Car not starting randomly? READ THIS (30 Amp Starter Relay Mod)

[3p141592654]

A repost of something I said in another thread...

This circuit is so trivial everyone should be embarrassed if they cannot diagnose it.

Things to check in no particular order:
1) Starter (starter relay contacts wear out)
2) Ignition Switch (using voltage drop test under load)
3) Starter relay used for theft deterrent & clutch interlock functions (using voltage drop test under load)
4) Neutral Start Sw A/T only (using voltage drop test under load)

 

[JDMMA70]

I haven't posted in this thread in some time. Having my father help me out we found that the main wire going to the starter from the harness was sending intermittent voltage. Sometimes just by wiggling the wire the car would start. During my engine swap we replaced that wire about 5ft id guess up the harness with gxl high temp wire, a new connector and terminal on the starter end. Since then I haven't had a single issue with starting like I used too. *knock on wood*

 

[Dan_Gyoba]

@Pi: Of course we still have people insisting that they get better light output by using the headlight relay to trigger another relay for positive power to the headlights, and the headlight positive power circuit is as simple as it gets.

@JDMMA70: Wire broken internally? You shouldn't need to replace a length of it, you should identify the break point. It's likely to be either where the wire exits the harness, or right near the connector. At the connector is usually the highest stress point on a wire, and the most likely break point.

Perhaps if people are too intimidated to use Google or something, a description of a voltage drop test, what it does, and what it means are in order.

A voltage drop test is checking 2 points in the same circuit for a difference in overall voltage. A higher voltage potential across something like the ignition switch means that the switch itself is acting as a limiter. So if you check the ignition switch harness, and put the + lead of your voltmeter on terminal 4, and the - lead on terminal 1, you should read battery voltage with the switch in normal position. (Terminal 1 will ground through the starter relay.) Hit the starter, and it SHOULD drop to near zero, since the ignition switch should be providing a good current path between pins 4 and 1. It will probably have SOME value, but anything under about 0.5V should be acceptable, with about 0.1 being pretty close to ideal.

Next, do the same thing to the relay, across pins 2 and 4. In this case, the switch in normal position should read 0V or really close to it, it should not rise much when the starter is triggered.

The auto neutral start switch also needs to be checked under load, but the manual clutch switch is only used to interrupt triggering the relay, so as long as it gets contact, it should be fine.

Lastly, the starter solenoid itself can be checked by testing between the big battery cable and the starter solenoid contact.

 

[JDMMA70]

Yes there was a break halfway up the decided to replace whole section dunno how it happened

 

[Satan]

... Not arguing, just a counterpoint.

Replace entire span of ~20 year old wire, or find the break and hope it doesn't break again somewhere else, (or be corroded) elsewhere in the circuit. Sometimes the same time to troubleshoot is way-more, than the same time to replace and be sure. I think it's worth ~$1 a foot to replace ~5 feet of problematic 20+year-old-wire, instead.

Only example I could give is when a computer tech tells me it took him 5 hours to troubleshoot a PC and still did not have problem fixed, or root cause (for a $$$$ salesperson). I look at him like he's crazy... It's a standard PC image for the Sales team. Take 30 minutes to backup the data and re-image it. Problem solved (goal of "usability" is met, in much less time). Salesman making $$. I don't care what got corrupted or why. It is a non-issue w/a fresh install of the OS.

Someone else may choose to use Toyota wire, or whatever. Matters not... you just eliminate that entire span as the issue (especially when it's overdue). I appreciate those folks who "really" want to identify the issue, but it's just not that necessary. I just want the issue resolved and usability returned to normal.

 

[jdmfreak]

True but I for one don't like "shotgun fixing" or one by one "blindly" replacing parts that COULD be the issue... If that wire is the problem then yes replace the whole wire it's a better fix anyway not having another splice increasing resistance... Without proper diagnosis you could end up replacing all the wire, relay, ignition switch, etc when the problem was simply the starter...... or replacing all the wire, the ignition and the starter when it simply needed a relay...

In order to fix something correctly one must first understand the problem

 

[3p141592654]

I agree, RC-CA. Root cause then corrective action. For the computer example, if the problem was the hardware and not the software then you've solved nothing and still have the same problem.

