1UZ Toyota Ignition Conversion

[VL400]

Right, the time has come to seriously get stuck into running a delco on my VL. The issue is the toyota V8 has twin dizzies and twin coils. Its not twin spark, only one plug per cylinder, but rather alternates between each coil. One coil feeds one 4cyl dizzy and the other coil the other 4cyl dizzy. Each dizzy then feeds two cyl on each bank. Its straight forward as it just fires alternate coils but nothing in delco land is direct compatible.

There are two cam sensors that give a pulse every engine revolution, although most people only use one sensor when converting to an aftermarket management system so it gives one pulse per engine cycle. The crank sensor gives off 12 pulses per engine rev. Both cam and crank sensors are reluctor syle.



So there are a few options....
1) Continue developing my ignition sequencer - i have had it running on the bench in rough form and in theory it should work how i want. Just comes down to time really to get it finished off. It takes the factory cam and crank reluctor signals and spits out a signal that keeps the delco happy. The signal that normally goes to the ignition module goes into my sequencer and gets sent to the correct coil via a standard bosch module. During cranking the sequencer handles spark just like a factory ignition module.
2) Convert to chev cap and rotor - it works, its been proven even under boost to be ok but does not look factory due to having to cut up the engine covers
3) Use Dicks magic black box (if he still makes them?) - It uses two holden V8 modules and sorts out the signals from the cam and crank sensors. Works well but costs a bomb by the time everything is purchased.
4) Northstar V8 DIS - again an expensive option and uses a crazy dual crank pickup which is hard to setup. But its a factory module and plugs right in to the delco.



Anyone else got any ideas for a simple cost effective solution?? I would really like to get rid of the dizzies and go to either a coil on plug or wasted spark system. I will be adding a turbo so need the spark power, but if there is a setup that means the dizzies stay and it is simple i will go down that path to begin with and just wait till the ignition system can no longer cut it

[delcowizzid]

why not make a single coil with a twin dizzy setup with the coil driving two leads one to each dizzy.but i spose it may rev to fast to get a decient amount of dwell to charge the coil.or 2 delco's one for each dizzy and for 4 cylinders each this has been investigated to run V12 jag before

[VL400]

A similar kinda thing was done by firing both coils at the same time. This meant that one rotor was pointing to a post in the dizzy and in the other dizzy it was was halfway between two posts. To stop the cross arcing, and to dispose of this extra spark, some ground posts were added between each post inside the dizzy. This partially worked but cross sparking was then happening from the rotor to the ground post when it was meant to be firing to a cylinder.

Two delcos would work, and would be cool as you have the ultimate limp home mode! But what a nightmare to tune!! That jag V12 conversion was impressive Some random useless trivia, toyota ran two independant ECUs on the V12 in the luxury century.

[delcowizzid]

yeah you would have to take into account in ecu age and variations in the driver condition etc and sensors would have to be smack on matching values then as long as they both had the same tune on each one you would be sweet but yeah dodgy way round it.but hey you could say to people twin throttles thats nothing ive got twin delcos

[antus]

this is a bit of a long shot.. but is there any 4 cylinder reluctor pickups you can put in the distributors? and if so, and if you wired them together, would that give a workable 8 pulses per cycle?

[delcowizzid]

yeah and link the signal to ecu together so it sees 8 pulses down the one wireas long as they were timed right but it would probably still fire both at once.you could do that with the 2 cam sensors if they are hall effect and trigger a VN V8 module if the 2 sensors were set evenly apart but i spose that give the same issue of crossfiring too .maybe just swap out one dizzy for a v8 version in its place instead.that would be by far the easiest.

[VL400]

The twin camira module thing was basically how it was done when the cross arcing was happening, even with the grounding rings. This was tried by Alan Gibbs from kalmaker. He now uses the black box from Dick on his 1UZ powered merc.

The chev V8 dizzy conversion is certainly the easiest in terms of keeping the delco happy and using factory ignition parts, but it means no engine covers which i would like to try and keep if possible.

I will get some pics up of what the toyota dizzies look like, they are mounted inside the end of the cam pullies.

[antus]

How far away from completion is your module? Ive been thinking further about this, and think that it might not be too hard to make up some kind of a module with a basic MCU and application. If you took the 1 pulse per engine rotation and fed it to an IRQ pin on an MCU with some kind of internal counter/timer hardware then you could save the time the last engine rotation took, and sync to the start of the engine cycle too. Then you could divide this duration by 8 and output pulses toggling between two output pins to drive two modules and the two distributors and solve the wasted spark problem. Sure, the timing would not be perfect, but I cant see that any given rotation could be that different in speed to the previous rotation, and if it was a concern, then you could also take the signal from the 12x sensor, time 3 pulses, then divide that duration by 2 so your timing is more correct - every 1st pulse would sync to a distributor and every other generated pulse would sync to the other with a smaller error margin. You'd still need the 1 pulse per rotation to sync up the engine position and thus which distributor to fire, though.

Or what about modifying both position sensors to send 4 pulses each, half out of sync?

[wake77]

I am using the northstar module on a commodore and have it running very nicely now. the cost wasn't too much i think i payed about $250.00 for the module.
the dificult bit is the crank trigger but i have the cad files of the trigger wheels. mounting the sensors exactly 27 degrees apart was the trickiest part for me.
I now have a new project and i am putting a 1uz into a hx 1 tonner, i have a trigger wheel already cut to mount on the back of the balancer and just have to make the sensor mounts for it. will be a little while untill i get it running stil getting the engine mounted in the car at the moment.

[VL400]

The module is at the PCB design stage, but have bench tested using a dev board and a basic circuit. It uses a dsPIC as the main processor mainly due its 4 one shot PWM channels. My sequencer syncs off the cam pulse and takes in the 12x pulse but outputs only the ones that are 60deg BTDC as the ref pulse to the ECM. Also outputs the cam pulse if the PCM will accept it. The ECM sends the dwell/spark pulse which gets sent to the correct ignition module to fire the correct cylinder- as determined by the sequencer which is managing the sync pulse and counts each cylinder.

The 4 one shot PWM channels are so i can extend the dwell fairly easily, that way the dwell can start before a previous spark event. I have a prediction algorithm that looks at engine RPM and its rate of change, so predicts where dwell needs to start if the same spark advance is commanded from the last event. It then sets up the one shots and adjusts them as required.