AFR correct on gauge but not MS3
 

Hold on, this isn't the usual "my WB and MS3 are showing different values" post...

I recently installed an AEM 30-0300 WB to replace an MTX-L on my MS3 setup. After the install, I verified that the AEM and MS3 (in TunerStudio) were seeing the same AFR. Ran one track event with no issues. A couple months later, I had to redo some wiring to add a battery kill switch to the car. While verifying that everything still worked after the install, I noticed that when I start and idle the car, the AEM seems to read the correct AFR, but the value in TS is totally wrong. After the gauge finishes heating, TS either displays a constant AFR of 19.1, or sometimes, it jumps between 19.1 and ~15. During this time, the AEM reads a steady value of 12-14 (depending on WUE) and the car idles normally. All other gauges seem to read the correct values in TS.

I've verified that the AEM has good power and ground connections. Double checked the calibration table for the gauge in TS (set as custom linear WB with 0.5V=8.5 and 4.5V=18.5 per the AEM manual). I've wiggled around all of the wires between the AEM and ECU at every point where there is a connection/solder joint/etc., but can't seem to change the behavior of the readings in TS.

It seems to me like I've got a bad wire somewhere between the gauge and the MS, so my next step is going to be replacing all the wires between the gauge and the ECU. Are there any other troubleshooting steps I should try before I do that? It's odd to me that the car seems to run completely normal - if I wasn't looking at the reading in TS, I don't think I would have any idea that there is anything wrong. Also odd that the gauge seems to spend most of its time at exactly 19.1:1, which corresponds to the 5V sensor voltage exactly, based on my calibration table.

Tune and datalog of the car idling are attached. Thanks for any suggestions!

I ran the analog output on my 30-0300 for a few weeks until one day the tunerstudio reading shifted a full point from what my WB display showed. A quick look didn't reveal an obvious reason so I chose to switch to running it over can bus.

It's amazing. I'll never go back. Knowing it's always right is confidence inspiring.

A quick search should pull up a couple of threads that cover the setup.

Thanks for the advice. That does sound awesome, but based on the threads I found, it doesn't sound like something I'll be able to do before the race next weekend. Based on what you're saying though, it sounds like I may have a noise/interference issue on that input. I'll try soldering all of the wire connections (instead of crimps) on the WBO2 and sensor ground wires to the AEM and see if that helps at all. Would creating some physical separation between those signals and other power wires (ex. the switched +12V for the ECU) also help the noise issue?

If you're planning on running new wires anyway you should run the CANBUS wires and program the MS to read over CANBUS. All the fiddling you do with crimps and solders will take longer than it'd take to get it to talk over CANBUS.

But don't I need something like this to get it to work?

MSLabs CAN Wideband Module

Or does the MS3 have a CAN bus built in?

CANbus is awesome for accuracy.

MS3 has CANBUS built in and reading from that specific wideband is supported directly. You have all the hardware you need to get it working.

What firmware version are you running? The switch from 1.4 firmware, which is harder to get working, to the 1.5 firmware involved the introduction of "driving modes" which can create unexpected behavior if you're not familiar with them. This means that just updating from 1.4 to 1.5 may create issues.
Megasquirt Support Forum (MSEXTRA) ? MS3 firmware 1.5.1 released - use this (View topic)

Well that's pretty exciting. So based on the info in this thread, it sounds like I need to do the following:

1. Wire the AEM to CAN-H and CAN-L inputs (pins 3 and 4 on the DB37). Mine is a Brain-built MS3X. Does anyone know if Brain internally hooks up pins 3 and 4 to the CAN circuit, or do I need to open up the ECU?
2. Update to FW 1.5.1 (I'll check which version I'm currently running, but my guess is 1.4 or even earlier)
3. Follow the CANBus configuration settings that were suggested in the thread above.
4. Hope that engine states don't break my tune. I am not a tuner and my knowledge of how this thing works is very basic. I see the engine states explained on Page 76 here, and it seems pretty straightforward. Does anyone have some baseline engine state settings that are working well?

Anything else I'm missing? This seems pretty risky to do the week of a race... can I easily roll the firmware/tune back if it doesn't work?

I think you've got everything.
I have successfully rolled back firmware on an MS3 but am unsure about being able to do that in all cases. Back up your tune beforehand.

You seem to know well enough about tuning to be able to follow the directions for the set up necessary for setting states. Would you agree?

1. You most likely need to do the mods inside the case. I don't think Braineack wires CAN up.

You can down grade firmwares no problem. There is a boot jumper inside the case you may have to jump, but I am doubtful of that. I believe you should be able to upgrade and downgrade the firmware straight from TS now.

Engine states are sort of car specific. I think you will have to adjust some of those no matter what.

I haven't tuned engine states yet, but the directions look simple enough. Anyone have opinions on it? Is this something the OP could accomplish in an afternoon with some datalogs?

Might have to bug Bytevenom to help with this when he gets back to town lol. I like the idea of running WB on canbus. I hate fiddlin' with the calibrations...

Cool, thanks for the guidance everybody. I may try one or two more things to see if I can get the analog input fixed this weekend (I'm giving up on the race), but I do want to see if I can get this working over CAN as well. Seems like it would be much more reliable going forward.

If I wanted to switch to CAN afr with a UEGO and a DIYPNP I'd need that module though right?

Well I just went for it, and I think I got the CANbus working. I followed all the steps from my previous post, and did the FW update to 1.5.1 in two steps (1.3.4 -> 1.4.1 -> 1.5.1). Double-checked everything and then configured the CAN settings the same as this post.

Fired the car up and it was running pretty rich, but I just realized as I was writing this that it's probably because I didn't hook up the hose to the MAP sensor. :rolleyes: The AFR readings on the gauge and TS are dead on now, though! I still need to tune the engine states, but it seems like I'll be able to get the car up and running like this. Thanks again for the help everyone!

Tune attached in case anyone else needs to see the settings to get the CANbus working.

Can the AEM 30-0300 wideband communicate over CAN with the MSLabs MS3 Basic when it is updated to the 1.5.1 Firmware update? Or is this only for MS3Pro?

Yes it can and this is the exact setup I'm running. CAN AFR is the only way to go.