I noticed a new problem where the car cuts while I'm boosting hard. It feels like it's lean, but the logs indicate the crank trigger and wideband are dropping out. It cruises fine and it's even OK boosting not as hard.

Before I go any further, I have a Haltech Elite 1500, FFE Hall Effect crank trigger, and Haltech WBC-1 wideband. I checked the trigger wheel and there is nothing bent, and I measured the gap at what looked to be 0.7mm. I'm using a Haltech flying lead harness and I believe I remember the CAS wires being grounded to a pin in the ECU connector.

Now for some background. This issue seemed to pop up just within the last week or so. Although I haven't driven the car before that since about November. I haven't touched anything in the engine bay all winter, and a cursory look didn't reveal any damage to wires from rodents or anything. I can't find any blown IC couplers, no vacuum leaks to be found.

There are only a few things I can think of that it could possibly be related to.

1. Back in October/November, I left the car at my neighbor's house while we were doing some renovations. I usually have it plugged in on a tender but didn't do that at their house. I have an Anti-gravity ATX-30, and after sitting for a couple weeks I went to take the car out and it was dead. I jumped the battery with a lithium jumper and drove the car. It had an overvoltage condition and popped the circuit breaker, and also blew one of my SPA Techniques digital gauges. It has had an OV condition a few more times since then too, but it hasn't happened the last couple times I've driven the car. When the CB pops, the car stays running, but my battery voltage is all over the place of course. I just close the breaker and it goes back to normal and doesn't typically pop again during that drive. I drove the car again about a month and a half later and had to jump it again, but the battery remains topped off all the time now. I don't think I have any other issues than the CB popping occasionally.

2. Since the SPA gauge blew, I started making my own digital display using a Raspberry Pi and OneGauge hub. The hub connects to the ECU via CAN. I made a harness and verified the wiring is correct and plugged it in. I did not remember to assign a digital dash to the Haltech CAN system in NSP though, so I'm wondering if that might be a problem. I disconnected that harness though and drove it yesterday. It did seem like the problem wasn't as bad, but it didn't go away completely. Although this could be a coincidence that it wasn't as bad this time. I believe I'm doing a pull in 2nd gear mostly when it happens, and that's the gear I was in yesterday, not sure if that matters but I want to make sure I'm doing everything the same. I have the wideband, thermocouple box, and that OneGauge hub harness connected through the same CAN hub, so I'm wondering if not assigning the digital display in NSP would cause this problem.

3. I have most of the warning lights plugged into the OneGauge hub as well, and I wasn't sure if some of them were activated by 12V or ground. I went through the wiring diagram and most of them I know for sure, but a few others I just assumed. I'm wondering if something might be connected to the opposite and causing electrical issues.
parking brake - ground
seatbelt - ground
cruise control - ground?
fuel level - ground through rheostat
rear defrost - 12v?
left blinker - 12v
right blinker - 12v
high beams - 12v

Has anyone else encountered this before?

 

Can you easily rev up to say 7k rpm when in neutral? If it revs up with bucking and hesitation then your really have poor trigger signal, might be bad ground, or extreme noise in trigger wires.
What's the amp rating of your circuit breaker? It should not pop randomly, at all. Unless there is a short somewhere. It is also possible you have introduced ground loops with your extra ground points although this is hard to diagnose and prove, the cure is combine ground points together when possible.

 

I'll try to check that tonight and report back, thanks for the heads up!

It's 300a, so yeah, it really shouldn't be popping. It's only done this since the battery issue came up, or when I cranked the car for too long on times when it hasn't wanted to start.

I added a whole ground system about 12 years ago when I was diagnosing another issue back then. I have a 4ga wire running all the way around the front of the cabin, connected to a buss bar under the ECU and another one on the driver's side, and I connect all grounds from any electronics to those buss bars. The only thing I don't have grounded there is the ECU itself, which goes directly to the battery through its own dedicated wiring.

 

I just ran the car and held it above 7k, seemed to be fine, no sputters or hesitation. I logged it just in case though.

