Relays clicking rapidly with key in RUN position; no start
 

My car was running just fine until last night when I went to start it and as soon as I put the key in "RUN" before I've even tried to turn the engine, I just hear a bunch of relays clicking in unison. If I try to start the car, it will crank but not fire (presumably because it's not getting fuel or spark).

Is this a relay failure mode, or should I look elsewhere? I am also running MegaSquirt PNP, so that's another variable in the equation. What I notice while the relays are clicking is that one light on the MegaSquirt (the stuff that lights up the logo) stays on and the other one flashes with the relays.

Also the radiator fan comes on and stays on while the relays are clicking. I suspect that is just a symptom of the MegaSquirt not being in a fully ready state because it's not getting the correct power to start up? (hence why the car doesn't start)

I did some searching, and found a few references to "clicking under the cowl" being the fuel pump relay, which sort of makes sense. The MegaSquirt normally tries the prime the fuel pump when you switch to RUN, and if that's not succeeding, maybe it keeps restarting or something?

do you have a boost solenoid?

Yes, that is one of the clicking sources. However, there are others. The boost control solenoid didn't used to click unless I press the throttle (0 duty cycle at 0 throttle and whatnot). Also if I disconnect the boost control solenoid, then I still hear clicking... just not from that solenoid.

edit: If I disconnect the boost control solenoid and the main radiator fan (to reduce noise) I hear clicking coming from the steering column/cowl area and the ignition coil area. I think the coil is actually sparking just with the key in RUN or something. If I pull one of the plug wires the sound changes a little and if i bring it close to a plug I can hear what sounds like sparking. I measured the resistance of the coil pairs at the high voltage towers like the Haynes manual says and got around 10k Ohms for both pairs which is within the spec. I'm now going to hunt under the dash for the fuel pump relay (?)

edit2: disconnecting the noisy relay under the dash leaves me with only clicking from the coil and the megasquirt's light flashing. Plugging megatune into the megasquirt gives me spurts of communication (gauges show data then go dead repeatedly in time with the flashing/clicking).

edit3: disconnecting everything that was making noise leaves me with the megasquirt alone in sort of a power cycle loop I guess. Replacing the megasquirt with the stock ECU (which I think works) and plugging the fuel pump relay back in gives me no clicking. When I put the megasquirt back in, it's still doing its powercycle loop, but now the fuel pump relay only clicks intermittently...

edit4: so I opened up the case and noticed that the LED on the MSPNP daughterboard stays on constantly. The 3 LEDs on the mainboard behave as follows:
D16: flashing steadily and quickly
D15: flashing intermittently; when it flashes i hear the relay under the dash click with it; also it only spurts data to megatune when D15 flashes
D14: flashing with D16
When I put the boot jumper on, only the Daughterboard LED stays on. The other 3 on the mainboard all stay off. With the boot jumper connected, I am unable to flash the firmware as it can't even connect. With the boot jumper disconnected, I see communication when D15 flashes.

Sounds like your firmware is corrupted. Try reloading the firmware and then reload the base map. Does that fix it?

The trouble is that so far I've been unable to flash the firmware since I can't connect (as stated in edit4 above). Should I be doing something besides setting the boot jumper?

edit: Oh the other interesting piece of information was that this happened once before a few weeks ago, but before i had a chance to try and diagnose it it just fixed itself. It was only like a span of 10 minutes between me first trying to start the car and failing to the car starting up.

Also what would cause my firmware to just suddenly corrupt with no action from me? I had corrupted firmware once before but it happened while I was actually trying to load a map. This time it just seems to have happened on its own.

What error message do you get when you try loading firmware?

Just that it can't connect on the COM port (I don't have the exact error, I will try again when I get home)...
Oh but now I can't remember if I closed Megatune when I tried to do it or not. I might have left it open and so it might have been trying to communicate? I'll try to do a firmware flash with the boot jumper again tonight.

I'm pretty sure I had the same problem. It all of a sudden happened on its own, flashing the led lights along with clicking relays, and communication with Megatune is on/off with the clicking/flashing. I could still reflash the firmware though during the spurts. I guess it was rapid enough to keep going and not time out. I sent it in to diyautotune, and they said there was some loose parts that may be grounding out, but unable to test it since their testbed is having issues... Perhaps check that or give yours a quick shake make sure nothing is rattling. I'm by no-mean any experienced with this unit but you could give that a try. Go over the circuits and make sure something didn't come out of its solder or something...Good luck.

Thanks aphongt, that makes sense. I did give it a quick (gentle) shake and didn't hear anything loose.

Okay, I tried to flash the firmware again tonight and here's what happened:
1. first attempt, with boot jumper installed, firmware flash successful, but at the end of the firmware flash you get instructions telling you to load a map immediately. Also it reported my battery voltage at 7v even though my battery measures 12v, and 11.7 w/ the key in RUN. So I fired up Megatune, but it gave me an error message while starting saying something like "-->Boot???" but I neglected to save it. It also said that table corruption would occur.
2. So I closed down Megatune and shut the Megasquirt off.
3. Removed boot jumper
4. Start up Megasquirt; now only one LED on the mainboard is flashing, I think D15.
5. Communication is still pulsing on and off in spurts.
6. I try to load the map now, but it says controller voltage too low for every page so I tell it to cancel
7. I swap out one of the Megatune gauges for battery voltage which looks to be oscillating from 0 to 11 with some values in between. However measured battery voltage is still 11.7ish
8. Power cycle the car and try to load the firmware again, (with boot jumper installed) but now I just get:
Now Megatune won't communicate with Megasquirt at all.
Subsequent attempts to load the firmware give the same error message.

