leds, ems car shuts off

This is my first post here. I am having a problem that doesn't make sense to me and I don't know how to narrow down. Here's a breakdown of what I have (since any of this could be affecting the issue.)

I have a 2005 Scion XB. I've added an AEM F/IC (Fuel/Injector Controller for the turbo) piggyback Engine Management System that connects to the stock ECU via a plug-n-play harness.

I've installed a Pioneer Avic-z1 with sirius, ipod, bluetooth, etc. etc. (they were all working when I started this project, but now the headunit doesn't respond. I can trace this, but thought I'd mention it.)

I added Autometer Nexxus gauges to the system and a Autometer tach that connects to the tach signal wire coming off the back of the OBD2 port.

I added a Tein EDFC controller.

I added a secondary oil pump that just has a +/-.

I installed LEDs in the interior, engine bay, and underbody.

I installed an Auterra DashDyno SPD that just connects to the OBD2 port. (Just like a ScanGauge)

The ACC wires of the gauges, shift light, and stereo hook into the fuse panel using add-a-fuse to a known(tested with multimeter) open ACC port.

The oil pump is in an add-a-fuse tap to a known(tested) open "Ign. ON" port.

The power (12v) for the gauges, shift light,  and EDFC run to a 10 AMP fuse at the battery + (this fuse is shared... more on that next)

There are four total wires coming off the side battery + post that share the 10 AMP fuse. One of these goes to the above mentioned. The other three individually go to 20 AMP switches for LEDs.

The LEDs are wired as such:
Switch 1 : 22 ga wire to fused power distribution block. 5 AMP fuses for each accessory. Slot 1 = single LEDs in vents, Slot 2 = Cathodes under dash/footwell, Slot 3 = single LED in cup holders.

Switch 2 : 22 ga wire to fused power distribution block. 5 AMP fuses for each accessory. Slot 1 = LED strips in grab handles, Slot 2 = LED strips in engine bay.

Switch 3 : 22 ga wire to underbody controller.

All of these items (except the radio and F/IC since they use harnesses) are grounded to a ground distribution block hooked to a factory chassis ground under the dash.

Everything was working wonderfully for about a week. Last night I tried switch 3 and it worked. I tried switch 2 and got nothing. Checked the fuses and they were good. Here's where it gets WEIRD and my main reason for the post:

With the car at idle, if I flip switch 1 for the interior LEDs, the car stalls!?!?

So here's a crude layout of how this is wired.
Power: Batt > 10 amp fuse > 4 (Four) 22 ga. wires, one of which goes to switch 1 that's rated 20 amp  > 22 ga. to fused distribution block > 5 amp fuse for each accessory > 22 ga. to accessory.

Ground: Ground distribution block to factory chassis ground.

Any help at all is appreciated. Thanks!
Jason

You have what we call an electrical nightmare.   :)

I would start by ignoring switch 2 - even unhook it at the actual switch.

Then, check to see if switch 1 still stalls the car.  If it doesn't stall focus on switch 2.

If it does stall remove one fuse from the distribution box on switch one and throw the switch again and see if the car stalls.  If it doesn't, you've identified the circuit that needs to be inspected.  If the car doesn't stall pull another fuse and repeat until all three fuses are pulled out.  If all three fuses are pulled out and the car still stalls you can inspect the wiring between the fuse block and the switch.

Do you have any indication as to why the car is stalling (no air, no fuel, no spark, etc)?  Have you tried bypassing the AEM and seeing if that changes anything?  I wouldn't think it would, since it is plug and play, but it may be worth trying.  At idle you shouldn't have any issues with running lean and damaging the motor, just don't boost the car with it bypassed.  I would also unplug anything from the OBDII port and seeing if that "fixes" anything.  The more things you can disconnect and isolate the easier it will be to discover the true problem.   

Sorry I did not answer your question. I dont think it sounds like anything to do with the lights does the service

light come on before it stalls or after?

Thanks for all of the suggestions! I followed Kevin's advice and tried to isolate the issue. After stepping through the current flow with my multimeter at different locations I found that my continuity was horrible on one of the wires feeding accessories on switch 2. I pulled it completely out and replaced it with a new wire and switch two worked again. I looked over the wire and found a "crimped" section with a tiny split in the insulation where it crossed the main feed of switch 1. I had wrapped every wire I ran in electrical tape (I always do this the full length of the wire to help it keep that OEM look.) except these wires for the switches. They are now wrapped. :) I tried them running without any drop in voltage and no noticeable difference in idling performance.

