One for car electronic types.

Jun 28, 2021 Last reply: 5 years ago 7 Replies

Social meja pal living in NI has a Range Rover. Circa 1990. Injection is Lucas 14 CUX. Ignition self contained with distributor. The only connection between the ignition and injection is a tach (RPM) signal obtained from the coil negative, and fed to the Lucas ECU via an inline



6K8 resistor.

He decided to get rid of the distributor and go wasted spark. There are several aftermarket systems to do this and he chose Nodiz. Using the wasted spark coils from the later Thor version of the Rover V8. And a 36-1 trigger wheel (fitted to the crank pulley) and VR sensor. All pretty standard aftermarket stuff.



He had problems with misfiring. After exhausting all the possibilities, decided to ditch the Nodiz. (It's not a common make and found support difficult)



Some time back he'd tried MegaSquirt to replace both ignition and injection. Had problems with that too - so stuck it on a shelf. I suggested he try it for the ignition only. It should be equally happy doing fuel only, sparks only, or both.



First results were encouraging. It ran well - misfire gone. Then decided to rationalise the wiring.



Nodiz said to derive the tach signal needed for the Lucas injection by making up a diode network taking a pulse from each wasted spark coil. (these fly back pulses are about 50v peak)



MS has a number of spare outputs you can configure in software to what you want. So easy to set one to the correct number of pulses for an 8 cylinder. Add a driver switching NPN transistor, and you have a 12v pulse (rather than 50V) Remove the 6K8 series resistor feeding the Lucas unit and you have a nice neat way of driving it. This is how I've seem it done on others. Leave the resistor in circuit and it drops the pulse level too much for the ECU to read.



When attempting a start like this, the switching transistor literally exploded. And the track on the PCB melted.



Only way I can explain this is 12v on the Lucas input - and at high current. Despite lots of Googling, I can't find a schematic for that Lucas ECU, although can for the earlier 4CU.



Any better guesses?


Why not reinstate the original ignition system to provide the pulses and the MegaSquirt to provide the spark?

Is this the switching transistor for the Lucas injection, or the Ignition drive from the MegaSquirt?

Power input? If it's driving injector(s) then I guess it will be high current.

More info needed.

Sounds to me like some current is going via a link it should not be to cause that sort fo damage, miswired somewhere. Brian

Crikey. Distributor ignition was about the least reliable part of any car. Hence no modern one using a distributor.

As I said, you can get the required pulses (known as a tach signal) from all four coils using a diode from each negative and paralleling the outputs. But it's not the neatest way of doing it, since MS can provide the same pulse frequency (but not voltage) easily. The diodes method is most often used to drive a dash revcounter - where the original was connected to the coil negative and got that high voltage flyback pulse.

You'd not normally make a direct connection to a processor port on MS. You add a buffer (if there's not one already there) And that can give you a

12v pulse, rather than the 5v one from the processor. All you need is a single transistor and a couple of resistors. MS does just this to drive LEDs, fuel pump relays, etc. All the output ports on the processor being the same - but configured in software as to what function they have.

The only connection between the two is the tach input to the Lucas ECU. And on earlier Lucas injection, a logic level pulse drives it happily, with the external 6K8 resistor bypassed.

It's actually all there, but needs careful reading.

Agreed. Many years ago (decades even) I added a home built CDI system to my car. It utilised the old contacts which were of a self-cleaning sliding variety which moved as per vacuum assist servo demanded. To make the rev counter work I had to drive the contacts through a resistor to

+24V. A supply that was convenient to make in the CDI electronics.

For the past 20 years I have only driven diesels. I get the idea of combining the pulses. It seems a messy way.

Ok, so essentially a bare Motorola MPU.

Can't the MegaSquirt produce the pulses required 2 x crank speed?

Its a shame the MegaSquirt doesn't use a more popular, faster MPU. I see the MegaSquirt supplies the source code. Are you able to compile, link and load this onto the MPU?

You can program it for any possible choice.

Which MS? There are at least three processors in use. Not much point in paying for the latest if you don't need the extra facilities or speed. They are generally used to replace early injection systems, so even the oldest and slowest likely superior.

I understand the first MS was a Motorola MC68HC variety, an 8 bit device, the rest are the Motorola 16bit MC9S processors of 24 and 100MHz.

I have found software and debugging facilities for Motorola devices to be expensive. Hence I've moved to the ARM processors where the whole development IDE, for say ST, is free and very well supported, which works with an inexpensive JTAG interface.

Hence the question if you were able to compile, load and even debug the software.

There are loads and loads of aftermarket ECUs out there. Most of which keep their secrets. And usually at a higher cost than MS which is aimed at the DIY market, with pretty decent support for that. And as with all computers, the software is often more important than the hardware.

The software author for MS is from the UK. And helps out with any problems, and listens to sugestions for improvements. Nothing to stop you modifying the software - but it's not open source.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required