The V8 Owners Forum -

Hi all.

Just looking for some advice on a, hopefully small, issue I'm having with a project to change to a distributorless ignition system on my RV8.
I had a couple of posts on this subject late last year and earlier this year while I was implementing the sytem on the car, now the weather has improved I've put the car on the road and I'm getting some issues. I still have the distributor, coil and amplifier mounted on the car (with separate set of plug leads) so it's possible to switch between the systems and quite quick to do.

When I run with the distributor the engine runs OK as it always has and is smooth throughout the RPM range under light and heavy load.
When I switch to the distributorless system (I'm using Nodiz) there is a noticeable improvement in idle and running under load but when I'm on lighter loads the engine seems to 'stumble' now and again, I can feel this in the power delivery and see it on the tachometer.

I'm not suggesting there is any issue with the Nodiz system, I'm currently looking as possible issues with the setup.
One issue which has always bothered me is the link up between the ignition and EFi systems, one wire from the 14CUX EFi to the negative side of the coil through a 6k7(ish) ohm resistor. My understanding is that when the coil is triggered the breakdown of the magnetic field in the coil causes a 'spike' on this line. This pulse is what the EFi uses to trigger the fuel control. This is the same system used to drive the dashboard tachometer.

Obviously the single coil on the distributor system 'sees' all ignition events (4 per engine revolution) but due to the distributorless system using 4 coils each one only fires once per rev'. To get a pulse per ingition event I've linked the 4 coil negative terminals together via diodes and then run this through a zener diode (a circuit suggested in the NODIZ notes), this seems to do the job as the engine starts and runs (very well in certain circumstances) but I'd like to eliminate it as a possible cause of the 'stumbling' so I have an idea but just looking for advice on whether there is any downside to what I'm suggesting.

What I'd like to do is also power up the distributor system but only to provide the ignition pulse, the sparks to run the engine will still be supplied by NODIZ. Obviously/presumably I need to do something with the spark from the single coil liked to the dizzy, so the question is:

If a clamp a spark plug to the side of my engine (or chassis) and run a lead to this from the single coil am I likely to suffer any negative consequences simply having this plug sparking away in fresh air under the bonnet?
I'm thinking of possible EMI issues but also any negative effects on the coil or amplifier? I don't see why I would as I can't see that system 'knowing' the difference between a single HT lead with one plug on the end and going through the distribtion system in hte distributor.

Bear in mind I'm only looking to run this as a test to eliminate (or otherwise) the diode circuit I'm using.

As always all input greatly appreciated.

On the older 4CU, the injection triggers happily from a standard 5 volt signal if you remove the series resistor. I know this because I use my MegaSquirt Stim to test a suspect 4CU ECU.

I assume your Nodiz has a tach output for a modern rev-counter that doesn't need the high volts pulses from a coil? If so, use that.

DaveEFI wrote: ↑Wed May 05, 2021 5:39 pm On the older 4CU, the injection triggers happily from a standard 5 volt signal if you remove the series resistor. I know this because I use my MegaSquirt Stim to test a suspect 4CU ECU.

I assume your Nodiz has a tach output for a modern rev-counter that doesn't need the high volts pulses from a coil? If so, use that.

Hi Dave, thanks for taking the time to read all that, only just realised how long it was!
That’s possibly worth a try, I believe the tach output is actually 12v but presume that should make little difference?

The only issue I have with that is when I tried feeding the dashboard tachometer with it on the initial setup it read half speed, don’t know if this was an issue with the way the tachometer works or if the signal was half frequency (there is a setting in the software for 2 or 4 pulses per rev but does not work).
So basically I’d like to work out if this pulse is the issue or not by simply continuing to run it of the distributor (because I know it’s working well). If that proves to be the issue then I can work on getting a more appropriate pulse, if not then I need to chase down what is causing the issue.

