I have strange problem now.
Car randomly dies (under high load, when everything gets hot in cabin and engine bay) and then it does not want to restart.
Trying to start right away yields no results. I pulled the plug right after it happened and it was dry, nothing out of ordinary...

In the log, everything seems to be correct, all connections verified, relays cleaned up.
Only show of dying in log is sudden engine death followed by 100% lean AFR.

lean AFR and plug leads me to idea of overheating injector driver

I have 3 relays for 6 coils (each relay has its own fuse and does power coil and ignitor)
I have 1 relay for all 6 injectors.

My injectors are :
stock high impedance 550 cc primaries (14ohm)
INJ1
INJ2
INJ3

low impedance - Bosch 0280150846 that has a 4.7 ohm coil and white top.
INJ4
INJ5
INJ6

In ECU manager i can choose :
Setup > Main setup > Fuel : Injector Resistance

There are 3 options >
1-3 ohm
3-8 ohm
High
My current setting is 3-8 ohm (maybe it should be 1-3 ohm?)

I had one day car dying on me and other day it was running no problem at all. Today its dying again.
I suspect something (ECU, Ignitors, Coils overheating)
ECU does not show any sign of error, and keeps logging during event like that. So, maybe its just injector driver in ECU?

My first idea was too rich mixture (10,5 AFR) and flooding the spark plugs, but it wouldnt show lean AFR, plugs would not be dry and when trying to restart, it would atleast try to catch up....
When its running, it pulls strong, all values are ok etc..

Attaching map and log when the stall happened.....

When i let it cool down a bit, it does normally restart and runs like nothing has happened..by cooling down i mean 5 minutes or so, coolant temperature or intake air temp does not change significantly.

Looking at battery voltage, starting to rise with the moment when engine stalls - either ignition or injection stopped drawing power....

Any ideas ?

Thanks!

 

Does the software have different options for each bank of injectors? If not that is your problem.

You need to be able to set High-Z for the bank the primaries are on, and 3-8 ohms for your secondaries.

When you set 'high' you should get the pure pulse width out. When you set either of other 2 options you should get the pulse width further modulated by another frequency say 20KHz at whatever duty cycle Haltech deem necessary for your hardware after the initial 'turn on time-period'. Instead of the frequency method the Haltech might alternatively have a linear injector driver (wastes the excess heat across itself). Or an IC version of a switchmode similar to the ECU method of current limiting.

jbperf.com makes a driver board. You could set all the outputs to high then and only use the 3 destined to be secondary injectors through the driver board.

 

Sounds like you're on the right track. Get it to repeat the problem. When it does and it won't immediately restart, crank it and see if it's the fuel or the spark that quit. Should be easy enough since you say you have a 5 minute window.

FWIW, I just set a car up with a similar injector set. High impedance primarys and the Bosch 4.7 ohm 1600 secondarys. Haltech told me to use the 3-8 ohm setting. My specific concern was, with the peak and hold logic of the two low ohm settings, the driver would not allow enough current to keep the high impedance primary injectors open during the hold phase. We used the 3-8 setting and it worked fine. We have since switched to four high impedance injectors.

Not specifically sure what the hold amperage limit is for the two settings, but it might be that the high impedance injectors are hitting that limit?

 

I plan on wiring in diodes to both power and ground parts of primary injector wiring and watch them whan problem occurs.....if failure should be on the power delivery to injectors (relay/wirign), first diode will no longer stays on, if failure should be caused by injector drivers in ECU, second diode will stop blinking.....

 

Today i started measuring everything and i noticed on INJ1 wire (ECU unplugged) i have 4.6V
Little bit more research and i found source of it. Its my old Apexi AVC-R boost controller. It uses INJ wire for measuring duty cycle, but for some unknown reason it does put +4,6V power into injector ground wire (INJ1 in my case).
Could this be causing some overloading of injector drivers and stopping ECU from injecting?

Or is it normal (that boost controller sends power to the ECU ground wire?) and my problem is somewhere else?

