Probably not that simple of a question but what is "Noise" and where would it be generated from on our car.

IF YOU WANT JUST THE MAIN QUESTION and not the reason why the skip to the last bit as its a long thread.. sorry guys.

Basically as a test our Megasquirt install worked perfectly with the CPS and coil pack etc wired up on connector blocks untill we knew the polarity was correct, with very minimal shielding of the wiring intact etc.
It was then all soldired and shrink wrapped properly and the car ran for over a month, during which time we had issues with the spark plug electrodes breaking internally, and eventually making it sound like a Subaru.. but it still ran and after replacing the offending plugs it was fine.

One day after a bit of updating and setting of the MS in a laybye on the laptop and the engine possibly getting a bit hot as the electric fan had accidently been dis-connected it conked.
I let it cool down and tried to get it re-started trying all sorts with no luck so it was recovered home.. when with a fresh set of eyes and trying all the spark plugs not just one there was no sparks on cylinders 2&3.

We tested it all out and found MS wasn't giving out the Earth signal to the coil pack to make it spark, so suspecting the coil drivers we sent it back. We were told nothing was found to be wrong with it but the coil drivers had been changed regardless.
When we recieved it back it would spark on cylinders 2&3 what appeared to be at the corect time the first time you turned the engine over from cold.. then every time after that it would spark at random and even constantly ark on those plugs.

So with it working from cold we suspected something couldn't take the load compared to being tested on a stimulator or lighting a LED for example so back it went again and we were assured it had ran all night firing an actual coil pack and plugs.

We've since got it back and sent off various log files etc and were being told its "noise" on the CPS signal... we've been advised to fit a capacitor on the +feed to the coil pack. We've tried new Coil pack, HT leads,Spark plugs, electrical plugs, a new CPS, moved it closer, a new lead for it, we've even made up a new DB37 incase there was a fault mm's from the plug or something daft like that and nothing seems to have made it any better.

So basically where is this "noise" likely to be coming from?, is there a way to test for it?(I know if its really bad it would make the raido buzz etc which it doesn't)
And why has it suddenly appeared when I was driving down the road it if was the reason for the original fault that made me break down! and it never affected it when first wired up roughly with connector blocks.

P.S as you can probably tell by the length of this post and the Time that its really getting to me now!

Electrical noise can be one of the nastiest things to locate and get rid of, whether it's in a hi-fi or a car! An oscilloscope is the best way to check for it, but they're not the most common tool in a garage...

There are a few things you could look at:

* Signal wires (from sensors) routed too close to high-power electrical equipment - alternators, coils, ignition leads, and their wiring.

* Screened cable from sensors earthed at both ends - you should only earth one end.

* Power leads to the ignition system should be as short as possible. The capacitor they suggest would help to reduce unwanted noise pick-up, as would a ferrite ring just before the control unit.

* All earths for the control unit, coil and so on should also be as short as possible.

That should keep you busy for a while...


HTH,
David

[Edited on 30/12/07 by David Jenkins]

Thank you for your suggestions,

We've tried quite a few things on it like moving the MS earth directly to the battery etc,

There standard ford CPS screened cables with a plug on the opposite end therefore should only be earthed at the MS end.

Thats the only screened cable on the car, but I do have some more would it be any use using the screened cable to wire the coil pack up with? and earth it at the engine block end.

Can anything be done really temporary like wrapping wires or components in tinfoil for example, I know how stupid that sounds... but thats how desperate I am!

It never had this "noise" before so I can't understand where its come from.

The other thing you could check is whether everything else in the car still has good connections! Earths and power.

True, I haven't noticed anything that doesn't work or is dim etc.

I suppose the main culprit would be the starter motor earth, but that turns over a treat, and wouldn't have been in use when the car originally faltered.

Are there any other tip/tricks?
I vaguely remember something about if signal and high power leads have to cross they should be at 90 Deg.

It's gonna be very hard to trace. Even with an scope all your gonna find out is whether or not you have noise. Your still gonna have to work out where it's coming from. Good luck.