FWIW, I never got the diodes trick to work 100% with the rev counter on my SD1, and found there were several circuits to do this job. But didn't experiment, as found another way to do this which was neater. So can't really advise on problems with it. Did you buy it ready made, or a DIY unit? It's possible a diode etc is iffy, I suppose.

Have you asked Nodiz why the x2 feature on the tach output doesn't work?

However, I don't think you need a spark plug in place to get the primary flyback pulse from a coil.
But if you replace the coil with a dummy load resistor, you should get a 12v pulse from the distributor. Which you could feed to the ECU with no series resistor.

DaveEFI wrote: ↑Thu May 06, 2021 8:56 am FWIW, I never got the diodes trick to work 100% with the rev counter on my SD1, and found there were several circuits to do this job. But didn't experiment, as found another way to do this which was neater. So can't really advise on problems with it. Did you buy it ready made, or a DIY unit? It's possible a diode etc is iffy, I suppose.

Interesting, I have seen a couple of other circuits including one suggested on my previous thread when I started this project. I think it was suggested by your good self in fact. However all I’m looking to do at the moment is work out if the problem is in this area, if it is then I’ll explore the other options for a”production” solution. If not then I’ll look at other areas.

DaveEFI wrote: ↑Thu May 06, 2021 8:56 am Have you asked Nodiz why the x2 feature on the tach output doesn't work?

I am also exploring this route. Don’t know if there is actually an issue there (the setting may now be part of selecting an 8 cylinder engine), ideally I’d like to put a ‘scope on the signal so I know if there is a problem there or not, unfortunately I don’t have access to one at the moment.
Using this signal would be my preferred route with a “controlled” signal form rather than a “spikey” signal of unknown voltage level and form.
But as I say I’ll be focusing on that once I’ve worked out if that is the problem area.

DaveEFI wrote: ↑Thu May 06, 2021 8:56 am However, I don't think you need a spark plug in place to get the primary flyback pulse from a coil.
But if you replace the coil with a dummy load resistor, you should get a 12v pulse from the distributor. Which you could feed to the ECU with no series resistor.

That is possible but I’m looking at this as a diagnostic procedure only, no a full time solution (I’m only expecting to run this configuration for 30 minutes or so), I know that the distributor is working well in this area so I want to use “as-is” any changes I make could introduce additional problems. For instance if I change the coil for a different type of load I have no way of knowing what effect this may have on the fly back pulse.
I know it can be argued that what I’m suggesting is a change but I can’t see how not splitting the HT signal with the rotor arm will influence the coil, but as I think is obvious, this is not my area of expertise!

OK, in the absence of anyone telling me the planet will stop turning if I do this I decided to “suck it and see” (nt literally obviously, that would be dangerous with that kind of voltage!).

Not been able to test the car in action due to weather (no roof on this thing) but it starts up and idles which will not happen if 14CUX doesn’t get this signal.
Didn’t actually put a spark plug on in the end, simply put a plug lead into the coil and put the rubber boot on the other end onto a bolt head with a gap to the electrode.
Interestingly (or not) I’m still running the tachometer from the diode setup on the NODIZ as I’d like to see if I still get a “stumble” there even if the engine runs OK.
If the problem is still there I’ll remove the diodes completely just in case they are causing issues with the coils or NODIZ unit.
Fingers crossed!

I sort of assumed you'd traced the problem to the connection to the 14 CU. But it sounds like it could within the Nodiz itself. Can it log the trigger? Does that look good? Does a log show an interruption of any sort at the stumble?

Are you using resistor plugs? They seem to be essential with after market programmable systems.

The other likely one is the VR sensor and output. Is the cable from the VR sensor well clear of any plug lead? Is the gap pretty consistent at about 1mm?

DaveEFI wrote: ↑Fri May 07, 2021 8:41 am I sort of assumed you'd traced the problem to the connection to the 14 CU. But it sounds like it could within the Nodiz itself. Can it log the trigger? Does that look good? Does a log show an interruption of any sort at the stumble?

