Hey y'all,

I'm stuck in a rut here, and I'd like to open a thread for this issue, as it's pretty obscure, and I can't seem to find any other cases exactly like mine in the forums.

Some context: I'm building a 1999 Miata with an MS3PNP, Kraken turbo kit, Cobalt coolant reroute, GM IAT sensor, and AMPEFI EBC. I got the car running fine on the MS3 naturally aspirated first. Then I spent a loooong time installing the turbo kit. Most of the issues I had were intercooler- and charge piping- fitment related, but those are most likely irrelevant to this problem.
During the turbo kit install, I installed a new starter. I suspected my old one was the original unit, and it seemed a little tired during cranking, so I wanted to update it while I was in there, as it's a real PITA to access the starter once the intake manifold is on. (I took the intake manifold off completely to install my injectors. Did the starter install while the IM was off.)
I buttoned everything up, and went to do a first-start. Now, I have a no-crank. I don't even hear the starter solenoid clicking. The Main relay DOES click when the key is turned to "Start".

I suspected the starter immediately, as it's a unit from Amazon (d'oh!). The starter solenoid simply will not click.
- I tried bridging the 12V and GND studs on the starter to see if that would spin it -- no dice, besides a lovely spark show.
- I even tried injecting 12VDC to the starter solenoid using a DC power supply, grounded to the oil pan -- no dice again. I should note that I didn't have anyone turning the key to "Start" during this process -- I don't think that would defeat the purpose of this test, as, according to the wiring diagram, the only purpose of
the "Start" position in the ignition switch is to connect the WH/BK wire at the ignition switch to 12V WH/RD, which goes directly through the clutch interlock to the starter solenoid BK/RD wire.
- I replaced my battery with a new O'Reilly AGM unit. Did not help, though I was probably gonna need to do this soon anyway, as my current battery was MFG in 2021.

Now, the real intriguing and fun bit, and why I wanted to post here: I measured the voltage at the BK/RD starter solenoid wire (disconnected from starter solenoid), and while the key is turned to "START", this wire only sees ~3.6VDC. Not 12VDC. Uh-oh.
In addition, the resistance from this BK/RD wire to ground (when not connected to starter solenoid, and when key is in "Off" position) is only ~76 ohms. BIG uh-oh. I don't think it should be nearly this low, as I can't find anything in the wiring diagrams suggesting the BK/RD wire is in any way connected to ground. I would think the resistance should read as an open circuit from this wire to GND. Please correct me if I'm wrong in that assumption.

Things I've done so far, to no avail:
- Verified that I get battery voltage at the positive terminal of the starter, relative to ground at the oil pan.
- Cleaned both grounds at the front of the engine bay, beneath the headlights.
- Verified ground at back of PPF (visual inspection only)
- Verified ground in trunk next to battery (visual inspection only)
- Checked ST.SIG fuse for continuity, and it's good (though, according to the diagram below, this fuse is not even necessary to deliver 12VDC to the starter solenoid)
- Verified operation of my clutch interlock switch using a continuity test. It appears to work. I've been running the car with a bypass clip for a long time and never had any issues starting. It seems very unlikely that there would be a serious voltage drop within this switch, according to forum posts.
- Tried removing ignition switch from key barrel and actuating it with a screwdriver -- still nothing.
- Removed and cleaned ignition switch (there was a bit of corrosion in there, but really not that much.

• One note here: Before and after cleaning, it appears that when the ignition switch is turned to "Start", the WH/BK lead only connects to one of the two WH/RD leads (did a continuity test with a multimeter). There are TWO constant +12VDC WH/RD wires at the ignition switch connector. The connection between WH/BK and the second WH/RD is extremely spotty, barely there. I'm not sure if this WH/BK lead needs to connect to BOTH WH/RD wires at once in order to facilitate starting, and I can't find any information online to answer this question. All of my sensors and gauges function correctly, so I don't think I'm missing any power here -- and I would think that if only one of the two 12VDC wires was connected to the starter solenoid wire, that I'd still see 12VDC at the starter solenoid wire regardless, albeit with half the current.

Things I have NOT done yet:
- Re-clean the GND next to the dipstick (I already cleaned and installed it during the turbo kit installation). The BK/RD wire at the starter measures 76 ohms to ground with respect to the oil pan AND with respect to this ground, so I don't think this ground is an issue.
- Remove the clutch interlock switch completely and bypass it at the connector using some spade clips. I will try this tonight when I get home.
- Remove the Amazon starter and bench test it. Again, I will be doing this tonight when I get home if the above solution fails.
- Measure the voltage at the WH/BK wire coming out of the ignition switch when it is in the "Start" position. Kind of dumb that I've overlooked that so far, but I will be doing that tonight.
- Shorting one of the WH/RD leads to the WH/BK lead at the ignition switch connector. This just seems like a bad test, as none of the other systems in the car would be powered.


