nienberg.11

This is just what I need. I looked up that thread on ecomodder: http://ecomodder.com/forum/showthread.php/how-get-instant-fuel-consumption-megasquirt-5474.html

I'd also like to see MPG info in Megatune. I noticed that the speedometer gauge tries to guess my speed, but of course it's wrong since I'm not running any speed input to the MS. If Megatune's speedo could somehow be calibrated for the car's highest gear and thus read correctly in cruise, the MPG gauge should also read accurately in cruise. Does anyone know if it's possible to enter a couple of RPM/Speed pairs to calibrate the speedo for one gear?

Thanks for the suggestions z-ya and softopz, I've figured it out. I used the method I described earlier and ran my tach off of the negative side of the coils. I disconnected the fuel pump and verified that the tach would operate during cranking. It showed 200-300 rpm while cranking, matching the megasquirt. Having done that, I hooked up the laptop, took the car for a drive and waited for it to stall. Once it did, I hit the starter and confirmed that not only was the MS getting no tach signal, the tach was showing no rise while cranking. Therefore, the coils were not firing and this was not a PIP/SAW communication issue between EDIS and the MS. That meant that the problem was either in the VR sensor, the EDIS module, or the wiring between the two. The wiring checked out, so I replaced the VR sensor this morning. I've driven the car for about an hour total since then with no further stalls. While having a timing light would have narrowed this problem down in the same way, I'm on a fixed budget and had the tach diode circuit ready to install. It made no sense to buy more equipment when I could use what I already had available to me. There was no reason to suspect that the timing had changed, as the car ran normally right up until each stall. Also, using this diode circuit instead of the tach output on the EDIS module will let me know instantly if I have spark or not when diagnosing future issues.

I'm not sure what part of "the timing has already been checked" isn't getting through. Let me make myself perfectly clear. After installing EDIS, I used a timing light to verify that the actual advance matched what Megasquirt was showing. In my current situation, I could blow money on a timing light I don't need, or hook up my tachometer which I need to do anyway, to get the same diagnostic information. As for cleaning my db37, I'll use whatever works. As long as you aren't concerned about the finish of plastic parts, brake cleaner is harmless to them. And yes, I use conductive grease on my electrical connections to prevent corrosion. If you layer some over the female side of the db37 and wipe away the excess, all of the pins get a small amount and cross talk isn't an issue. Thanks for these redundant warnings. I know you're quite the MS/EDIS guru since you've had yours up and running for a whopping three weeks.

Rolling, - The timing has already been checked with a timing light. I just don't own one. - If you read my last post, the diode circuit can't be the issue since it isn't in the car yet. - I read over that link, and it says that grounding the cap 'may be safely omitted.' There's nothing to suggest that having it grounded causes problems. I'm fairly certain my coil pack doesn't have that cap anyway.

Z-ya: I have had broken solder connections on the back of the db37 in the past. I'll take a look, but the fact that the signal comes and goes makes me think it's something else. Rolling Parts: That's interesting. I had always heard that EDIS modules were nearly infallible. I don't have a timing light, but I had the same thought. I just built the diode circuit to run the 280z tach from the EDIS coils. When I get that installed, I can check to see if the tach still registers rpms as the car coasts to a stop after the MS cuts fuel. If so, the EDIS is still firing and I have a PIP/SAW connection issue. Otherwise it's the VR or the module.

Good thinking, z-ya. I sprayed the db37 out with brake cleaner and coated the pins in conductive grease. Unfortunately this wasn't the problem. I took the car for a drive and after 10 minutes it stalled. 3 minutes later, it started back up. Another 30 seconds and it was stalled again. 5 minutes after that it was back up. I headed toward home and started datalogging at that point. Right as I pulled into my space, it stalled again. I'm still thinking the VR sensor may be faulty somehow. I've attached the datalog in case it might shed any light. The interesting portion is at the bottom where the rpms go to zero. rpm cut-out.zip

First, let me give some background on my setup. I run MSnSE with EDIS. The car ran fine for weeks on this setup, but lately it's been stalling out unexpectedly or not starting. Whenever the car stalls, it starts back up after sitting for a few minutes. After using a laptop to diagnose the stalling, I saw that the MS was losing the rpm signal from the EDIS module entirely and cutting fuel. I double checked the clearance of the VR sensor, and it's at .020" from the toothed wheel which is well within the normal range. I also checked for corrosion in the connectors at the coil pack, VR sensor, and EDIS module. They all checked out. I'm almost stumped at this point. My only thought is that the VR sensor may be damaged. It was knocked around pretty hard when my toothed wheel broke loose from the crank pulley at one point due to poor welding. Has anyone else had a similar experience? Do you have any insights as to what would cause this intermittent RPM signal?