No. This process is to try and rule in/out whether the feed to the 14CUX is causing the problem. If not then work in other areas to see what is causing the issue.

Don't know if you're talking about the 'trigger' to the NODIZ from the 36-1 wheel/VR or the trigger to the 14CUX from the ignition.

Looking at the 14CUX trigger, I can output a log file from RoverGauge however I'm not sure if it will be easy to extract any meaningful information even it it is there. I'd expect this to show up as a drop or raise in RPM, but may be difficult to tell if it's 'real' or a "stumble".

I don't believe there is any logging capability in the NODIZ software although what I should be able to see is a "Lost Sync Count" which I'd assume would increase if it was seeing any issues in the pulse stream from the VR.
Need to rig up a laptop in the car for this, problem is out of 3 available only one has a half decent battery and that is only good for about 30min's. Should be enough though.

DaveEFI wrote: ↑Fri May 07, 2021 8:41 am Are you using resistor plugs? They seem to be essential with after market programmable systems.

I am using resistor plugs, NGK BPR6ES at the moment.

DaveEFI wrote: ↑Fri May 07, 2021 8:41 am The other likely one is the VR sensor and output. Is the cable from the VR sensor well clear of any plug lead? Is the gap pretty consistent at about 1mm?

Once I can rule out the interface to the 14CUX this will be one of the items on my list.
As I've mentioned above I may be able to get some indication of this being a problem monitoring the Lost Sync Count in the NODIZ software.
The gap is good at around 0.25mm and consistent. Do you know of anyway I could monitor the output of that? Obviously a 'scope is the obvious way, but I'm thinking if there were some software available?


Anyhow, progress (or lack thereof) so far.
I ran the car this morning with the 14CUX powered from the original single coil as described above.
Still no improvement, although the car did start and run so 14CUX was happy with that signal.
This tells me that the "stumble" is being caused by the ignition rather than the fuelling - I think!

Is my diaode network causing a problem.
I decided to cut this out completely so that NODIZ is not running anything unnecessary, this seemed to improve matters.
There were no major stumbles (where it feels like the engine is about to die) although there were some minor ones.
These mainly seem to be on throttle transitions, particularly increasing slowly.
These could be down to mapping (which is still only at the starting point) or the MAP signal which I notice does seem to suffer a lot of ‘activity’ on the gauge in the NODIZ software. I’m going to look at fitting some sort of damper to try and address that.
Although I still need a bit more testing this would seem to indicate that either the NODIZ or the GEMS coils don’t like the diode system being in place (or there is a problem with my circuit).

Still not completely satisfied, I need a longer drive but unfortunately weather stopped play.

I did also try feeding the TACH wire from the NODIZ directly to the igntion trigger of the 14CUX but this did not work, the fuel pump did not even re-start on cranking which happens as soon as this pulse is received.

The manual describes this output as:
The NODIZ Pro™ features a 12V @ 50% duty tachometer output that can be used to drive modern or aftermarket tachometers. Note for 4 Cylinder Modes, it will pulse TWICE per revolution, for 6 Cylinder Modes, it will pulse THREE times per revolution and for 8 Cylinder Modes it will pulse FOUR timers per revolution

I have mine set at 8 cylinder mode so should be getting a square wave from 0v to 12v with equal times high and low with 4 cycles per engine revolution. Which I work out should give me around 53hz at the idle speed of 800RPM (unless my math's is letting me down again), I really wish I had a silly-scope to test this!

I'm going to try a curcuit to use this signal, such as descrived in this thread viewtopic.php?f=11&t=15611&start=30 at 14-Oct from yourself Dave. I've also found some information where this whole circuit was built into the relay case which looks like a neat way to do it.