For reference, here is the 1999 Starter circuit, taken from the Mellens wiring document.



My initial prognosis--assuming the Amazon starter bench tests correctly -- is that I have a wiring issue somewhere between the WH/BK wire coming out of the ignition switch, and the RD/BK wire in the engine bay.
The lack of voltage, and the low resistance to ground at the BK/RD wire, makes me consider the possibility that a wire got pinched in the harness somewhere, and shorted to ground. I would think, though, that if this were the case, I'd get a near-zero resistance to ground at the BK/RD wire, and basically no voltage, as it will all be shorted to ground. Countering this idea is the possibility that a pinched wire could cause a voltage divider circuit, where only MOST of the voltage gets shorted to ground at the pinch, and the rest could get through. I see this as unlikely, but wiring gremlins are mysterious beasts.
I checked the loom sitting above the starter (where the BK/RD wire is), and it moves freely. It does not feel like it's pinched.

I'll now pull a Columbo "just one more thing" on you and mention that after my turbo install, I get no voltage at my A/C condenser fan connector. I believe that passes through the same section of loom as the starter solenoid wire, but I haven't done a deep dive into this problem yet. I just want to get the damn car cranking first. I also can't really see any relation between this connector and the starter solenoid wire, but please correct me if I'm wrong


Sorry for the humongous wall of text. I wanted to present as much evidence as possible, as I really have fought this problem hard. I greatly appreciate your time and any suggestions you may have.
If all else fails, and nothing will work, I'm going to do what I saw in another thread, and simply splice a new solenoid wire at the ignition switch WH/BK connector. I hope I don't have to do that, as wire splices are ugly and running the wire sounds like another PITA.

 

Following up after some research over lunch:

I mentioned that the WH/BK wire at the ignition switch only connects to one of the WH/RD wires in the "Start" position.
Per this diagram, that is completely normal. I don't really expect my ignition switch to be the issue.

 

Long story short, I made a new wire to deliver voltage to my starter solenoid when the ignition is turned to "Start". Spoiler alert: it worked. Here is my process:


So the clutch interlock switch has two BK/GN wires that go to a connector on the left side of the driver-side footwell. This connector can be seen here (you can just barely make out the BK/GN wire coming out of it:




This connector mates up with another connector in the left side of the driver-side footwell which has a WH/BK wire and a BK/RD wire. THIS connector is plugged into the chassis with a plastic retainer, and I kinda had to break it to get easy access to the connector. This connector can be seen here:





I should note that my car does NOT have cruise control. Mellens diagrams mention another brown wire involved with the clutch interlock switch, which is not present on my vehicle.

The clutch interlock switch only does one thing on my car. When the switch is pressed down, the two BK/GN wires are connected to each other. This then connects the WH/BK wire to the BK/RD wire at the mating connector. As I stated before, the WH/BK wire connects to 12V at the ignition switch when the key is turned to "Start".

Thus, the order of operation is: Clutch interlock switch pressed in, key turned to "Start", then 12VDC goes through the WH/BK wire to the BK/RD wire.

This BK/RD wire in the footwell connector SHOULD go directly to the BK/RD wire at the starter in the passenger side of the engine bay. This BK/RD wire has a connector on the end of it which locks onto the spade terminal of the starter -- this is the starter solenoid.

In reality, I did a continuity test between the BK/RD at the footwell connector and the BK/RD connector at the starter, and I got no continuity. My wire is clearly broken at some point along its run from the footwell to the starter. At this point, I committed to making a new wire, because I wanted to get this car started, and I did NOT have any desire to remove my dash or the engine to start picking apart the wiring loom.

This picture shows the BK/RD wire, its connector, and the starter spade terminal in the background (I cut the connector off the BK/RD wire to make my own.)




So, I made a longass single wire with the starter solenoid connector soldered onto one end, and a male spade terminal crimped onto the other. Plugged the male spade into the WH/BK at the footwell connector, and the starter solenoid connector...... connected to the starter solenoid.

Turned the key to "Start" and HEY PRESTO, WE HAVE CRANK!


In case you run into the same situation as me in the future, and you've proven that you have 12V at the ignition switch and footwell connector when the ignition is turned to "Start", just know that it will be much easier to run a new wire than to start taking apart the car to get at the loom to try and find where the break is.