I did the flyback mod to Q9 and Q12, and added a noise eliminator to the MS power source. The car runs excellently now with no stumbling or lurching. I re-logged the same route as before. Here are the results: As you can see, it's still not perfect but it's improved enough that I'm not going to worry about it. It seemed like those two lone spikes must have been disregarded by the MS, because there was no roughness of any kind during that run, or later when I took the car on a longer drive.

Thanks again Datman. I looked for the rising/falling edge setting and couldn't find it anywhere in Megatune. I'm thinking that might be something that shows up with Megasquirt II and not MSnSE. While I'm posting, here's a graph of my RPM signal from a datalog I took last night:

DatMan, The link that HizandHerz posted isn't working anymore, and now I'm looking into that noise mod again. The fix that I used didn't completely solve my problems and I still have some roughness and voltage spikes. If you could point me toward the injector noise mod, I'd appreciate it.

Thanks Datman! The car isn't available for me to look at right now since I'm at school, but I'm pretty certain that I had all of the opposing cylinders wrong. The rising/falling edge setting is most likely part of my woes as well, since I followed the instructions I found to the letter. Thanks for the offer on the plug pinouts, but I'm pretty sure I have that part right.

Guys, I recently got my Z running on EDIS, but there are a few kinks that need to be worked out. First, there is a discrepancy between what the Megasquirt/EDIS online instructions say about how the plug wires should be arranged on the coil pack and how HybridZers like Z-ya have found it to actually work. Since my car runs I assume that I have the firing order correct but I was unable to find which cylinders are 'opposing' on the L28. To settle this issue once and for all, could someone please fill out the following picture with the cylinder numbers that correspond to each set of coils... Secondly, when using an inductive timing light to check my timing, I did an experiment and used the light on each of the six plug wires. Obviously the light would only read my timing from cylinder 1, but what concerns me is that on three of the cylinders, the light flashes only about 1/4 as frequently as on the other 3, making me think that I'm not getting nearly as many sparks on those cyls. Anyone have any wisdom here? Finally, the car runs well at idle but poorly on the road. It's very weak and stutters constantly. The MS shows occasional spikes in RPM and voltage. In reality there are a lot of sketchy aspects of my setup that may have gone wrong here, ranging from injector current limit spikes reacting badly with my new tach input circuit, to the fact that I have both my PIP and SAW within the same shielding. There's also the matter of my plug wires nearly touching my new EDIS harness because of where the coil pack is mounted. Any help here is appreciated.

Here's what I did: I left all of msextra's EDIS output mods in place (shown above). The saw didn't work right without them. I then compared msextra's EDIS output mods with the hall circuit assembly, step 50a from the MS v3 assembly guide at megamanual.com/ms2/v3assemble.htm That led me to these steps to revert to the normal Hall input from msextra's circuit: -Leave C30 out and remove the jumper between it and S12C -Jumper C1 and C2 -Install R12 -Remove the jumper between TACHSELECT and XG1 -Jumper XG1 to XG2 -Jumper TACHSELECT to OPTOIN -Jumper TSEL to OPTO OUT -Install U3 -Install C11 -Install R13 This left me with a Hall tach input on DB37 pin 24 and a SAW output from LED17 through DB37 pin 36. I'm considering adding in C12 for noise reduction as well. I'm still getting lone tach spikes here and there.

It turns out that it had nothing to do with the injector drivers. The tach input circuit that msextra.com instructed me to build was the issue. I kept the SAW output mods but undid their input mods and went back to the standard hall input circuit. This cleared the issue up completely.