 

The injector outputs will show positive voltage when open. Can't remember how much voltage off the top of my head. Might be 12V. They pull to ground when triggering the injector.

 

Please re-read my last post.

Injectors are normally powered by 12V (supplied by relay). When i pulled out this relay, there should be 0. But in my case, i had 4.6V there, which is sourced by Apexi AVC-R (connected on INJ1 signal wire). Unplugging AVC-R got me 0V as i expected.

Voltage then gets passed by connected injectors (i unplugged one secondary to put probes on, others injector plugs were connected to injectors)

By reading AVC-R manual, this seems to be normal, question is, does it interfere with Haltech ECU in any way?

I made test diodes now, will make some runs tomorrow.....and see.
One diode is constant 12V sourced directly at injector, should be always on with key on.
Other diode uses same 12 source, but uses INJ1 ground, should be blinking only when engine running.
If engine will stall again, it should be able to tell very fast if its injectors or ignition issue..


I will as well prepare stock FC resistors to plug in and will try high impedance setting in ECU Manager then......

Secondaries are 4,7ohm , mazda injector resistors are 6,5 ohm, should make around 11 ohm then and all 6 injectors will be in high impedance range.

 

Maybe something is getting lost in the translation or maybe you need to re-read what I wrote. I'm not talking about the battery voltage supply through the fuse and relay.

When you measure voltage on the injectors outputs (INJ 1-6 in your case), you will see 12V+. When the ECU triggers the injectors it pulls the output to ground. You're only seeing 4.x volts. That's not correct.

 

Ok, you are the one who is getting lost in translation.

When i measure voltage on injectors with everything plugged in, i see 12V+.
When i unplug INJ relay, i see 4,6V which comes from AVC-R connected on INJ1 signal wire.


quote from AVC-R manual :
http://www.alltrac.net/manuals/AVC.htm

Although this unit may be installed on virtually any turbo vehicle ranging from normal, modified engines, to turbo upgrades, the unit utilizes the injector signal for its self learning function which limits its usage to Electronically Controlled Fuel Injection System Vehicles Running On Gasoline. Please be aware that the Self- Learning mode cannot have an accurate reading if the injector pulse display does not work and if the installation vehicle has an electrical current controller For the injector signal. (For instance: PZ 31, R 31)


My wiring is correct, my readings are explainable.
Question is - does AVC-R by putting its 4,6V in affects Haltech?

 

Problem first occured on previous firmware, now i am on latest firmware/ECU manager and it was same case.....

Problem occured on higher boost turned on (ie, load on secondary injectors being bigger, with their longer injection times .....)

 

Your not losing coolant are you?

Was your engine recently rebuilt? I had a similar problem with my 20b late last year. Every time I shut the engine off after it reached operating temp, it wouldn't start. I had to let it sit for 5 mins or so so it would fire back up. This only happened after it reached operating temp. Cold starts it would always start easily. Since I break my engine down all the time for my experiments, I noticed several broken inner coolant seal valley tabs on my intermittent iron and one broken tab on the thick iron. All broken tabs were towards the center rotor. Because of all the broken tabs, my inner coolant seals had no support underneath and were dipping into the coolant passage. What was happening was that once my engine reached operating temp and the thermostat opened, the coolant system really begins to pressurize. When I shut the engine off hot, the pressure would cause coolant flooding of the center rotor. I've since fixed my problems and have had ZERO hot starting problems.


I know your problem is different, so I was just putting this out there as some 20b engines have had bad castings from the factory causing internal coolant leaks when you least expect it. You never know, you could very well have broken tabs also.

 

No, i am not loosing coolant and my engine starts everytime no problem, hot or cold.

Only when it shuts itself (due to overheated injector driver or something of that nature) and restarts as easy as any other time once the injector driver cools down - 2-3 minutes (while engine still maintains its temperature)

I have used the car on the racetrack (and abused it, drifting all day, hot day.....went thru 4 full fuel tanks) and water temps stays under 100C , same does the oil temp. Coolant level is still same .....