I'm taking a shot that it was causing some type of ground loop fault scenario and since the EMS relies on voltage from sensors to adjust injector cycle and timing it could have theoretically seen incorrect values when I sent power to that first switch.

All is working fine, no longer shutting down, and all lights are functioning. 

A couple of good points I'd like to make in case anyone else searches this topic.

About the OBDII conversation, any car running OBDII (any US car since 96 and some 94-95 models) can function completely fine with any kind of scan tool/monitoring tool plugged in to them. This is a one way data communication (unless you've got a lot of money in your tools) so data is being sent to that port, whether you are reading it or not. It will not change or affect performance of the car. (ScanGauge II is a great example of this. The Auterra a-501 I have is similar as well as a lot of palm/pocket pc/laptop software products.) Now, having to t-tap the tach wire like I did is a little different. If I had not made a good connection and severed the wire between the harness and the ECU then we'd have a problem. But, for anyone wanting to datalog, check codes, monitor fuel consumption, etc. it is perfectly fine to use the OBDII port for this.

As far as resetting the ECU, you should reset it anytime you add any performance part. (ie: intake, header, exhaust, pulley, turbo, s/c, etc.) I know that's a wide range and seems a little rediculous, but most newer cars have faster/larger processors in them. Rare are the days of finding 8 bit ECUs. They almost all are "Learning ECU's" that adjust different aspects of the vehicle based on your driving habits. It thinks it can make itself run wonderfully and save you some money on gas. For the most part it can. But, if you go changing anything on the car, give your ECU a break and let it start the learning process over. It will make a difference in the long run.

Anyways, now that I've got that out... thanks again for the quick responses and helping me stay on track guys!
Jason 

KPierson wrote:

---

It's common practice these days to use the OBDII port for all sorts of things - from guages to carPCs to various logging tools.  I've never heard of anyone having problems from doing this - in fact I drove on a three hour trip last weekend with my laptop plugged in to my OBDII port the entire time with no issues at all.

---

Come on people I am quite aware that you can plug laptops and scan tools into OBDll ports you can also

change the pramaters of the motor to get more preformance, But if you dont know what your doing its not a

good idea to attach anything to this plug. And I am not sayin' this person does not know what he is doing!!

I have seen it before all it takes is a kid watchin' the next import racin' movie a laptop and a program

and they think there a speedracer until they try to start there car then there parients call the shop to see when

we can get it in.

I don't see a problem with that scenario.  Kid watches TV, does something he sees as an example, and you get another customer. 

BTW, the next person who blames TV, Movies, Music, or Video Games for the stupidity of thier children are gonna get the royal smackdown.

OBDII connections can and will work.  If you know what you are doing, there's no harm in tapping into the OBDII.  It's not unlike anything else in a vehicle or anywhere else for that matter, if you don't know what you are connecting into you're gonna mess something up.

jayson33 wrote:

---

Come on people I am quite aware that you can plug laptops and scan tools into OBDll ports you can also

change the pramaters of the motor to get more preformance, But if you dont know what your doing its not a

good idea to attach anything to this plug. And I am not sayin' this person does not know what he is doing!!

I have seen it before all it takes is a kid watchin' the next import racin' movie a laptop and a program

and they think there a speedracer until they try to start there car then there parients call the shop to see when

we can get it in.

---

He said he put a shift light on the tach wire at the OBDII port and a "scan guage".  Where did he ever say he connected anything that could change any parameters of the ECU?  With what he connected there is virtually no risk of harm to the vehicle.

jdlong wrote:

---

I'm taking a shot that it was causing some type of ground loop fault scenario and since the EMS relies on voltage from sensors to adjust injector cycle and timing it could have theoretically seen incorrect values when I sent power to that first switch.

---

Just a quick question about your "shot" as to why things weren't working - weren't you switching the + of the wire through the switch?  If so, how could that cause a ground loop?  And, also weren't you using a power wire directly from the battery?  If so, the signal paths for all the engine sensors should be on seperate fuses/circuits and you shouldn't have been able to harm them.  I think there was something more going on, but since you got it fixed that is all that matters!