Looking at your Tooth trigger logs you made for me using MTx the noise is clearly on the trigger cable, or it is down to the engine cranking over at a very erratic rate. It may be worth linking 2 batteries up to see if that helps to give a constant cranking RPM. Also as has been suggested, get new plugs and leads, connect ECU ground to the engine and ensure engine is very very well grounded!!
Phil

The car's got a brand new Calcium battery on that should be man enough for the job, but I seem to recall something about the standard starter motors on the pinto stuggeling on high compression engines, so maybe it does slow slightly when its coming up to fire.

It does seem to turnover fine tho and doesn't struggle half as much as my Vauxhall XE starter struggles on the tin top.

I doubt i'm going to get in to the garage today, or certainly untill later.

Its already had brand new:
Halfords HT leads (specifit to the coil pack)
Bosch Sierra injection plugs 7R's (slightly higher heat range)

And alternative:
Coil pack
Wire to coil pack
CPS
Wire to CPS

The CPS has been moved to 0.75mm and direct in the centerline of the toothed wheel and stays true.

We've had the MS connected direct to the battery with both the +ve and -ve and it didn't appear to make any difference.

I've just been thinking aloud.. with regards the CPS, if 1 wire is an earth to the CPS and the other is an induced signal back...
If the earth to it is poor then the induced signal back from it is also likely to be poor.
This earth comes from the MS if for whatever reason in the plug or internal MS wiring, soldierd tracks etc it was a little weak could that cause it?.

Can I put the CPS signal earth straight to the battery -ve, or to the Chassis very close to the CPS?

Point to note, on shielded cables you should only earth one end of the shield, not both.

Other thought, disconnect or isolate anything else that uses power so knock off the fan belt to isolate the alternator, unplug the dash etc, the only cause of noise then becomes the EFI system.

I,m feeling for you, this has been months ...

Regards Mark

quote:

---

Originally posted by mark chandler
Point to note, on shielded cables you should only earth one end of the shield, not both.

Other thought, disconnect or isolate anything else that uses power so knock off the fan belt to isolate the alternator, unplug the dash etc, the only cause of noise then becomes the EFI system.

I,m feeling for you, this has been months ...

Regards Mark

---

Yup a good suggestion Mark and your sympathies are welcomed lol, I was just thinking of unplugging the alternator, but taking the belt off would be better.

Being of the 7 type there shouldn't be many things that are live with the igintion untill there switched on, but I can always pop the fuses out to be sure.

Then if sucessful plug things back in untill it returns

I don't suppose anybody has any examples of what a Tooth logger graph should look like to know what i'm aiming for? I've had a quick look but not found anything.

[Edited on 30/12/07 by MkIndy7]

We've been out and tried the things that have been suggested and still no luck.

There is now only the bare minimum of electrics connected, isolated by pulling all fuses that aren't nessacery to start the car, the alternator has been totally removed and unplugged as well.
There is also No HT side connected or coil pack plugged in.

Were now testing purely by the Mexatunix software that show's what its reading from the CPS via the Megasquirt, we've also tried adjusting the MS pot.

We've tried with a big Jump lead direct from the battery -Ve to the block, and also putting the MS direct to the block.

We've also tried connecting the MS to different battery completely.

Could the noise be coming from the starter motor itself? thats about the only thing left! were just leaving it to cool at the moment as its got a little warm with keeping turning the engine over.

Is there any way to shield such a thing? wrap in it tin foil, cover it in a baking tray, anything!.

Were also now thinking of taking it to a friends house that has a working Megajolt install, connecting it to his CPS and running the MS off its own battery and seeing what we get.

UPDATE:

We've been and fitted the Megasquirt to our friends working Megajolt setup.

First turnover of the car with the plugs in the engine the car fired!
But the owner turned it off as was running really rough and probably well out of time as it was a Triumph Spitfire.

Then every sucessive attempt after that it wouldn't start and the CPS signal was showing the same pattern and noise as when fitted to our car.