 

I'd be checking engine/loom earth's and connectors. If it's doing it on load it could be a combination of b
poor contacts and physical strain on wiring & connectors too.

 

Thank you! I'll inspect those.

 

Oh, I forgot to mention, the ECU had a CEL stored the next day, P1301 if I remember correctly. I erased it because I've had issues with a tire pressure sensor and I thought that was the code it put up, but when I looked it up on the Haltech website, it says it's a trigger reference error. But I haven't gotten that CEL since and I've had the problem still.

 

Check ecu power and ground too.

 

I was looking at the grounds and found an issue. I have continuity between sensor grounds and battery grounds in the ECU connectors. I have a connector between the battery ground wires on the engine harness and the battery ground itself, and when I disconnect that, I don't have continuity to ground on any of the ground wires at the ECU connectors anymore, but I do have continuity between the sensor grounds and the battery side of the harness, which means they're grounding through the chassis somewhere. Which is odd because I don't see any damaged wires, and I know I didn't wire any sensor ground wires to the chassis or any battery ground when I made the harness.

I disconnected all the indicator/warning light wires from the OneGauge hub and it didn't change anything with the grounding issue. So I guess that's good and bad lol.

I'm going to disconnect all the IGN1A coils tomorrow to see if it goes away, because I believe those connectors are the only place on the harness that grounds to the block. I'll probably take the car out with the OG hub wires disconnected just to see if the problem is still there.

Another thing is I have the battery ground wires from the connector going straight to the battery, and my tuner/buddy told me they should be going straight to the engine block or it will cause ground loops and other weird grounding issues.

 

Assuming you have sensor ground as the zero reference for the coil triggering? I think they were supposed to be separated on the coils from the coil main ground but perhaps they aren't so it's getting through the coils?

 

Maybe your problem is the battery itself? Who knows what a bad battery does to an aftermarket ecu and custom wiring... At this stage I think your best bet is excluding things one by one, like trying another battery etc.. A side note, 300A CB is too much (unless you have a big sound amp or so). I'd suggest a 150A one, and it would catch high current anomalies rather sooner.
Your ground loop situation should be fine as you connect grounds at a single point and preferably at the battery. At least your car used to drive fine, right? So something changed/broke/cut somewhere at some point.

 

I discovered tonight that it's grounding through the knock sensor. Apparently I've had the wires swapped this whole time, because the wire going from the knock sensor for sensor ground is connected to the terminal in the sensor that is grounded to the block. So I'll swap that tomorrow and that should completely solve the grounding issue. I wired the coil grounds to be separate, but I thought that might be the case too, that maybe I forgot and wired the sensor and block grounds together, but it was good when I checked it.

I'm planning on swapping back to an AGM battery anyway, mainly because I don't trust it now that I've seen what it's done so far. Yeah I need to swap to another CB anyway, so I'll get one that's 150a this time. Thanks for the suggestion! I had read somewhere years ago to get a 300a, but it makes more sense to have one that's lower amperage. I think 120a is the highest amp fuse in the car anyway. Regarding the grounds, I disconnected the extra ground wires from the battery tonight, leaving only the 0ga going straight to the transmission. We'll see if that helps any.

Another possible issue I found is odd wiring of the CAS. I can see the yellow and green wires coming from the CAS bundle and going into the connector, but the red and blue wires are clipped and sealed off with heat shrink. I don't remember doing that, but maybe I did. So it appears I wired it for VR instead of HE, which is what I have. For HE, you're supposed to use 12v (red), signal ground (blue), and trigger (yellow) wire. I ohmed out yellow and blue from the CAS connector to ECU connector. The green wire didn't ohm at all, but it did from the connector to B11, which is 12V fused power. Actually looking at it again, I think it's right. But I need to make absolutely sure. I'm not sure how it wouldn't be right if it's been driving this whole time like that. But yeah, it was working fine up until recently, so maybe it is the battery. Time to go shopping.

 

I assume if using a 150A breaker the starter isn't on that circuit?

