ECU Code 12 and strange timing

[svpmx83]

Norbie had the same issue, and boost will not be an indicator of health as Limp home mode gives tha same result as Anti-Lag in rally cars, retarded timing and rich mixtures will spin a turbo wonderfully.
The sensors have to come from Sydney so I probably wont get them till later on in the week, I hope this fixes it.

 

[svpmx83]

Ok, the sensors arrived today, installed this evening with no change :shrug: I'm now completely confused. Ive disconnected each cam sensor in combintaion
both unplugged - wont start
just one unplugged either side - runs in limp mode and gives error 12
both plugged in - runs in limp mode and gives error 12
wiring tests to correct values all the way back to the ecu
no shorts - no open circuits - shielding is correctly grounded.
I'm starting to think that I have 2 ecu's with exactly the same fault.
I'll get someone else to recheck my wiring just in case I'm going blind but other than power I can't see anything in the body connection plugs that could even remotely cause limp mode. Bummer.

 

[Zuffen]

I think you will find it's wiring.

If the engine ran OK in the half cut and won't now it's either a wiring error or a sensor you've damaged during the install.

So far I haven't found an engine that ran well in the half cut have a sensor failure during the swap.

Go back over your wiring and recheck step by step. Don't just look at it check every wire has a good connection to its terminal and the terminals are all clean AND connected.

Have someone else assist as they may see something you don't.

How neat is the job? The neater it is the easier it is to troubleshoot.

 

[svpmx83]

Thanks Rod,
I have rewired it twice now and honestly have got to the point where the wiring is spot on or I'm making the same mistake by checking my own work.
I cannot say how well the cut ran before it coming to me, so cannot eliminate a pre existing fault. It will definitely need a fresh pair of eyes to verify my work though.

 

[MacroP]

Hi steve,
I was thinking about what could actually bring up the code 12. Other than faulty cam sensors and wiring, which you have changed and verified, to me it must be the cam's actual poistion, either one side is out or both. To me the ECU would use one cam for intial reference(and use the other for redundancy) then count the teeth on the crank(24) and expect to see a cam pulse at a certain period, then make an assumption of position, using time and rpm(maybe acceleration too), to fire the plugs with some sort of accuracy. The intial reference would be locked in(a timing table perhaps) and any errors from startup would be there to stay until shutdown. The ECU could compare the crank pulses count between each cam pulse. If it's one tooth out, and each cam tab is 180 degrees apart then there would be 24 crank pulses between each cam pulse. If it's one tooth out the the sequence would be 23,25,23,25 etc etc. Surely this would produce an error in the ECU. Sorry I'm thinking out loud.

A handy thing to do would be find a local electronics technician(TV repair - they're cheap) with a 3-4 channel Digital storage scope and get them to monitor all three sensors at once. At a constant idle you should be able to count the pulses and watch them relative to the others taking note of the crank pulses between each cam pulse to see if they are symmetrical. I assume they should be. The 4th channel could monitor the tacho output to get an idea of actual timing relative to the others.

Another thing, in regards to your meter measuring the frequency of the pulses. I'm thinking that a meter would have a hard time measuring the cam sensor output accurately as the duty cycle of the cam pulse is so small, ie one small tooth per entire rev. Tomorrow, I'll check my cam outputs with some different meters and compare your results. Unfortunately, I can't test my scope theory myself cause work won't allow me to loan their scopes for a few days - I've tried before.

Good luck. Hope you find the answer soon. If you need any electrical test measuremant done for comparison just ask. My ECU and wiring is still dangling out under the dash.

 

[svpmx83]

Marco, Think away!
I'm seriously going to owe you and andrew dinner next time I'm in Perth.
I was also thinking of comparing the signals on a scope, and as it happens my brother is an electronics engineer and has the equipment at his home, so a few beers later for him and I should be able to eliminate signal issues. Physical cam timing has been checked by me, by both comparing static timing marks and also by using the timing light on the cam pulleys while the engine is running. I also had toyota check the cam timing at the lobes while they were doing the valve clearances. Also if the timing was out the ecu usually gives code 13 not 12.
I've since discovered that my test equipment is picking up interference from the high voltage side of the ignition and therefore giving inaccurate readings, static tests with a screwdiver over the sensor will happily produce 2hz readings but my digital frequency meter is too sensitive and not just picking up the cam pulse. I'm pretty sure it is seeing small voltage fluctuation in between the main pulses and counting those as well. So CRO it must be.
Rod may still be right and I've goofed in the wiring somewhere or I have two munted ECU's, so I'm going to get my brother to recheck everything again.
It runs quite well considering there is no timing advance, so I'm going to start from scratch an rewire it again and then get it rechecked and the signals verified, but it's probably something as stupid as a bad earth.
I refuse to admit defeat!!!

 

[Peewee]

If you get REALLY pissed off I have a complete UZZ30 loom and ECU/ignitors etc here you can try out.

 

[Zuffen]

I don't know what engine you have. Is it a Crown (1 ECU + 1ECT) or Soarer/Celsior (combined ECU/ECT)?

If it's the Crown I've heard you can plug an ECT cable into the ECU by error (the plug is the same) and apparently it will run?

I don't know if that's correct but worth a look.