We have also noticed this strange senario where it will fire properly for the first 5-10 secs from dead cold, and its happened ever since it faulted.

This isn't unusual but no-one seems to take any notice at the forum, except the posters who have the same problem.

I have the same, wasted spark on MS V3 and tower B has either weak or intermittant spark.

This with everything changed twice, all grounds checked and renewed.

It works fine on single spark output A with direct coil drive and distributer but change back to wasted spark and the problem returns.

I've tried scoping everything and there is no noise present on any signals. MTx now shows the triggers and they are spot on.

I've yet to take the MS from the car to investigate the ground arrangements on the board but something does not seem right when quite a few report problems with twin coil output B being weak or non existant.

quote:

---

Originally posted by Rob Lane
This isn't unusual but no-one seems to take any notice at the forum, except the posters who have the same problem.

I have the same, wasted spark on MS V3 and tower B has either weak or intermittant spark.

This with everything changed twice, all grounds checked and renewed.

It works fine on single spark output A with direct coil drive and distributer but change back to wasted spark and the problem returns.

I've tried scoping everything and there is no noise present on any signals. MTx now shows the triggers and they are spot on.

I've yet to take the MS from the car to investigate the ground arrangements on the board but something does not seem right when quite a few report problems with twin coil output B being weak or non existant.

---

Thanks Phil,

I'm unable to just at the moment but anyway they were exactly as the manual picture. Classic text book.

Rob

Phil,
email extract :

Went out to garage tonight with my dual beam scope, this allows me to see both input waveform and output waveform onscreen at the same time.

Cranking with no coilpack connected. Superb text book VR signal in. Superb sq wave output to input of VB921. Switched to overlay mode and the trigger point was correct, going positive to negative. All well so far.

Connected coilpack again and plugs. Attempt to start the engine and lo and behold 2,3 not firing again.

Scope channel 1 connected to Spark A output, scope channel 2 to spark B output. Wow !! super noise signals! Many combined AC signal waveforms present superimposed on each other !! Spark B worst affected, signal is there but equal level noise signal present masking it, hence no fire from coil. The coil would probably be full on all the time and unable to discharge, hence coil would heat up as well.

Spark A has noise but at one third of the level so there is still a discernable trigger for the coil leftover. Hence it will fire.

Poor running and misfiring obviously caused by occasional synching of spark B when noise does not mask signal completely.

I'm almost certain that I could see the VR input signal waveform mixed in there !!! There is also a severe pulse, one shot, waveform every now and then, that also created inlet backfires.

More investigation needed.

I can't believe I've not seen this before ? Often though if weathers bad I won't push car out of garage and the fumes otherwise choke me. So I then test at cranking. Maybe that's why I've missed it.

Rob Lane

Tried again with alternator disconnected.

Pics are poor, I don't have anyone here to give me a hand at the moment. It's hard trying to keep it going on two cylinders and take pics and adjust scope.

Spark A VB921 pin1 input

[Edited on 4/1/08 by Rob Lane]
Rescued attachment SparkA_VB921pin1.jpg

Spark A VB921 pin 2 output
Rescued attachment SparkA_VB921pin2.jpg

Spark B VB921 pin 2 output
Rescued attachment SparkB_VB921pin2.jpg

Spark B VB921 pin 2 output different timebase
Rescued attachment SparkB_VB921pin2-2.jpg

I never noticed but it looks like I left the combiner on in the scope, hence straight trace.

Hi both of you now

pop both those MS's to me and I will try something on them, the VR input components changed a while ago, I wonder if this is a problem, Ive only found out recently that they changed, so lets try putting the old ones in it and see what it does, email me for my address
philip.ringwood(at)ntlworld.com
Of course it wont cost you anything, lets get to the bottom of this,
Phil

[Edited on 5/1/08 by daxtojeiro]

Thanks for the offer Phil but I will plough on.

The problem is diagnosed at the moment, a new coilpack is required. Tower B is defunct, it's dead, it is no more, it's shuffled off this mortal toil etc etc.