I've had a look at the Nodiz manual, and that does say anything under 1mm should be OK as an air gap. Thing with a VR sensor is the output level rises with speed - a bit like a dynamo. And since your problem is at higher revs, I'd first try increasing the gap (if this is easy) as it costs nothing. (In the Megasquirt manual it mentions having to add a series resistor if the output from the VR sensor is too high - but of course MS has to cope with a wide range of VR sensors, and I'm assuming you're using the Nodiz recommendation) I'm also assuming since Nodiz comes with a loom, it has screened cable between sensor and processor, with the screen only connected at the processor end.

I do have a very good but ancient 'scope in the workshop, which is far too large to use in a car. So bought a very basic chassis only battery one from Ebay for under 20 quid (and put it in a case) which is fine for looking at such signals on the car.

As regards building the high voltage pulse generator just to drive the 14CU, it would seem to me using a sledgehammer to crack a nut. It's different with a rev counter designed for this, but the 14CU is clearly not, given the external series resistor.

DaveEFI wrote: ↑Fri May 07, 2021 1:57 pm I've had a look at the Nodiz manual, and that does say anything under 1mm should be OK as an air gap. Thing with a VR sensor is the output level rises with speed - a bit like a dynamo. And since your problem is at higher revs, I'd first try increasing the gap (if this is easy) as it costs nothing. (In the Megasquirt manual it mentions having to add a series resistor if the output from the VR sensor is too high - but of course MS has to cope with a wide range of VR sensors, and I'm assuming you're using the Nodiz recommendation) I'm also assuming since Nodiz comes with a loom, it has screened cable between sensor and processor, with the screen only connected at the processor end.

The connection between NODIZ and VR is indeed a screened cable (as supplied with kit) I've kept this as short as practical (less than 1m) and as far as I can from sources of interference.
I can increase the gap but this is not easy to do so I'll leave that down the list a bit.
I think the manual also mentions the gap being the thinkness of a piece of paper. 1mm is thick paper in my experience, more like card!
"The gap should be less than 1mm between sensor and teeth - the sheet of paper test is the usual method to confirm this, "

BTW, the problem is not only at high RPM (sorry if I gave that impression) it can be at any RPM and my testing has mainly been 1-3000.

DaveEFI wrote: ↑Fri May 07, 2021 1:57 pm I do have a very good but ancient 'scope in the workshop, which is far too large to use in a car. So bought a very basic chassis only battery one from Ebay for under 20 quid (and put it in a case) which is fine for looking at such signals on the car.

I guess I should look out for something simlar, would cut out a lot of assumption.

DaveEFI wrote: ↑Fri May 07, 2021 1:57 pm As regards building the high voltage pulse generator just to drive the 14CU, it would seem to me using a sledgehammer to crack a nut. It's different with a rev counter designed for this, but the 14CU is clearly not, given the external series resistor.

By "high voltage pulse generator " are you referring to the circuit I linked to? Doesn't seem like such a chore, a relay, resistor and transistor is all it uses.

In my simple world way of looking at things the original single coil/distributor system uses the same signal for both the tachometer and the 14CUX ingnition pulse so anything I substitute this should likewise be able to do both. Am I looking at that too simply?

In fact driving the tachometer seems more difficult that the 14CUX.
The circuit in the NODIZ manual indicates a diode from each coil (they only show 2 as its for a 4 cylinder application) followed by a Zener diode (in opposite direction).
When I originally put the system together last December I simply used 4 diodes, one on each coil, didn't put the Zener in there, linked this to the 14CUX and tried to start, the car started no problem. Later when I added the feed to the tachometer it did not work (car still started), I put the Zener in there and then the tachometer did. That suggests to me the tachometer is less tolerant of signal.

Also from the tests I've done today it seems it is not the signal to the 14CUX which is causing the problem but having the diode circuit attached, including the tachometer, once that was removed things seemed to improve (although I still want to test further).
For this reason I want to look at how I can get a suitable signal to replace the distributor/coil system ideally using the TACH output from NODIZ instead of this diode circuit. If that coil-resistor-transistor setup will drive my tachometer I'd hope it would also drive the 14CUX.