I'll admit at this point that my trials and tribulations may not be over -- It's entirely possible that what happened to my BK/RD wire could have also happened to another wire in the loom. But we'll burn that bridge when we get to it, and I'm sure I'll write another wall of text if/when that happens.

Keep on crankin'.

 

Didn't read 100% of that, but I'm not sure you've got the correct grasp on how the solenoid works. It's not a ground and positive terminal, one is constant 12v, the solenoid connects that 12v battery source to the other, pushing out the starter gear engaging the motor. The starter gets it's ground from the body of the starter to trans bellhousing.

You can easily check a starter's function two ways. One is a little safer than the other. You can short the constant positive terminal to the solenoid terminal, which is a ~5amp load. This tests both the solenoid and starter motor. The other method is bridging the two large posts, bypassing the solenoid, and is engaging the ~40-60amp starter motor with a screw driver, hence the large amount of sparks you experienced. You can do either of these test with the key removed, ECU removed, ST fuse removed, even ignition fuse removed. As you can see from the diagram you posted above, the black positive wire goes directly from the battery to starter, with no fuse or relay interrupting it.

Next step is diagnoising the lack of 12v when the key is turned to START. Unplug the solenoid. Check the white/red at the ignition switch for 12v. Then hold it in the start position and check for 12v at the white/black wire. Check again at the clutch switch white/red. Press the clutch or install your bypass bracket and check for 12v on the black/red wire at the clutch switch, and again at the starter motor. Depending on where you loose 12v is where to start looking. IIRC, there's a connection at the engine bay fuse box that connects the chassis harness to charge/batt harness. I realize this is more or less what you did, but thought I'd summarize.

I've rarely seen breaks in the chassis harness, all wires are RTV'd into the firewall grommet, they're quite robust. I'd be looking for a shitty aftermarket alarm or remote start install, or oil leaks causing damage by the starter/charge harness. There's not a lot of plastic split loom running from starter to fuse box, it's not too difficult to remove it all and inspect that wire to the fuse box.

 

Hey man, thank you for your response and summary.
I now understand what you mean about the lower starter stud not being a direct ground. I misunderstood my starter last time I looked at it, and I initially thought that the wire on the lower stud was soldered to the body of the starter. I now see that the wire on that lower stud goes into the solenoid, so I appreciate the clarification.

Your comment about a shitty aftermarket alarm is foreboding, and unfortunately, possibly on the money. My car DOES have an aftermarket alarm/lock/unlock unit in it that was installed a long time before I bought the car, and the key fob for this system stopped working last year. There were no starting issues to speak of between that day and now. Possibly, it was only the fob which failed, and the box in my footwell might still be fine.

How big of a world of hurt do you think I'm in if I try to uninstall it? It seems like these systems tap into a good bit of the OEM wiring (I know my box has a 8 wires coming out of its main connector), and if I wanted to remove the system, I might have to remove the dash to get enough access to return everything to factory wiring. I'd like to not do that during the racing season if I can help it. I'm a very slow worker, and I can see that project putting me out of commission for a couple weeks easily if I make any mistakes.

I know every system is different, and you couldn't possibly know exactly what my unit taps into, but in your experience, when these aftermarket alarm systems go bad, what else in the car might they take offline?
I'll try and monitor the car for any other electrical gremlins in the future. For now, I have my starter functioning with the new wire. It's 18AWG, which should be able to handle that 5A load you mentioned, but if you think it's too wimpy of a wire, it'll only be an hour's worth of work for me to upgrade.

Again, thank you for your response. I appreciate the support!

(Edited: I took another look at the remote alarm box, and its connector doesn't have nearly as many wires as I remembered, it has 8 wires. Seems like a doable task to remove it)

 

The stereo shop that installed it in probably less than an hour or two definitely didn't remove the dash, so I would assume you could remove it likewise, it's just cramped and not very fun, obviously. Then again, pulling a dash takes me about ~45 minutes, so your idea of "easy" is probably different than mine Taking your time in your own garage is probably better for your mental health though!

Alarms usually tap into the starter, constant batt, lights, and switched batt. Some are spliced in, some are cut and spliced into, if that makes sense. Then they'll have outputs to things like after market door lock solenoids or horns, and inputs like antennas. Best is to find a part number of model number on your box and google it, I usually find instructions that way. Or just dive in and remove one wire at a time. Wouldn't be surprised if it interrupted the start signal somewhere.

 

Thanks for the input, Curly! I will probably try to take it out after I get this thing tuned. I only hope that whatever splices they made will be easily reversible. I know the fused power to this unit from the fusebox had wires twisted together and soldered instead of lap soldered, I hope that they didn't use that technique on the OEM wiring harness...