Haltech + 7MGTE CPS issues

[paradox616]

Hi guys,

i've got a haltech e6x running on a 7mgte, its been to the tuners once, they managed to do a base map and made some good power with a miss, (284rwhp @ 8psi)

turns out the miss was the Gapping on the CPS it was 0.4mm or 16thou
i regapped it to 0.2mm or 8thou (lower recommended in the TSRM still within spec)

the car ran awesome! was a little hard to start and didnt want to idle until warm... but did not miss a beat whilst driving!
as were moving into winter in Australia its getting colder (round 10deg C once the sun goes down)

it wont actually start when cold, i re-gapped the CPS to 0.3mm and it starts but has the same miss issue. i've tried a few different cp's with no luck....

I'm not the biggest fan of electromagnetic reluctors... its analogue after all


I've done a search on this and found not much
has anyone had luck converting it to hall effect and how did they do so? i think Duane did it?

I'd love to know how/see pictures. are there any other cam sensors that may bolt in? if needed ill fabricate a slotted wheel but id like to know if there is a better way.

 

[gennro]

Well I would suggest getting rid of the cps all together. You can use this http://www.diyautotune.com/catalog/trigger-wheels-c-48.html and mount it to your crank pulley and use your choice of VR or Hall sensors with a homemade bracket.

Or you can try this http://www.supramania.com/forums/showthread.php?121812-CAS-mod-for-MS-II-(Complete-Writeup).

 

[paradox616]

Thanks! the modified stock CPS looks alright!

i've got a spare CPS so i might give it a shot, if that dosnt give me the desired result ill look at something a bit more hardcore such as Hall effect.

 

[whitemike]

I would LOVE for someone to do a write up on a mod like this. Ditching the CPS. Part #'s and prices and stuff.. I would be so down.

-Mike

 

[gennro]

I would LOVE for someone to do a write up on a mod like this. Ditching the CPS. Part #'s and prices and stuff.. I would be so down.

-Mike

There is one already its called a 2JZ =)

You need to be running a standalone to get rid of the CPS.

 

[ml43]

anyway you could do screen shots of your trigger setup page on the haltech, also pictures of how you have it wired(if possible)

 

[paradox616]

Sure, Here is the trigger setup, when warm. timing reads 10deg BTDC or over 1000rpm when cold.

 

[ml43]

have you tried messing with the gain and filter to get a more stable signal?

also, what type of ignition are you running? MSD? HKS?

 

[paradox616]

yeah i have, gain makes things worse

my ignition system is as follows

Haltech 3 channel ignitor into stock coil packs into 0.7mm gapped plugs NGK

 

[ml43]

k, pretty sure the trigger pickup is supposed to be set to falling edge, rising is normally only used for aftermarket systems such as MSD or HKS

sounds like you don't dont have a good shield, or have a bad connection somewhere on the signal wires, or it could just be a bad CPS, but if you have tried multiple ones with the same result, then I highly suggest double checking the wiring and shielding, and possibly redoing it if possible

 

[paradox616]

k, pretty sure the trigger pickup is supposed to be set to falling edge, rising is normally only used for aftermarket systems such as MSD or HKS

sounds like you don't dont have a good shield, or have a bad connection somewhere on the signal wires, or it could just be a bad CPS, but if you have tried multiple ones with the same result, then I highly suggest double checking the wiring and shielding, and possibly redoing it if possible

I've checked the shielding, shielding grounding which is at ECU, its the factory haltech loom, there are no issues with it, no resistance on the wires and they don't pass HT leads.

Any idea about the trigger offset?

 

[ml43]

where and how do you have the ecu ground(ed)?

tooth offset and trigger angle are just ways of adjust the sensor to the motor and getting the timing spot on

you use tooth offset to make large adjustments, and the angle to get it perfect

basically, if you have 24 teeth on the trigger, the offset can be anywhere between 1-24, 12 being one full revolution, 24 being a full engine cycle

also, your timing should be set to zero, so when you/your tuner tunes, you/he is seeing actual values rather than having to calculate values

 

[paradox616]

where and how do you have the ecu ground(ed)?
tooth offset and trigger angle are just ways of adjust the sensor to the motor and getting the timing spot on
you use tooth offset to make large adjustments, and the angle to get it perfect
basically, if you have 24 teeth on the trigger, the offset can be anywhere between 1-24, 12 being one full revolution, 24 being a full engine cycle
also, your timing should be set to zero, so when you/your tuner tunes, you/he is seeing actual values rather than having to calculate values

1) there is no soldering on the ground, it goes STRAIGHT to the ECU ground which is bolted to the chassis in the engine bay, once again tested resistance from the sensor end to the ecu ground and found no issues and continuity
2) thanks for the offset info, that makes sense to me.
3) @ lock timing 10deg it reads 10deg... at 0 actual timing it will read 0 or 20 will read 20 etc

This issues really just got me stumped, im going to try reducing the gain on the trigger to 0 and keeping the gain on my home signal as 2. i get consistant rpm of 125 on crank and trigger signals too, injection pulses as well, all sensors are reading normally. as i said earlier its like its missing home signals


As per a earlier question about the "motronic filter" it only applies when there are missing teeth, so i haven't fiddled with that.

 

[ml43]

mmm, forgot to ask, when gaping the CPS did you use a brass feeler gauge? also did you check multiple teeth for variation

also make sure you have the polarity correct going to the sensors

 

[paradox616]

mmm, forgot to ask, when gaping the CPS did you use a brass feeler gauge? also did you check multiple teeth for variation

also make sure you have the polarity correct going to the sensors

1) no i did not use brass feeler gauges, i believe they're stainless steel, is this an issue?
2) i did check for variations with different teeth, all were the same
3) did i check the polarity? i wired it up according to how others had running the same Ecu, ill have to check how its wired up and get back to you on that.

 

[ml43]

yeah, you can damage the sensor if you use a ferrous feeler gauge, its not a definite thing, but it does happen

exactly how do you have yours wired?

 

[paradox616]

ok just checked,

NE = Yellow (trigger)
G1 = Green (home)
G2 = not wired to anything
G- = Ground (red+blue wires on haltech)

swapped the trigger edge and i'm getting erratic signals on crank, it spikes up to 600rpm at even intervals, ie a home event. at this point timing jumps about -10deg...


found this in the haltech manual


"reluctor triggers produce a different signal
trace and the ECU uses signal conditioning to produce the square wave signal required by the
processor. When a reluctor trigger is being used, the Trigger and Home edge should be set to
Rising edge triggered, unless trigger edge problems occur."

they have an example of a toyota 24+1 trigger and it looks like the only point the signals match up is on a rising edge.

 

[paradox616]

Ok i think my home edge was wrong, set trigger to rising and home to falling edge and it starts up easily with minimal gain. the notch on the crank does not hop now its VERY steady. car drives good aswell


found this also in the haltech manual

"The common symptoms for trigger edge problems are a sudden jump in ignition timing angle,
of usually one reluctor tooth value.
Ie, with a 24 tooth reluctor cam sensor, one tooth value = (720 deg / 24 teeth) = 30 crankshaft
deg. If this problem occurs, try changing the edges of the Home signal. Please make sure you
check the base timing of the engine after this change is made."

 

[paradox616]

went for a test drive and its fixed, no misses car starts first go every time rpm's don't jump and makes good power on boost.

Thanks for the help guys it sent me looking in the right direction!

From what i can tell the issue was a combination of CPS gap and home edge incorrectly set.