 

[Dan_Gyoba]

Right. Because (for example) a bad RAM chip NEVER causes OS instability...

Unless damaged by something external, the age of wire doesn't matter. The insulation may eventually become brittle, getting cracks, which then start to corrode. Unless there's repeated flexing at a point, the wire doesn't generally break. That's why most wire breaks are right at the connectors, or at junction points. If the break in the wire were at the terminal 50 junction point, I'd probably put a new connector on it and call it a day. If it were back where the wire joined the harness, I would be more likely to splice in new wire back away from the strain point, and install some sort of strain relief at the weak point to prevent it from happening again.

For my '90, I expect that what I need to do is clean and put some fresh dielectric grease in the starter switch. A new ignition switch is only $35 from Rock Auto, so I might even just do that instead. I'm not going to do that until I diagnose the switch though.

 

[Satan]

Okay... I guess my point has been missed. :-/ I will have $30 into new wire, distribution block, and relay to starter. Maybe 60-90 minutes, including beer-breaks. My starter is "new-ish." Replaced about 10 months ago, when I thought it was the issue.

 

[Dan_Gyoba]

Not missed. If you're sure the problem is the wire. If, however, the root of the problem is the ignition switch, it's going to get worse, and you still haven't solved it, because you haven't fixed the actual problem. Maybe in your case, the wire is broken, and replacing it fixes the problem. Hooray. But it's not going to be the case for most people, as wire is a relatively low incidence failure part, and when it DOES fail, it's generally right at the connector, and is reasonably obvious. Relays are also relatively low incidence failure parts, but it does happen. Even if it is $30, I'd rather not spend it, or the time installing it if it doesn't address the cause of the failure, and I'm back diagnosing the problem when I'm late for work and it's dark out.

60-90 minutes for the mod as listed is probably reasonable. Possibly doable if you're replacing the wire from the solenoid contact to the wire trunk as well. The point though is that it's not a guaranteed fix, because there's still other things that it could be. Maybe it makes the symptoms go away for a while, and they come back when I go to start the car next Spring, because now the stock relay is really fried. Or the ignition switch isn't making any contact at all now, or a completely different piece of wire somewhere has finally completely let go. In that case, I'm out the $30 and the hour to an hour and a half (During which time I could have properly diagnosed the problem, and at least known what the correct solution is.)

Done right, there's no guess work. I don't have to wonder if the problem will come back. I don't have to remind myself that the TEWD is no longer correct on that page for my car. I know that I had a specific problem with the car, and that I fixed it.

 

[Satan]

I gotcha (I got it many pages ago, in this thread). All I am saying or have said, is that I plan to replace most of the wiring/circuit anyways... was just something I already planned to do. It may not solve the issue, but it will address 90% of what it may have been and I want that "comfort" level. I may knock it out today if the distribution block arrives and Just post a couple of pics.

I personally think it is a combination of things in the wiring, relay, and switch (combined weakness over the stretch of the circuit), which is why this has been a tricky issue for most... Ever take a 7M harness apart? That green stuff is corrosion. I used to have ghost issues all over my 87T, until I got a new harness. I have seen the same with many other MKIIIs. After doing several 7M/1JZ/2JZ engine harnesses and swaps, I do not have the same confidence as you in my 20+ year old wiring. Enough so, that I know that fresh wire will let me sleep better, even if I still have the issue at any of the remaining components and have to eventually fix/replace them too. If the problem reappears in a year, I will know it's not old wire/relay/starter, etc. But, I am prepared to deal with it again, if I have to.

Believe me. I have learned that shortcuts with the MKIII will bite U. This is not a shortcut, or workaround. Much like the wire/circuit upgrades needed for the high-output alternator or the fuel pump 12v mod, it's just cheap insurance that I feel is worthwhile, in my case.

 

[Poodles]

@Pi: Of course we still have people insisting that they get better light output by using the headlight relay to trigger another relay for positive power to the headlights, and the headlight positive power circuit is as simple as it gets.

Entirely different. OEM's tend to not use heavy gauge wire to power the headlights. Add a couple decades on to that and you have issues with lighting.

The starter is a giant relay as the main power to it is not the solenoid wire.