To address the post that matters, Kevin's, you are correct. Everything was on seperate fuses/circuits (original harness/added circuits). But, the grounds for everything were sent to the chassis ground. I dont think that you understood that +/- were not on seperate circuits. Just +.

A little clarification of the issue may help here. When I mentioned that "Switch 2's insulation was cut where it crossed the main feed of Switch 1", here's what I meant. The fuse panel I'm using is the 10 output model found here: https://www.steinair.com/fuseblocks.htm You can see on that the main feed is an open bolt with no cover. The bare wire must have been in contact with that.

To substantiate my "guess" I'll use something slightly out of context as it's the best way I know to describe it. Taken from Wikipedia: https://en.wikipedia.org/wiki/Ground_loop_(electricity) : Low current wiring is particularly susceptible to ground loops. If two pieces of audio equipment are plugged into different power outlets, there will often be a difference in their respective ground potentials. If a signal is passed from one to the other via an audio connection with the ground wire intact, this potential difference causes a spurious current through the cables, creating an audible buzz at the AC mains base frequency (50 or 60 Hz) and the harmonics thereof (120 Hz, 240 Hz, and so on), called mains hum.

I know that starting out I didn't run the circuit in the most efficient way, star, but instead some of  the grounds tie together way before hitting the distribution block. I didn't feel with the low current LED's that the issue could outweigh the trouble and miles of wiring I'd be dealing with. So, my theory relies on  when Switch 1 was turned on and power was sent down the line,  Switch 2's accesories saw 'some' power (never measured it, but it wasn't a solid connection obviously) which went through the accesories causing them to in essence have a 'bad ground'. If this differential was taken up by the other grounds it shared it could have caused what I mentioned above. I'm not a scientist, so don't hold me to it. The main thing is what you said: At least it's working now.

Regarding the other comments, I'm a 30 year old web manager for a University. I hold a HTI+ certification (Home Technology Installation) and have enough knowledge to pass the CEDIA cert. (Just never took it because of changing fields.), I was a professional car audio installer for four years and have personally built/fully restored seven cars starting with a '65 Mustang when I was just 10. I don't claim to know everything and value the knowledge of others which is why I joined this forum. I hope to be able to give back to the community. It's a shame my first appearance had to be a question.

I definately don't want to give the wrong impression to younger (newer to the auto-electrical world) people out there. That's why I explained in detail what OBDII was useful for with no risk (scantools), where risk can be involved (with enough money to have a programmer), and the riskiest (connecting to the factory wiring harness). What I'm doing is way beyond installing an ECU chip or a preprogrammed reflash. This is total management of several systems of the car. The only other step being standalone and rewiring the entire thing myself which is a little much for an XB.

Anyways, thought I'd clear that up. Which I can't believe I got a post about movies and no one even questions why I'm running a piggyback EMS on a box. lol... Again thanks for the help guys. Hope I can return the favor.

You had already answered the "why am I running a piggyback EMS on a box" question - turbo.  

I still don't buy the ground loop and here is why.  In your example you have to pieces of equipment - referencing differeing grounds so there is a difference in the ground.  In your car, the ECU makes sure this can't happen by providing sensor grounds to each dependent sensor.  Your MAF/MAP, IAT, coolent temp, cam/crank angle, etc sensors will not be tied directly to ground, instead they will have a ground wire coming directly from the ECU.  This will eliminate any chance for a ground loop and there is a reason why cars are built this way - so that ground loops won't make the car undrivable.

I've been thinking about this for a while now and I can't think of what might have caused this.  I've done some pretty stupid things to cars over the years (most of the time on purpose) and not many of them will stall the car out without popping a fuse or throwing a CEL.  Anyways, I hope the problem is gone and never comes back!  Any dyno numbers for the xB?

Not yet. And I'm not one to guess. But I've got most of the closed loop tuning done. Just a few Short Term Fuel Trims left from my logs to work on. Then I'll clamp the MAF voltage and the fun part of the tuning begins.

I wasn't aware that the ECU monitored grounds electronically. (kind of an ironic statement..) I just thought no matter what, the end result is your grounding to your altenator casing.

I guess the "kill switch" will always be a mystery. I'd rather it be a mystery and work again though, that's for sure.