 

The knock sensor pinout is a good catch, I've heard that some factory knock sensors have one of the pins connected to the body of the sensor which makes a connection to the engine ground.
https://www.hpacademy.com/forum/gene...sensor-wiring/

Regarding the crank sensor wiring, that is important to get correct. If the sensor has three wires it's usually a hall-effect or digital sensor, which needs to be powered (usually by 12V, sometimes by 8V or 5V). Most variable reluctance sensors (like the factory CAS for the FD) have two wires but some have a third wire for shield ground / chassis ground, I think I've seen this on German cars like VW or BMW.

I've seen most people on this forum use circuit breakers lower than 200A, but I agree that might not be the best way to do it. I've measured the starter current on my mostly-stock FD, with the factory battery location and wiring. A clamping DC current meter in the 'record max value' mode showed something around 200-250 amps, from what I remember That could be a brief spike of inrush current when the motor begins spinning, I didn't have a helper to crank the engine for me or record video of the current meter reading to see if it stabilizes out at something lower during sustained cranking.

 

Well, damn, I didn't think of that lol. I'm currently using a 300a, but it is on the starter, so I guess I'll keep it.

Totally missed it for two years lol. I copied Chris Ludwig's harness he made for the previous ECU, but I guess I mixed up those two wires somehow. I'm using a Bosch knock sensor like the one on the RX-8, and I verified that one terminal had continuity to the block. Thanks for the link! I'll check that out. Now I'm just hoping I didn't damage the ECU having it wired like that this whole time.

I have the battery positive going directly to an RX-8 series 2 starter. I haven't measured the current but I have a tool for that so maybe that's in order. Thanks for the idea!

 

It is ok to have a 150a cb. Starter motor inrush is too quick to heat up and trip the breaker. I have 150A at the battery positive terminal (not relocated), and many people have had 150a too from what I can see by a quick search in this forum.

Good catch on the wrong CAS wiring.

 

That's probably a good thing then. I'm used to dealing with mining/heavy industrial protection regimes now where switch timing is within a couple of AC power cycles, the LV DC stuff is obviously a lot more forgiving, unfortunately not enough mobile plant has full battery currently protection, seems silly to lose millions of dollars for a hundred dollar part.

 

My 300a breaker trips if I crank for more than about 15 seconds at a time, and it won't reset for about a minute. When I'm sitting in traffic, that's a huge turnoff lol. I might switch to a 150a, but I'll swap back to 300a if it doesn't work out well.

I just kept checking continuity and unplugging connectors until it went away lol. I'm just glad it wasn't the harness running over the passenger front wheel for the turbo speed sensor, boost control solenoid, and pre-turbo air temp thermocouple grounding out against the chassis. The knock sensor being the culprit did surprise me though, hadn't even considered that.

I contacted Chris Ludwig to ask about the knock sensors he installed, and he said to switch to a new version, Bosch 0261231188. He said this one doesn't ground to the block and it won't matter which way it's wired, so I won't need to switch the wires around in the connector, just swap out the sensor.

I read the link that Scotty posted, and watched a couple youtube videos and read the comments. I probably could have run just the signal wire and been fine, but I'll swap anyway. Some had said that the isolated (not grounded) sensors have a polarity and need to be wired a specific way, but that's not what Chris said about this one.

 

I got the new knock sensors today (thanks Amazon Prime!) and verified they don't ground to the block. I installed one of them, connected everything, started putting the UIM back on, and before I completely bolted it down, I tested again to make sure there wasn't continuity. YEP! IT'S BACK! LOL

I disconnected the knock sensor again, still there. I disconnected any other electrical connectors to the UIM that I reconnected, STILL THERE. I checked last night after reconnecting all the other connectors in the engine bay and in the cabin, and had no issues. I'm losing my mind here lol.

 

sounds like a ground issue to me, i almost never see over voltage charging issues. but how much are you seeing? if we're talking like 19+ volts you could be damaging electronics.

that to me would want to be figured out before continuing to drive the car, it likely is a symptom of your erratic electronics issues.