What I always found funny about these starter threads is how often people discounted it being the starter. These starters are used on a TON of cars (hell, they're one of the 2 main starter designs out there), and the good thing is, the main wear parts is absurd in it's ease of repair and replacement. They've also had issues over the years with the contacts wearing out too fast, that's why they use larger contacts these days (and yes, the parts retrofit).

Really, only in the supra world do people jump to ZOMG THE WIRING instead of the starter like the rest of the car world

/rant

 

[Dan_Gyoba]

@Poodles: My point there being that they address the positive power supply, but not the ground, which is way more wiring, longer wire runs, and complicated circuits. The power supply to the headlights is dirt simple. Main fusable link to the fuse box (If you have a problem with THAT wire, you've got bigger issues than headlights!) to the relay. Decent gauge wire to the fuses, then reasonable wire to the lights themselves. There's no wire that goes back further than the underhood fuse box. and the whole thing takes up a corner of the TEWD page on the lighting system. the ground circuits, on the other hand go into the cabin through the actual dimmer switch to the ground at the bottom of the junction box in the driver's side kickpanel. None of the wires going to the dimmer switch are what I'd call heavy gauge, either, yet the ground for both low and high beam go there. For what it's worth, it's also fairly trivial to just replace the wire from the headlight fuses to the lights themselves, too, but when I did a voltage drop test on my wiring in my '89, I discovered that there was no need. (What I needed to do was clean the dimmer switch, so that I got a decent ground for my high beams.)

That said, I agree that it appears to be an anomaly that the Supra community jumps to the conclusion that it must be the wiring.

 

[Poodles]

Eh, a relay harness bypasses the fun switched ground headlights Now if only I could get the high beam indicator to function reliably with such a system on different cars... (keep in mind the relay harness is usually for HID's in my case since the factory wiring isn't adequate).

 

[Satan]

I did temporarily rewire the circuit, awaiting the arrival of the distribution block (hopefully today). My starter has less than a couple hundred miles. I ran a wire from the Ignition switch (ST1) through the speedo cable hole, to a relay, to the starter. Grounded the relay to the body/fender, but may move to where the engine harness grounds at the intake lower runners. Basically totally bypassed/replaced the stock circuit.

Looks like only a few feet of wiring needed for the signal to the relay (I relocated to the engine bay). A couple of feet from the relay to the starter, another foot from relay to distribution block. Maybe another foot for ground. I am using a piece of the old signal wire from the harness for the ST1 signal to the relay, but am ordering/getting new wire to replace it (just did not have that gauge on hand). Starts up consistently, like it should. My aftermarket alarm seems to be fine with it also. Took about 30 min to move/reconnect/test. I'll probably spend another ~30 min tonight to finish installing the distribution block and make pretty.

 

[3p141592654]

I'm confused. I thought your root cause was a broken wire. Why the relay and such if the problem was just a broken wire?

 

[Satan]

My problem is at/before the stock relay. I did this mod (30-Amp-Starter-Relay-Mod) and it worked fine, but when I relocated near the ECU, it kinda stopped working after awhile. I suspect some of the circuit/wiring is bad (less than optimal performance) in that area. Shortened the run of the ST1 signal to the solenoid, moving the relay back near the starter. I simply replaced the circuit with newer, non-Toyota, wire/relay. The ignition switch is the only piece that's not been replaced.

 

[3p141592654]

It stopped working after a while because you never fixed the problem to begin with. Its hardly worth repeating, but if you troubleshoot it as described, you'll find the problem and you can put all your extra relays and wire away for good. Its a trivial circuit.

 

[Satan]

Trivial circuit with old HW. Have U been reading the thread? From the beginning? We all agree on the "book" answer (you just repeated it for the kajillionth time).


Please read post #191. I am not doing the mod, as described here. I have replaced most of the circuit and it is done (it is something I wanted to do ANYWAYS). About the same time it takes for the mod, to troubleshoot/repair, or to post obvious rhetoric.

The mod is not a complete fix for all, but may buy you some time. I bought time w/the mod, then eventually did it like I wanted.

Standard disclaimer from other's posts in this thread - PLEASE READ FROM PAGE 1.

 

[Dan_Gyoba]

The ignition switch is the only piece that's not been replaced.

And if the ignition switch is the root cause of the problem? Do you know that it isn't?