OMP: Prototype Tester

Casey · August 23, 2021, 08:42:17 PM

I experienced a P1688 error code earlier this year and my RX-8 went into a (very unpleasant) "limp mode". The error code pointed to the ECU detecting the OMP not working correctly on a start-up check. I checked all the electrical connections were OK, then checked the 4 stepper motor coils, and position sensor resistances were all OK, which they were.

To get my car back on the road quickly I decided to purchase a replacement OMP supplied by Boston Rotary, which fixed the problem. I recorded my DIY test and replacement of the OMP in a video, which needs some editing before publishing.

However, there seem to be an increasing number of reported failures of OMP's, so I was curious to see if they could be tested and maybe fixed. Some research led to several useful documents on the operation of the OMP, the most useful of which was this one, which describes the check sequence the ECU goes through to determine if the OMP is working correctly.

Casey · August 23, 2021, 08:43:04 PM

I needed a bit of a technical challenge at the moment, so set about putting together some hardware (based on the cheap Arduino controller) and some software to simulate the ECU test sequence. The worst bit was the small bolts attaching the OMP top cover were seized solid, so I had to use a Dremel to cut/grind them off to enabled me to check the mechanical arm inside was moving as it should. Here's what it looks like inside (not my pic in this case):

Casey · August 23, 2021, 08:43:53 PM

Here is the result of a few hours work (note I have not edited this, so please excuse DIY nature of the footage!):

https://www.youtube.com/watch?v=y9TyjN8NBdE&t=5s

Casey · August 23, 2021, 08:46:47 PM

The next stage was to convert this prototype into something that is portable and self-contained, so it can be plugged into an OMP in situ.

First, a breadboard version with display added for extra info:

Casey · August 23, 2021, 08:48:22 PM

And here we are, as of today.

Hardware now soldered onto small board and tested with the software. A Schottky diode added to provide simple protection against reverse polarity connection. A few bits to tidy up and a small "press to test" switch needed for the final version.

It looks like it will juuuust fit into the small black box pictured. A bigger one would make it easier, but I have a few of these lying around.

Casey · August 23, 2021, 08:53:37 PM

The slightly puzzling thing is that the unit I'm using for testing PASSES all the published Mazda tests; resistances all in spec and the stepper motor works and moves to correct positions. BUT, it's the unit taken from my car that the ECU kept reporting as faulty.

So, are Mazda doing any tests that are not published? Weird, as the unit appears fully functional (so why would Mazda need to do any further checks?), but it threw P1688 codes, despite resetting a couple of times...........

Mr Spiller · August 24, 2021, 08:44:26 AM

Shame I binned my broken one
You could have had that to test
Good work Malcolm

Sent from my iPhone using Tapatalk

boosted · August 25, 2021, 10:14:36 PM

Just 1 thing Casey...it has no oil init...could be stalling the stepper motor.

Sent from my SM-N950F using Tapatalk

Casey · August 26, 2021, 04:21:31 PM

Thanks Jon - unlikely I think, but I'll leave it on the "possibles" list.

If anything, the presence of oil would help to lubricate the moving parts in the oil volume control mechanism and ease any load on the stepper motor. On the bench, those parts have to moved "dry", with more friction, although, as this is a used unit, there will be an oil film still in there I reckon.

boosted · August 26, 2021, 04:26:23 PM

How about warming it up to engine temp..see if that effects it.
I lost 1.. then moved away from heat and no more probs. Clives has been under his dash for 60 races...

Sent from my SM-N950F using Tapatalk

Casey · August 26, 2021, 11:38:21 PM

Good thinking Jon. Again, worth adding to the list. The first failure was when the engine stalled a junction when hot.

Now, blowtorch or rig up a hot oil bath?

boosted · August 27, 2021, 12:27:08 AM

Old metal saucepan...bolt it to it and heat up water on a camping stove or such like.

Sent from my SM-N950F using Tapatalk

jinx · August 29, 2021, 04:34:19 PM

Quote from: boosted on August 27, 2021, 12:27:08 AM
Old metal saucepan...bolt it to it and heat up water on a camping stove or such like.

That's the sort of thinking that made this country great!

Eccentric shaft · August 29, 2021, 06:13:06 PM

Great video & impressive work!

Casey · September 06, 2021, 12:29:11 AM

The prototype is now complete and I hope to test a few more OMPs to see what results I get, trying to solve the mystery of why mine failed in service, yet passes all the published tests. I have an old saucepan ready to adapt for some high temperature tests too

I'm quite pleased with the final size and ease of operation - Sharpie pen for scale in pics

The loom is long enough to allow for connecting to the car battery (or any other 10-30v DC source that can deliver 600mA), and to connect to the OMP in situ. Reverse polarity protection is built in too