I have one on order.

It's a very common problem on Ford coils apparently. Doesn't go immediately but breaksdown over a period before failing completely.
The primary impedance changes during this time, so probably loading the VB921.

When the new one arrives I will dissect the old one.
I'm curious as to whether there is a 'boost' diode in secondary or what they have on the output to successfully divide the load to the two sparkplugs.

quote:

---

Originally posted by Rob Lane

I'm curious as to whether there is a 'boost' diode in secondary or what they have on the output to successfully divide the load to the two sparkplugs.

---

what is the reason for not using an standard EDIS 4 setup???

its commonly known that ou need some caps fitted to the coils.

also the edis wil have some internall...

Tks

EDIS setup cannot rev limit which is important to me when engine is capable of 10k but should stop at 8k.

It's also a bit of a compromise when in fact the MS is capable of doing its job.

New coil fitted and so far it's working OK.

All coils should have a cap fitted as they are inherent electrical noise generators.

It wouldn't have mattered about EDIS anyway in this case as it also fires the same coilpack and would have behaved the same.

EDIS is a very stable, reliable method of getting wasted spark and it is the route I try to get people to go as it simply works straight out of the box. Driving the coils from the ECU means you must use resistive plugs, a cap on the coil, etc, but that still functions fine and is very reliable too. Ive been running my V8 directly driven in wasted spark for a few years now and Ive done loads of cars the same all without problems. Ive even directly driven a 27L V12 with 24 spark plugs, so 12 spark outputs (wasted spark) and that runs great at around 1500BHP!!
rev limitting is still possible using EDIS, you simply reduce advance and cut fuel.
Phil

quote:

---

Originally posted by Rob Lane
EDIS setup cannot rev limit which is important to me when engine is capable of 10k but should stop at 8k.

It's also a bit of a compromise when in fact the MS is capable of doing its job.

New coil fitted and so far it's working OK.

All coils should have a cap fitted as they are inherent electrical noise generators.

It wouldn't have mattered about EDIS anyway in this case as it also fires the same coilpack and would have behaved the same.

---

yep, you can cut fuel only, or fuel and spark or spark only.
Phil

Evening good people of the Locostbuilders again.

Well the MS has been off for another good looking at and repairs and it looks like its either a new one or its had some quite major changes so Phils definately pulling out all the stops to try and get this thing working again!. (thanks Phil)

Its now sparking alot better and not continualy sparking after 5 secs or so of cranking but it still won't start the engine .

In the past I have written about it being better when the plugs weren't in the engine, but this was before I knew about the MegaTunix tooth logger.
I've just had a go with this again tonight and here are the logs:
Plugs in:

plugs in
Plugs out:

no plugs .

Soooooo why is the compression on the engine effecting the signal so much?

There is absoloutely no electronics connected other than those required to turn the engine over and power the MS and no HT side or even the coil pack connected.

The MS unit has in the past been fed from its own 12V stand alone battery so it shouldn't be voltage drop when cranking.

I was thinking maybe the noise could be from the starter itself but the MS has been tried on a friends car that has a working Megajolt setup and shows the same noise using his CPS etc.

The battery could do with a charge as its only turning over at a steady and quite constant 300 ish RPM but i'm sure its been higher.
The only other thing I can think to try is some new battery cables to the starter, but the existing ones are Industrial Fork Lift truck cables so should be man enough and we've had the jump leads from -ve to the block as well.

One last observation, and its true for 2 CPS leads i've got, the Gound pin (no2?) shows continuity with the Shielding and also the cars earth, not sure if this is correct or not?

As ever any observations, hints, tips, questions etc are welcome.

Ian

Hi Ian,
Im afriad youll have to continue this with me using EMail as I have a rediculous TRADE status now and a warning under each of my posts. Ive tried to contact the owner twice and had no responce so untill its sorted Im not helping out on the forum, I have other forums that I help out on and run and dont have a shouty label under them. If anyone knows the owner of ths site could you give him a poke please,
Phil