[tryingbe]

Hey guys I've got everything plugged in and running well with the Arduino IMA delete, minus I have the p1646 code. Double checked all my connections. Any tips? Which wire specifically? Pin 12 out of the Arduino? Red/yellow what where? Do I need to update the code?

P1646 is MOTSTB (YEL/RED) signal not present/signal error.
Look at the wiring diagram.

 

[dsteele713]

NO. Take them all off!

http://www.thelostartof.net/tryingbe/insight/imabypass/imabypassinstalled.jpg

I'm finished with the first part of the build (arduino w/ soldered connections on harness so I can revert back to stock at a later date, but I was having trouble finding parts for the fused DC/DC high voltage loop. Home depot had some 250V fuses that looked like they were intended for AC current, but I assume as long as they handle the voltage then it's fine. What I didn't see was that fuse case. Where did you get yours from?

Also, for those wires you made, is it just standard ring connectors with some heat shrink tubing over the connector?

 

[mudder]

Your assumptions on fuse ratings are incorrect. Breaking a DC spark gap is MUCH more difficult than an AC, because the AC signal has numerous zero crossings per second (e.g. 120 per second at 60 Hz). Each time the voltage crosses zero, the spark will extinguish, which will allow the ionization in the air to disperse, which will increase the breakdown voltage. DC doesn't get this benefit, which is why most high voltage DC fuses have an infused silica layer. In short, DO NOT USE AC-ONLY RATED FUSES IN DC SYSTEMS!

Use the DCDC fuse (you already have) that's bolted to the junction board. It even has eyelets you can directly bolt to. Place it in a PVC pipe to keep it insulated.

 

[dsteele713]

Your assumptions on fuse ratings are incorrect. Breaking a DC spark gap is MUCH more difficult than an AC, because the AC signal has numerous zero crossings per second (e.g. 120 per second at 60 Hz). Each time the voltage crosses zero, the spark will extinguish, which will allow the ionization in the air to disperse, which will increase the breakdown voltage. DC doesn't get this benefit, which is why most high voltage DC fuses have an infused silica layer. In short, DO NOT USE AC-ONLY RATED FUSES IN DC SYSTEMS!

Use the DCDC fuse (you already have) that's bolted to the junction board. It even has eyelets you can directly bolt to. Place it in a PVC pipe to keep it insulated.

Wew, glad I didn't actually put it in, then. I was trying to avoid doing anything to the system that would be a pain to fix later, but I guess replacing that fuse will be pretty easy. When I go to take the fuse off, is there anything hot anywhere on the junction board once you've flipped the breaker switch on the battery? I'm pretty sure I've seen a "no" to that question somewhere, but pointing google at IC.net isn't returning any answers, and I like not being dead.

 

[mudder]

Flipping the switch to "off" breaks the pack voltage in half. Since half the pack voltage exceeds 60 VDC (the point above which DC voltage can break down the insulating properties of human skin), then yes there are still hazardous voltages on the junction board.

When you look at the junction board, you will find that there are four bus bars that bolt to the actual battery cells. Two of these bus bars go to the switch you turned off. The other two are the pack negative and positive. Since the entire pack voltage is isolated from the car frame, when the switch of off, that means you can safely, simultaneously touch any of the following without harm:
-Any ONE (single) conductor on the junction board & the car frame, OR;
-both conductors on either side of the switch, OR;
-the pack positive & negative terminals, OR;
-the pack positive & negative terminals & the car frame, OR;
-the pack positive & negative mid-stack conductors, OR:
-the pack negative & positive mid-stack conductors.

However, you CANNOT touch any of the following:
-the pack positive & the positive mid-stack conductors (this will apply half the pack voltage across your body), OR;
-the pack negative & the negative mid-stack conductors (this will apply half the pack voltage across your body).

As you can see, flipping the switch off makes the pack safer, but you CAN STILL ELECTROCUTE YOURSELF if you touch certain wires simultaneously. I like to remember it this way: In a properly functioning car that isn't throwing IMA error codes (most notably IMA insulation breakdown), flipping the switch to off allows you to touch the body (ground) AND up to ONE conductor on the IMA pack without electrocuting yourself. Therefore, with the switch off you can place a metal socket wrench on ANY bolt on the junction board without electrocuting yourself.

It's still best practice to initially touch the conductor with the back of your hand/finger... that way if you do get zapped your muscles will tense and pull away from the conductor... if you grab a socket wrench and put it onto a bolt (and someone don't follow one of the rules above), then you will not be able to let go of the wrench, as your muscles will contract due to the current flowing through your body. You be dead.

 

[dsteele713]

Edit: Decided to just start from scratch and cut the harness. Not worth all this headache when there are harnesses you can source from junk cars if need be.

OK, I'm getting a persistent P1640 IMA code, which it seems like no one else in the thread has, which seemed odd, so I wondered what I was doing differently. As mentioned before, I'm trying to tap into the existing harness rather than cutting it, because I want to keep it intact. So my solution was to pull the pins out of the connectors, solder on a small wire towards the back where there was a little extra space, and slide the old wires back into the connector. For the extra pins that aren't part of the new "harness", I just pulled them and wrapped them in a bundle in saran wrap and electrical tape.

I did that for all wires on the schematic, then soldered the other end to the respective holes on the board (I'm using a mega 2560 because that was what was on hand, but it doesn't look like that should matter). I then checked continuity from the pins on the board to the connectors, from pins on different connectors to each other, and then also checked that one pair of pins to verify a short. As far as I can tell I have full continuity from all pins to each other and to all board pins as the schematic calls for.