 

[svpmx83]

Its a single unit ecu like the soarer so no problems with wrong plugs unfortunately. And I'm really hoping I don't have to resort to a complete loom swap :bigeyes: but will put that down as Plan C.

It HAS to be something stupid

 

[Zuffen]

Steve,

My car ran like a dog.

Toyota couldn't pull a code to try and sort it out.

One night I ended up going to the car with a pair of wire cutters and walked away with the ECU. plugs and small loom I had made and started again.

Wired it up on the Kitchen table and made it as neat as possible so I could find any faults.

Plugged it all back in the car runs perfectly.

I still don't know what was wrong but I do know it WAS my fault as it was obviously a wiring error.

 

[svpmx83]

Thanks Gentlemen, I now have ignition advance.

I've damn well fixed it :veryhappy
The lexus sc400 wiring diagrams i have show STA being fed from the starter solenoid, and as the engine and ecu plugs are the same - Silly Me assumed that the engine harness would be the same - WRONG. The UZS141 crown takes STA from the positive side of the neutral start switch which feeds the crowns starter relay, and as the MX83 doesn't have a starter relay I fed this from ignition positive withe the original MX83 heavy start wire going through the relay contacts to the 1UZ starter feed in the body plug.
This caused the ecu to constantly think the starter was engaged.
So yes i'ts MY fault - thanks Rod for the suggestion to attack with wire cutters and start again. I went at it distrusting everything, and I traced EVERY wire with a multimeter before reconnecting.
Oh and I now have a small fuel starvation problem before the fuel pump ecu kicks in to full speed - but otherwise it goes very v e r y well.
Now back to the garage to finish the wiring so I can have my cruise control, sat nav and ABS back.

 

[Peewee]

You'll find the SC400 has the exact same wiring as you originally guessed it would.

If you have a look at this diagram
http://conceptual.net.au/~peewee/wiring/uzz31_into_mx73_wiring_1.gif
you'll see I have changed the diagram to suit what is correct.

When I first looked at the diagram before I had my loom, I could tell that who ever drew the diagram, drew it incorrectly, and so I changed it straight away.

 

[MacroP]

Awesome dude, so the ECU heavily retards on start hey and stays there. Also the ECU gets confused when it thinks the engine is cranking but is seeing a heap of revs bringing up a code 12?. I'll lock that one into the memory bank.

In my setup, I've actually put in a starter relay in the engine bay. Originally, I wasn't going to do this but I thought the immobiliser wouldn't handle the solenoid current so I put in a relay setup like the original Soarer loom. Very lucky I did this.

Don't worry about the ABS, sit down and crack a tinny I say.

 

[Zuffen]

SVP,

Good to see it sorted.

I hadn't realised you were using a miss-matched ECU and loom.

So far everyone who has had a wiring problem has gone back and found the error/slip, or whatever, and fixed it themselves.

Think how much the collective knowledge of EFI improves through these sorts of problems.

Now go out and enjoy it.

 

[svpmx83]

Thanks again Gents, I'm over the moon that it has finally been resolved.

The ECU and loom are both off the UZS141 crown and have the same ecu pinouts as the soarer/sc400 just the neutral switch wire colours are reversed. SO after fixing that I now also have the transmission modulating shift pressure correctly to the point that you can't really feel the gear changes.

Something you may want to try if you are running the auto box is grounding the NSW signal with a SPDT relay when the ect switch is in PWR mode as it pretty much eliminates the modulated clutch slip. My wiring stuff up had STA constantly seeing 12v and NSW always to ground. By fixing the STA signal only I corrected the engine control and still had firm shifts, but no cruise control as the cruise monitors NSW.

 

[sideshow]

i had same prob on a 1j

i have said people to check this before

i fitted a scanner to it and read the live data

cause it was recieving a start signal all the time due to a 1j behind a 6m auto gbox
it was gettin backfeed thru inhibitor so on scanner the timing stuck at 10 degrees

all the time due to the start signal being on like an ignition signal

alot of people stuff up the start inhibitor wiring due to they dont know how it goes


heres a quick run down


start signal (thick wire)from key goes to pin 30 and to black/blue(thin) wire on inhibitor and also to NSW on ecu

coming back from inhibitor is a black/white (thin) this then goes to sta on ecu and also to pin 85 on relay

86 on relay should goto earth

and 70 goes to starter solenoid

so basically nsw is start signal before inhibitor
and sta is start signal after inhibitor


so if gear stick is in 1 2 d or r and u crank it wont turn but u will get 12v to nsw only when cranking

when in park and neut u should get 12v to nsw and sta only when cranking

 

[hi-lex]

my prob was exactly the same ,only had 2.3v backfeed but thats enuf to trigger the input, cut the start input wire while motor running an instantly ran better, silly me thought I could get away without a start relay, yes I was wrong, damn I hate that, thanks for the tip sideshow, you allways got the answers

 

[sideshow]

the start signal can stuff alot of things up if not wired correctly

on cressidas when fitting 1j or 1uz u have to do some small mods to get the cruise control to work

cressidas have thick inhibitor wires so start signal goes thru it

but a 1j and a 1uz has thin wires so u need a relay but then u still get backfeed to the cruise

they are a big pain in the butt

but alll u need to know is nsw is start sig before inhibitor
and sta is start signal after inhibitor