Just about the only serious glitch with the AEM EMS is that it can't drive the temperature gauge on our instrument cluster.

Apparently the signal out of the ECU is some sort of pulse or signal embedded with other signals to the cluster and the AEM can't reproduce it.

What I'd like to do is co-ordinate all the AEM EMS users to get a project together to investigate and solve this issue.

Ideally the solution would come from a member. MacGyver and Modifry come to mind as they have frequently shown ability with electronics. And Modifry has an SC so this may be of interest.

If you're an AEM EMS user or plan to be at some time, or have experience in the sort of electronics involved please express interest here.

 

Aus, this might be better in the "Electronics" forum. If you don't get the proper responses there, ask to have it moved back.

A link will be left here in UTH.

 

Yeah, I was in two minds about where to post. My thinking was that AEM users are probably more often here and most of the electronic whizzes seem to frequent both.

I expect this would be a doddle for any electronics enthusiast with a bit of experience but to make it a reality we need to get enough people wanting it.

 

GaryRudolph has some insight with this matter.

 

Will GaryRudolph share that insight?!

If the link between the ECU and gauge is entirely electronic, a microcontroller could easily snoop that datastream and pick out the data and recreate. All for a few bucks in hardware, albeit, quite a bit of work identifying the data structure.

I find that hard to believe, though. It would have been much easier for Honda to simply pass that signal through. I assume it goes to the ECU in an attempt to eliminate the second sensor that's needed so that the ECU AND the gauge can work, this way they both share the same sensor. It would be easier to share the signal rather than modify and regurgitate it.

 

I'd help if I had the time, but even something like taking the dash apart to check out the signals back there is a PITA... probably easiest to pull it directly from the ECU, anyway. If someone can scope out the ECU wire and let me know what kind of signal is there, I'm sure a solution is simple to think up.

 

Tomorrow I'll check out the wiring and do some comparisons between the stock ECU output and whatever the AEM is outputting and see what I can find.

What exactly should I expect to find? A voltage level (doubt it)? Pulse signal (more likely)? I have a multimeter and a handheld oscilloscope so I should be good to go.

 

Use the o-scope... if it's a DC voltage, the scope will show it, but the multi-meter won't show an AC signal. It could be anything, although some options are more likely than others. It could be a digital communications channel, but I highly doubt it. Something more reasonable would be a resistance change over a specific range, or a variable voltage level... possible a PWM signal, but that seems overkill for such a slowly changing signal as temp.

 

I'm not sure what you guys are thinking, but...

The stock sensor is resistive. Resistance changes with temp. You can find that data in the manual. The EMS will be set up to read that just like on any other Honda

What I am not sure of is what the stock ECU outputs to the gauge cluster. If there is a dedicated line for temp from the ECU to the cluster I bet you $5 it's the exact same signal. From what I hear it will not be, and the link between the gauges and ECU is a digital line. You will need a logic analyzer to capture and decode that data. This is probably best done with the engine off and key on, and using a pot to vary the signal fed into the ECU. This way you know all other inputs are static and you are dealing only with the temp signal. You can snoop and decode this data.

 

Does the AEM have an output for the ECT gauge that doesn't work, or does it not have an output at all?

I assume it has an input for the ECT sensor?