When car dies, it will do in the gear, under full load and it would not catch back up untill some electrical gremlin gets cooler.....

In your case of loosing coolant, you probably had problems to starts when center rotor was flooded with coolant, but other 2 rotors would probably atleast try to start (ie, you can hear its catching up, but not totally.....)

In my case, you can hear there is no attempt at all, just like when you turn the injection off in ECU manager.

 

The boost controller is not putting voltage "in". You should see 12V. How many times do I have to say that? Claudio told you the same thing on the Haltech forum.

Remove the boost controller and go for a drive. How hard is that? It's the obvious anomaly. Eliminate it and test.

If that is the problem, move the boost controller pickup to the tach output. It's just looking for an rpm reference anyway. The tach output will give the same type of signal; switched to ground with a 12V pull-up.

 

C. Ludwig

Apparently, AVC-R does that and by looking into its wiring diagram and reading installation manual, you would understand that. I am not blind, i can read multimeter.

AVC-R has one (1) blue wire to measure injector duty. That wire splices to INJ wire. That wire is normally open, when ECU sends signal to injector, it gets grounded inside ECU. To be able measure injector duty, AVC-R has to send current into that wire. That is my 4,6V.
When i unplug that blue wire from ECU harness, i get 0V with relay out and 12V+ with relay in. If its still unclear to you, i can create wiring diagram.

Up to this point, its all clear.

Question is, does Haltech have a problem with AVC-R putting this 4,6V there?

Already been quoted , but just reminder :
Please be aware that the Self- Learning mode cannot have an accurate reading if the injector pulse display does not work and if the installation vehicle has an electrical current controller For the injector signal. (For instance: PZ 31, R 31)

Is this the case? Does haltech have electrical current controller for injector signal?

I wired in diodes and prepared resistor pack and wiring, going out for drive now.

 

Update

Been testing it out, injectors still set at 3-8ohm in ECU Manager.
Diodes hooked up. AVCR still connected as before.
Few runs and it died again. Quick look at the diodes showed up power is still there, but other diode was not blinking.

Popped the hood, removed bypass and plugged in Mazda FC3S OEM resistor pack, changed settings in ECU Manager to Injectors HIGH and back to testing. Have not died ever since, and i was trying atleast for hour (1-2-3 gear pulls, turn around and again...)

I have drift event on saturday, that will be good test as well....

 

Another update

Problem still not solved. Symptoms still same :

ECU gets hot in race conditions (u cant keep hand on it)

Injectors in primary rail 14ohm (stock), secondaries 4,7ohm+6,5ohm resistor pack.
Injectors impedance settings in haltech - HIGH.

It was still shutting down, so we changed settings to 3-8ohm and it helped for hour or two, i was boosting up to 1bar, drifting in 4th gear on 225/50/17 tires (stock FC transmission, FD torsen, stock FC 4.1:1 diff ratio), car run great, made a power but due increased heat in cabin and engine bay it started shutting down again, always diode with INJ ground signal dies, same issue.

I managed to win event in my class in meantime while it was running.

Key on/off usually helped but to keep it running, but once loaded (secondaries kicking in with longer injection times) it died again. We are sure its injector driver getting too hot and self shutting.

Now, i can probably try to wire in another resistor pack (in series, which will get me to secondaries injectors resistance of 6,5+6,5+4,7 = 17,7ohm and switch back to HIGH settings.
This should discharge more heat into resistors, instead of ECU. Not sure if 17 ohm is not too much of impedance for injectors in general (from ECU point of view)

 

my E11/20B car was the same way, intermittent failure of the ecu, you're lucky cause the e11 software would get all scrambled....

but anyways, where is the ecu? if its under the carpet, like stock, this is bad, that location is basically between the exhaust and the carpet, the glove box is better, and actually letting the ecu just dangle from like the rearview mirror was even better.

the E11 would also be hot to the touch, i had the e11 doing the fan/tach/boost control/etc so i was actually using almost ALL of the inputs and outputs, and IMO this was too much for the ecu, as this problem got better with a 2 rotor, which was fuel and ignition only.

try putting the ecu someplace cooler

 

Mike, i had problems with E11V1 back in 2004, and it was shutting itself off, but this is different.

