your ECU for the Rtek v2.x mod, which gives you the option of removing the OMP (don't forget to run premix!) and cancelling the limp mode, along with adding some fuel/spark tuning capability to the stock ECU, for about $450. If your mods are not too wild (<300 HP), and your stock wiring & sensors are in good shape, the Rtek isn't a bad investment.
3. Go with a stand-alone EMS, wiring harness & associated sensors as required.

[tweiss3]

Originally Posted by Pete_89T2

Are you certain it's the OMP is shot? I say that because the OMP is usually pretty reliable and because old/worn out OMP wiring harness issues are more likely and will also dump you into limp mode. The FSM has a few procedures to test the OMP itself and troubleshoot the system's wiring. Assuming you've done all that and verified that the OMP is indeed dead, your options are:

1. Pick up a used OMP & take your chances with it
2. Send out your ECU for the Rtek v2.x mod, which gives you the option of removing the OMP (don't forget to run premix!) and cancelling the limp mode, along with adding some fuel/spark tuning capability to the stock ECU, for about $450. If your mods are not too wild (<300 HP), and your stock wiring & sensors are in good shape, the Rtek isn't a bad investment.
3. Go with a stand-alone EMS, wiring harness & associated sensors as required.

I found a rebuilt for cheaper than new (about used price) but I don't have the cash right this second.

Originally Posted by RETed

It's s 4-pole stepper motor with feedback.
You'll need to fool the stock ECU into thinking that this is still the case.
You can easily fool the 4-pole stepper motor part with a bunch of matched resistors.
The hard part is to fool the feedback signal part.

With the prices of stand-alone EMS units coming down (under $1,000 in some cases easily), why bother messing with this?


-Ted

Thanks, for the replay, it makes a bunch of sense. What does the feedback signal look like? I'm curious now.

Anyways, I guess I have to weigh my options.

I could go standalone for less than the price of a new one from Mazda.
I could buy that reman unit.
I could make a simulator and delete it completely.

I already premix, just need to up the premix now and block it off.

[Pete_89T2]

Originally Posted by tweiss3

Thanks, for the replay, it makes a bunch of sense. What does the feedback signal look like? I'm curious now.

If you reverse engineer the FSM test procedure, it's easy to figure out that the OMP feedback loop is implemented with a potentiometer that varies it's resistance based on the stepper motor's position. The problem is the FSM doesn't elaborate on what that resistance is for each of the stepper motor's 4 possible positions - all it does is give you some min/max resistance values to measure across the OMP connector terminals when it's in a static condition (i.e., disconnected) to judge it as good/bad.

Also, the S5 FCs, all FDs & FE's ECUs vary OMP oil delivery settings as functions of not just RPM, but engine load, and perhaps some other conditions depending on the model year/ECU version. Without knowing precisely how Mazda mapped this function into the ECU (which sensor inputs matter, under what conditions?) it makes it very difficult to come up with a good feedback sim/stim to wire in and fool the ECU to stay out of limp mode. To reverse engineer this, one way might be to get someone with the same year/model '7 that still has a working OMP, and rig up a way to measure voltages across the OMP feedback potentiometer on a running car with a high-impedance DVM (high impedance so you don't effect the system's operation or the measurement). Stick the car on a dyno, and start data logging those voltages as functions of everything else you can possibly measure at the same time (RPM, MAP, etc.). Then you would have to analyze the data and try to reverse engineer the resistance vs. OMP setting curve.

Another thought might be testing by substitution. Get same '7 with working OMP, disconnect the 3 pin OMP feedback connector (I'm assuming S5 T2, don't know if the OMP wiring is same on FDs & FEs), and replace it in the circuit with a fixed value resistor. I'm thinking you get a bunch of fixed value 1/4 watt resistors that range in value from the min to the max values listed in the FSM OMP feedback test, start at the low end and work your way up to the high end. At each iteration, go for a test drive that runs across the car's RPM range under different loads. If/when it goes into limp mode, record (1) the resistor value you used (2) At what RPM it went limp, and at least have an estimate of engine load when it went limp - if you can monitor MAP (i.e., boost/vacuum gauge), that will work - high boost = high loads, deep vacuum = low loads. Based on the data you collect, it might be possible to design a sim/stim box to fool the ECU.

Of course all this crap is very time consuming, so the most expediant fix is just replacing the OMP with used/new after verifying that your wiring is still sound.

[RETed]

If you still want to further pursure this...
No need to systematically map the E-OMP.
The map is given in an SAE paper somewhere...


-Ted

[Viperx7]

You an also get the stage 1 for $140. It just takes out the eomp. If your stock or minimal mods, not too bad way to go. I did this on my 89 na. Pretty quick turn around.

[Boosted FC]

Rtek that shit. I have a handheld Palm if you need one for tuning purposes.

[tweiss3]

Ok, so I see Rtek Stage1 is what I am looking at for now.

In theory, even if the OMP pins on my ECU are fried (I don't think so because I parked it immediately) this will pull the car out of limp mode and ignore them, correct?

If that is the case, this is the route I am going in a few weeks.

[tweiss3]

Thanks so much for the help guys. I think I will be looking for that SAE paper just to be curious.

I'm going Rtek when I get the money, other car needs finished first