The board appears to have been flashed correctly (I've got an LED blinking maybe 5 times every 4 seconds), but when I plug this in and turn the key the only thing I get is what sounds like the fuel pump. The starter doesn't turn over at all. Note that it was working fine with the standard "unplug the BCM" bypass.

My first thought is maybe I damaged the arduino board when I was soldering, so even though the pins have continuity maybe a chip is burnt out or something. Another thought is maybe I was incorrectly assuming that the extra pins not used in this setup couldn't be connected to anything, so it was safe to have them all wrapped together. But there are some extra connectors that look like they might be splitters? If that's the case then in putting the rest together I may have been shorting something that shouldn't have been shorted.

Any thoughts?

 

[mudder]

Did you get it working?

 

[dsteele713]

Did you get it working?

Well, the car isn't starting, but the CEL goes away after several seconds. Is this the expected behavior since the board is powering on after the car starts? If so, then it seems like I finally have a correctly-working IMA bypass, but managed to fry something with my fail-harness.

 

[mudder]

Did the car start before? Difficult to answer the remaining questions until we know the answer.

 

[dsteele713]

Did the car start before? Difficult to answer the remaining questions until we know the answer.

Yeah, it was running on the 12V starter just fine. I had pulled those two BCM plugs for a standard battery bypass several months ago. Was running when I parked it to do the procedure. At this point I'm going to follow the owner's manual for diagnosing a car that doesn't crank. My first guess is the ECU isn't closing the starter cut relay properly, but I should know for sure in a few hours.

 

[dsteele713]

So I checked a few things under the dash. The stuff I've checked looks OK, but still got plenty to go. Just to make sure, I ran a little makeshift jumper wire down to the solenoid, and the car started right up on contact with the battery terminal. It was nice to see a dash with no error codes for the first time in a long time. Anyways, gonna pull it into the garage and see if I can check some more continuity.

 

[mudder]

It's possible if you mis-wired the bypass that you connected 12 volts to digital logic inputs (0:5 volts). If that happened, then there's a real chance that you damaged the ECM. Keep us posted on what you find. Note that you can remove all the electronics from the IMA bay and the car will still start (as long as charge remains in the 12 volt battery).

 

[dsteele713]

Keep us posted on what you find. Note that you can remove all the electronics from the IMA bay and the car will still start (as long as charge remains in the 12 volt battery).

After ruling out pretty much everything else, I bought a new ECM at a local junkyard. When I pulled the old one, this problem at least was immediately obvious:

Pins 1 (STC), 2 (ENGRDY), and 12 (IMOLMP) on ECM Connector A all had significant corrosion. I do remember reading a thread where someone had a similar problem at one point, so this is apparently a thing. I sincerely doubt it was the cause of my current problem since that corrosion has probably been there a good long while, but it does explain why sometimes the relay used to click a few times before the car would start even though I had replaced all ground wires and thoroughly cleaned all contacts.

The STC pin is what closes the starter cut relay so your car can start. Even though I don't believe this was the primary cause of my current complete inability to start, the clicking was annoying, and I'm sure if the corrosion were bad enough it would have kept me from starting the car altogether.. If anyone is having starting problems, you may want to visually inspect your ECU pins, because I'm at least the 2nd guy that had this problem.

I couldn't see any evidence of damage on the board, but there are a lot of tiny components on there, and the damage from an overvoltage event might not be visible anyways.

So while I was putting my new ECM in, I managed to set off the car alarm somehow. I think something must have been busted in the old ECM because in the 8.5 years and ~140,000 miles I've had this thing, I've never had that car alarm go off once. I've even pressed the panic button and gotten nothing.

Anyways, I rushed to turn the car on to cancel the alarm. Decided to try starting the car while I was there, and I got nothing. The ECM wasn't properly secured yet, but everything was plugged in, so it should have been OK, right? Except I forgot about the immobilizer. Correct me if I'm wrong, but the key is paired with the ECU, right? So swapping in a new ECU means I need to either desolder and swap the immobilizer chip, or add on an immobilizer bypass chip. Is that about right?

Lol, this has been a headache, but I totally brought this on myself. Always double-check your wiring, kids.

 

[Rainsux]

You are correct on the immobilizer chip.

Sam

 

[dsteele713]

Got it working! Came across this thread about modding your starter relay so you can start from the 12V system instantly (I modified the starter relay, IMA bypass), realized from the starter circuit diagram if I just did the relay bypass this would cut the ECM out of the circuit entirely and tie the starter straight to the ignition key position. Pulling the relay, popping it open, and wedging something in there to hold the relay closed was like a 10-minute operation. I popped the modded relay in and got immediate ignition. Now I just need to drop my battery off in storage for a while and I'll be 80lbs lighter and good as new. Thanks for the help, guys!

 

[Drtye]

I know i had to bypass the starter relay in the under dash fuse panel. The down side is that it also bypasses any clutch safety switch for the starter, but at least my car cranks.

 

[mudder]

You don't need to bypass the starter relay for this to work... you just have to hold the key down for 5 seconds.

 

[BabyDiesel]

I’ve almost got everything complete with the bypass, except that I cannot get the code to the Arduino. My computer will not recognize the board when it is hooked up. The “ON” light on the board goes off a half second after plugging it in. The board is warm to the touch near the plug in, so it is receiving power.

Any help would be appreciated!

 

[mudder]

If the board is warm to the touch, then something isn't connected properly. You've probably shorted out the power rails, hence the reason you can't communicate with it.

 

[tryingbe]

I’ve almost got everything complete with the bypass, except that I cannot get the code to the Arduino. My computer will not recognize the board when it is hooked up. The “ON” light on the board goes off a half second after plugging it in. The board is warm to the touch near the plug in, so it is receiving power.

Any help would be appreciated!

Sounds like the board is burnt.