Try reloading the firmware with the boot jumper in place and see if it makes the error message go away. Tell it you're upgrading from standard MS firmware.

Aphongt's real problem was that a previous owner of his MSPNP apparently tried to disassemble for some reason, got in over his head, and just tossed the parts back in the box without actually putting the pieces back together the way they should have been installed.

Sorry, looking at my post it wasn't clear... every attempt at loading the firmware WAS with the boot jumper installed.

You know, the intermittent nature of this problem (that it just started out of the blue and that it happened once before and cleared up without me doing anything, and now that it seems worse than before) makes me suspect that I have bad grounds in my car or something. I suppose I should test the pins of the ECU connector?

Another piece of information is that when I was tuning the MS regularly (soon after install), Megatune would sometimes say that the controller voltage wasn't high enough to load the map (usually when the car was hot). I would usually just be able to try a few times and eventually it would work. Is that at all pertinent?

Well, I managed to reload the firmware with the boot jumper installed and saying "Y" to upgrading from standard MS firmware. So now I'm back to this:
Megatune still gives me communication in spurts. Now it looks like one of the LEDs is only faintly on steadily (D16 i think), with D15 occasionally flashing. It says the controller voltage is too low to program the map into flash and if I say to try anyway, it fails. Does this sound like a flaky ground?

Check the main grounding straps, there's one going from the back of the head (exhaust side) to the firewall/shelf, and also check the main grounding cable back at the battery. A bad ground would make sense, if it's just sitting where its supposed to be bolted. That would give you 11.7 volts at the battery, but intermittent voltage at the MS, wouldn't be surprised if it averaged out 0 and 12v to find 7ish volts.

And that ground location is on my '93, they could've moved it. Just look for any flat braided steel straps. I think in the back it's a solid cable bolted to the chassis though. Think there's one on the PPF too. Then go to the little ground, as it's more than one system faulting.

I guess before you do all this check continuity between the MS harness and ground, cause it sounds like you have a voltmeter. Also check voltage at the harness.

I ran into this same problem when installing a WBO2 controller on a friends 240SX with MS in it. It was due to a rushed install. I hooked up the shared(between MS and the O2) switched power lead to the wrong spot and fed 12v to one of the inputs(maybe O2) of the MS :facepalm:. Im guessing it activated the flyback protection and made all the relays trigger on and off. I went over the wiring, spotted the problem and fixed it and the relay clicking went away. Hope this gives you some idea of the problem.

I would go over any wiring that you or someone else may have changed. Maybe a wire is frayed somewhere and grounding out. :2cents:

Curly, I've already gone through most of the grounds at the chassis side and I cleaned them up as best I could. The two black wires that go into the ECU (pin 1A and 1B?) seem to have pretty good continuity to ground.

I'm actually starting to suspect the main relay (fuel injection relay under the hood, right?)

Here's some new information:
With ECU unplugged and the key on, I measured the voltage on the white/red wire that goes into the ECU from the main relay (pin 2B?). This was pretty much reading 0v... though I expected 12v (this seems weird and I think I must've measured it wrong).
With the ECU plugged in and the key on, I measured this voltage again and found it floating between 4v and 8v.
The voltage from the blue/red wire at (i think) pin 2A is a steady 12v with the key off or ECU unplugged and around 11.5-11.7 with the key on and the ECU plugged in (this wire comes from the ROOM fuse). So this seems good.
Note that through all of this my measurements may have been sloppy so if things seem implausible, it's probably because I made a mistake.

Anyway, this leads me to the main relay. I can feel it click if I leave the key on and unplug/replug the main relay... so it seems to be semi-functional, but I'm guessing that it's got a high resistance at the contacts, causing a voltage drop when you actually try to draw current. This is inconsistent with my measurement of 0v with the ecu unplugged though.

Tomorrow I will try to do Joe's tests described here: https://www.miataturbo.net/forum/t4681/#post44737

I think my main relay is in fact done for.
I hooked up 12v straight from the battery to the signal pins of the relay, and while it does click, I get no continuity on the switch pins.

edit: Welp, I put a jumper in place of the main relay and was successfully able to load the firmware and a map. So that's good news. However, now my battery is so drained that it can't turn the starter... presumably because of all my diagnostics. I'm going to see if I can get a new main relay tomorrow and get my car jump started.

Awesome.. Hopefully you got it fixed. I had the exact same symptoms from beginning to end. I'm not sure if the loose parts caused my ignitor to blow but I ended up with having fuel without spark. The relays clicked non-stop, but as soon as I switched to my stock ecu it stopped, so I figured it was definately something is going wrong with my mspnp. I just wished my fix was something like a relay. Oh well, definately I learned alot from this thread.

holy driver, It seems like I'm having the same issue you were back in 2010. When you changed the Main Relay were you able to immediately connect to the MS to do a firmware upload, or were their other issue you had to fix as well?

Did anyone ever suggest that he needs to fix his flat battery?