ECU runs, does not disconnect and log shows no errors, it even shows injection times, rpm's and all other outputs/inputs works. Its just the injection driver which gets hot and shuts off.
And its load related (load on that injection driver to be specific).....u could be driving high rpm, second after it shut itself and was restarted, and it will not die unless you start boosting up and putting more current in (thru secondary injectors).

 

so its the output side? that's almost worse, as who knows what fuel/ignition does before it shuts down...

either way, try mounting the ECU in a cooler spot, or switch back to a 2 rotor....

 

We have a theory about one of my injectors (primary, as they are old , not new) is partially stuck, or just not working perfectly smooth.
That might be causing excessive draw on it and that will lead to excessive heat. In combination with race conditions, ECU locations etc, it will all add up.
Friend of me with quite large diagnostic equipement will visit me and we will run tests on injectors to see if there is something wrong with them.
If i had spare set of them, i would simply just try to swap them out, but i dont

I really like the PS2000 and i dont think unit to be faulty. It just seems to be responding normally to some excessive conditions and i am trying to find source of the problem.
All steps we made so far, were just confirming our ideas. Its just very very hard to replicate the conditions, as it will never happen in street/road usage, only after very hard usage in race conditions (for hours).

 

Does that image mean anything? It explains why you're seeing less than 12V on the injection output you have your boost controller connected to. As I said before, the boost controller can't "put" voltage into a circuit. The boost controller is applying a load and drawing what should be a 12V source down to 4.x volts. That same load is going to be present while the engine is running, and possibly create your overload issue.

Have you tried removing the boost controller as I suggested?

 

agreed, if there is some "injector" problem, and there is something in the circuit, it would come out really quick!

 

Specially for C. Ludwig, i made this video to show him that AVC-R does provide 4.6V at blue wire, which is connected to INJ wire at ECU. And does that, whenever the wire is connected to anything or not.

I honestly believe You know the basics of the Haltech wiring and do understand this :

1) All injectors have common power supply wire (one relay does provide power to all 6 injectors)

2) if you hook up DMM (digital multimeter) at either side of injector plug (when connected to injector), and one of the 2 wires is being powered by 12V+ from injector relay, you will see 12V+ at both wires, measured against engine ground.

3) If you remove INJ relay (or INJ fuse), you will cut the power to the injectors and either of the wires will show 0V against the engine ground.

Now, this all should be no discussion about, as those are the facts everyone dealing with engine / ecu wiring could very easily measure and would remember like basics.

At this point (INJ fuse or relay removed, no battery power applied to injectors wiring), we will connect the the AVCR's blue wire to the INJ1 wire of haltech's wiring. It does not matter if ECU itself is plugged in or not, INJ wire on haltech side is not providing positive voltage, only ground and only when the injector should open.

AVCR does provide source of +4,6V at blue wire, as you can see in attached video i just made for you.
And this is something which you will then measure at the injectors itself, obviously, as INJ1 wire will act like power wire (+4,6V power) and thru original power wire will get this reading at all 6 injectors (injectors plugged in, fuse / relay removed). This is the fact and truth, no matter how hard is it accepting it....

And to answer your question if i tried to disconnect it - YES, i have completely disconnected it when problem occured and it did not stopped it from happening it again. It eventually might be interfering with Haltech and lead into faster overheating of injector drivers, but when it will overheat, instant removing of AVCR and then boosting (loading of the inj drivers) lead into another overheating and shutting down.....

Enjoy the video

http://www.youtube.com/watch?v=243LZ59u97g

In video there are 4 wires connected to AVC-R (12V+ constant, 12V+ IGN switched by key, batery minus (ground) and blue wire (+4,6V output which connects to INJ wire of ECU)

 

kind of missing response from C.Ludwig...

 

Let us know when you fix it...