I've no idea why the TACH output from NODIZ won't drive the 14CUX (with the 6k7 resistor removed) but this is definitely not seen as a valid signal (regardless of accurate frequencey) as it does not even start the fuel pump. Any valid pulse should do that I think. Perhaps the 14CUX needs to see a 'spike' rather than a sustained signal like a square wave.

I've tried to find a schematic of the 14 CU online and failed. But do have one for the 4CU. And as I said that triggers happily off a low volts square wave. Be interested to see what the differences are on the inputs.

BTW, have you got the Nodiz grounded to the same place as the 14CU? At the rear of the cylinder head?

DaveEFI wrote: ↑Fri May 07, 2021 3:17 pm I've tried to find a schematic of the 14 CU online and failed. But do have one for the 4CU. And as I said that triggers happily off a low volts square wave. Be interested to see what the differences are on the inputs.

Bear in mind this is 14CUX I'm using, also known as Hotwire, 14CU is an earlier system and I'm not sure if it was used in the UK, I've only heard of it in regions such as USA, I don't know what the differences are or if these are in the area we're looking at.

If I've understood correctly 4CU is the flapper system which is quite a bit different as it's analogue rather than digital in which case the handling of all input signals would be very different.

DaveEFI wrote: ↑Fri May 07, 2021 3:17 pm BTW, have you got the Nodiz grounded to the same place as the 14CU? At the rear of the cylinder head?

My main ground is on the chassis with a short, thick cable going from that to the rear of the cylinder head. NODIZ is grounded there and a number of wires (3 or 4 I think linked to a single 'eye') from the 14CUX loom are also grounded there. The 14CUX also has ground wires at the front on the engine block, opposite (drivers) side.


Just did a bit of digging, looks like 14CU was used in 1989 only then superceded by 14CUX which seems is a bit more capable and also has OBD1.

Just joined forum because......
I have a Rover P6 V8. Has SD1 heads and front cover. 14CUX EFI system.
Just had the heads skimmed and refitted them with composite gaskets. Bought a Nodiz Pro and ford coilpacks.

GDcobra, any progress?

Geordie Jim wrote: ↑Fri May 21, 2021 6:43 pm Just joined forum because......
I have a Rover P6 V8. Has SD1 heads and front cover. 14CUX EFI system.
Just had the heads skimmed and refitted them with composite gaskets. Bought a Nodiz Pro and ford coilpacks.

GDcobra, any progress?

Hi Jim, good to hear of someone on the same/similar journey. Will be good to know if you progress better than me.

My unit is currently back with ME having the firmware re-flashed to address the “small” stumble I still have (the thought seems to be that this may have been “upset” due to using the beta version of the EZ-Tune software which is NLA), however I don’t think this is going to help with getting the ignition pulse for the EFi or tachometer. I’m still not sure what is going to be done about that so will be interested to hear how you get on. I have proved that this is due to using the diodes to try and get the ignition signal - with them removed the major”stumble” goes away. The re-flash May sort this but I’m not holding my breath.

I tried using the “relay coil & transistor”, this only gave a half speed reading on the tacho and the EFi did not run at all, didn’t even re-start the fuel pump. Don’t know if this was due to the incorrect pulse frequency or form, in any case it’s a non-starter as the frequency is wrong. It seems there is no way to set the pulse/rev’ to 4. So it looks like I’m back to trying the diodes again, I’ll try re-routing the wires to prevent any cross-talk between wires. If that doesn’t work then I’m out of ideas.
Maybe you will fair better with yours, perhaps the Ford coil packs will give a different result.

Are you using MAP or TPS for load?

Check your clocks, some will have a little potentiometer on them where you can adjust the rpm.

I know one of my old Ford clocks did.

Or if you're using Rover clocks...maybe hunt for a Rover tacho from a 4cyl car ?

Or this is a simple pulse adjuster ( I use it for my speedo...well the older DIY version. )

https://www.jaycar.co.uk/speedo-correct ... medium=web