 

I'm not sure what voltage gets to because the CB pops before I even notice it, although it hasn't happened the last couple times I drove the car. Either way, I'm planning on swapping the battery as soon as I find the one I want to swap to.

 

Here's the issue most likely when it comes to your grounding scheme:

An OEM RX-7 REF and SYNC sensor are a variable reluctor type. That means they have 2 wires. One is to REF+ (Trigger) and the Other to REF-. Then you have the SYNC which is + and - (home). The shielded part of the cable grounds to sensor ground. It would be spliced into the sensor ground pin on the ECU. It shares this with all of your 5V sensor types.

If you are using a Hall Effect sensor like FFE it jumps to a 3 wire sensor. You'd then have 12V, Signal, and Sensor Ground. You would NOT have a wire running to REF- or SYNC-. Again the sensor grounds would be spliced together and run to the same pin as before.

If you have continuity between your engine/battery ground on the engine and sensor ground... that means you have a wire either spliced incorrectly or you've wired something wrong. It will also ruin your signal because the sensors (all sensors!) are using the sensor ground as a reference. You can also do damage to your ECU.

The knock sensors on a factory car are single wire and are not wideband. If you have aftermarket wideband knock sensors, they are a 2 wire setup. On a Haltech, the signal wire is wired into the Knock pin (knock 1 or 2 etc). The second wire is run back to the ECU along with the shield to sensor ground pin. Again these are all spliced together with the previous sensors etc.

It is important to note that the shield MUST be grounded at only one end of the harness and that is back at the ECU. Otherwise you create a 3rd wire and the shield is not shielding the electromagnetic signal generated.

 

Oh I sorted out the CAS wiring already and it's good. I was just mistaken when I first started looking into this and thinking I wired it for VR because I couldn't remember how I wired it. I ohmed it out and I also found a pic from when I built the harness showing which wires I used. That's not to say that I don't still have the issues with the CAS and wideband dropping out, but I imagine that has to do with the grounding issues, or something else entirely.

I did track down the current problem tonight though. It's my CAN cable going to the OBDLink OBDII cable I'm using. As soon as I disconnect the OBDLink the continuity goes away. And that really sucks because it's the only thing I have for gauges right now lol. I'm guessing if I switch back to a bluetooth OBDII adapter it would probably be OK since it's not hard wired, but I haven't tried it. I'm not a fan of that fix either though since I switched to a cable connection due to the bluetooth disconnecting numerous times while driving.

So not only was the knock sensor grounding to the block, but the OBDII cable is grounding somehow as well.

 

I put my bluetooth OBDII device in the car and it didn't cause any issues with continuity, which is probably obvious. It did connect, but I had issues with it later in the drive with disconnecting and not reconnecting.

I didn't have any boost issues at first, so I thought the problem was fixed, but it did come back. I should probably note though, I only have this problem in second gear when I start boosting from a low speed, if I boost in 2nd any other time, or in any other gear, it pulls just fine with no issues. Actually I don't know about 1st gear, I don't remember if I've tried that. But I can't figure out why it's only happening in 2nd gear and only from lower speed.

 

I took the car up to my tuner buddy in the Denver area yesterday. He said after his first drive that it was happening in every gear, not just 2nd. We tried several different things with no luck. We verified wiring again, and voltage and ground. We swapped his HE sensor from his car to mine, and at first it looked better, but the next run it was back to what it's been doing. We swapped his sensor standoffs onto mine because they're a bit shorter, no luck. We added different thickness washers to the standoffs, no luck. He made some other changes in the calibration but those didn't work either.

We thought it might have been the TPS, but he checked that and the wiring and all that was good too. One of the trips out it threw an error code again for trigger reference error. On the way home I noticed that the wideband read in the 6s when I was pulling away from a light, and looking back at some logs today, it showed in the 9s as it dropped out and the car shut off, so I'm going to replace the wideband sensor and hope it fixes it. We're lost at this point.