This is something that's been bugging me for a while, and I think I may know the answer, but I thought it'd be worth asking others before I throw time+money at it.

Some time ago, I discovered that the 12v-ign wire that goes to the ECU (and also supplies the injectors and various 12V sensors/solenoids in the engine bay) in the factory wiring harness on my 97 Miata yields a lower voltage than the battery or system voltage. When the car is not running and battery voltage is 12.6 (tested not only at the battery, but also fuse panel and various other 12v connectors), the voltage on that particular circuit is closer to 11.8 volts. Similarly, when the car is running at 14-something volts, the voltage on that circuit is often 13-something volts.

Is this part of some intentional voltage regulating circuit for protecting the ECU, or do I have some excessive resistance somewhere (likely the main relay)?

Thanks!

 

What is the voltage at your injectors?

 

I didn't check it again today, but I seem to recall it matching my readings at the ECU the last time I tested it. Will try to test again soon to verify though.

Assuming that it is also lower, what are your assessment/suggestions? For that matter, if it is higher, what are your thoughts?

 

No. Ideally, the voltage on the white/red wire at the ECU (and the injectors, and the MAF sensor, etc) should be within a few mV of what is observed at the alternator / battery.



Stepping through the wiring harness outwards from the battery, one connection at a time, using the diagram on page 1 of this document: https://www.mediafire.com/?6086wa899380jt8

Also, test it with the engine both running and not running.



That your testing methodology is flawed, or that you are insane. Flawed would include being incautious / inconsistent as to the selection of ground points between tests.

 

OK, good. That pretty much confirms my expectations, which also indicates that my most likely culprits are either the fuse or the relay.

Similarly, I've discovered that there is a very minor voltage on that circuit (~0.8v) when the key is not in the ignition, which leads me to suspect an issue with the relay since I would not expect there to be any voltage until the key is turned to the ON position.

I will double check voltage at the injectors and other sensors to verify that it reads low there as well, and if the numbers match I will see about swapping out the fuel inj. fuse and/or relay.

 

That's odd, though having never measured the white/red circuit with the key off, I can't say whether it indicates a fault. I would not *expect* there to be any voltage there at all, but I can't say that definitively.

I would not suspect the relay in this, even slightly. Mechanical relays, unlike their solid-state counterparts, don't leak. They may exhibit high on resistance if the contacts are funky, but they don't pass trickle current when off.


I know that it sounds cliché, and I hate suggesting it because it's usually a red herring, but I'd look very closely at the condition of the three major ground straps- one from the engine to the body on the exhaust side, one from the PPF to the body near the rear on the tailpipe side, and one from the body to the battery inside the trunk. Pull 'em, clean 'em, flex 'em and make sure they're not used up, re-attach 'em with some grease.

 

verify they match. if they do, ignore it. there's always a votlage drop between the battery reading and the system voltage.

once you install your MS3, you should calibrate the voltage so that it matches the voltage reading at the injectors. Everything is based on your voltages as seen at the injectors.

 

Thanks for the advice, Joe. I am going to go ahead and swap out the fuse (because I've already got a handful) and the relay (because it's fairly inexpensive), but much like you I've never seen a relay behave in such a manner so I'm doing it more for peace-of-mind than expecting a home run. I will be sure to check the grounds, particularly at the engine-to-body strap since that's another of the potential culprits that I've considering as well.

I verified readings today, and indeed that entire circuit (post fuse+relay) has the lower voltage, as well as the small amount of voltage when "off". The MS does read voltage slightly lower than that circuit tests at, so I will get my hands on a second multimeter to verify correct voltage before attempting to set a calibration change on the MS3.

---

EDIT: My brain kicked in and I decided to do some more probing.

With the relay removed, I get 12.59 volts from the hot lead of the relay to chassis, but only 11.83 volts from the hot lead of the relay to the load side of the relay. Obviously one might expect some voltage drop as it's surely going through some resistance and such through sensors and/or circuits, but the fact that this matches the voltage at other points on the circuit even when powered up makes me think there may be a ground issue or unexpected resistance somewhere.

Additionally, I check resistance from the relay load side to the various connectors for things like injectors, iac, etc and there is at most 0.3-ohm resistance, so that seems to rule out some sort of inline resistor/regulator between the relay and the sensors/ecu.

So it's looking more and more like a grounding issue, right?

---

So on to the voltage leak thing...

I get 0.791 volts from the load side of the relay to the chassis, so I'm getting some resistance and some voltage feedback on that circuit from somewhere else, but definitely not because of the relay.

With more testing, I've determined that the voltage leak is coming from something backfeeding through the ECU on the "third" connector bundle (fans, a/c, headlights). I'm too tired to mess with it any more tonight, but I suppose more testing will be in order to see if I can determine just where it might be coming from...

 

I figured out what was causing the voltage backfeeding through the ECU, so I thought that I'd update this for anyone who runs into a similar issue. The source of the ~0.75 volt backfeed was my OBD2 scanner. I have a ScanGauge2 that I leave permanently mounted in the car and I'm guessing that the circuit it uses to auto-detect when the car is running is the culprit for the odd low-level voltage readings that I found running through the system when the car is turned off.

Now I just need to see about the voltage drop on my main power circuit.

 

OK, so with some more poking and prodding, here's what I've discovered:

Voltage drop across the relay/circuit appears to be ~0.3v when the car is not running.
When the car is running and system voltage hovers around 14.2v, the circuit voltage hovers around 13.6v--so an average of 0.6v drop. This seems like a little bit more than I would expect, but perhaps this is relatively normal? Anyone care to weigh in on this?

The MS3 sees an additional ~0.6v drop, which might actually be correct (because I still need to doublecheck some ground connections) but if not correct then I will need to calibrate voltage settings on the MS3.

I think I've got this problem almost licked. Thanks Brain and Joe!


Relevant Test Data

Key On Engine On:
BAT 14.20 (Across positive to negative terminal)
ECU 13.60 (@ ECU harness terminal to chassis)
OBD2 13.30 (@ diagnostics port to chassis)
MS3 13.0 (*As reported by device, not measured with DMM)
SG2 14.4 (*As reported by device, not measured with DMM)

Key On Engine Off:
BAT 12.19 (Across positive to negative terminal)
ECU 12.04 (@ ECU harness terminal to chassis)
OBD2 11.92 (@ diagnostics port to chassis)
MS3 11.5 (*As reported by device, not measured with DMM)
SG2 12.1 (*As reported by device, not measured with DMM)

 

Update: Verified grounds on MS3 are good-to-go and was still getting a ~0.6v discrepancy so I made some minor adjustments to max ADC battery calibration and everything seems to correspond properly now. Thanks again for pointing me in the right direction, guys!