off soon to try the old squirty thing on an airfield.
I STILL think there is going to be a spark issue. It seems to peak and fall but I'm hoping it's because the fuel is Waaay to rich and it's blowing the spark out.
trouble is at steady revs the in car rev counter reads 4K and the megatune one reads 2.5K revs
however in megalog viewer you can see the revs peaking and falling and the logviewer says I'm reaching up to 4.5K revs.
I'm getting really pissed off with this now and am seriously condidering going 45's and some sort of spark system (old cavvy 8valve dizzy has been suggested.)
Perhaps a rethink/rewire of the squirt using more shielded cables would be an answer as well but can I face the prospect of ripping it all out and re doing it??
MSQ and log of this afternoons fun to follow later with some photos as well perhaps. After that anyone care to comment on what perhaps I can do to solve it all.

[Edited on 24/1/09 by omega 24 v6]

a) what engine / application

b) have you got the VR sensor wired the right way round?

OK here goes with all the info for your criticism/advice/perusal.
Setup is
vauxhall xe redtop with gsxr throttle bodies (750s) and megasquirt v2.2 running extra code (029y3) spark is with a vr conditioner and 2 vb921's (well bip's now) which are within the squirt case as per the photo


Description

So it didn't go to well here's a couple of pics


Description



Description

It realy does look to me like the spark is not right. The car starts and idles beautifully no problems but as soon as there's a load on it or the throttle opens it's really all over the place. In megatune the rev counter (rpm gauge) does not read at all right IMHO.
I don't really know where to go from here although I must say todays outing has put a smile on my face.
My thoughts are
1. move the spark and vr sensor wiring/conditioner to a seperate box/location from the ecu.
2 Change the wiring so that the coil is wired with screen cable. Or perhaps wire everything with screened cable.
3 Junk it and put on carbs and dizzy although it'd be a shame to spend upwards of a grand to solve what I THINK is a spark problem.
4 Some of you will be able to analize my msq and datalog and tell me I'm missing something fundamental (I'm 99.9% sure it's a spark problem)

PLEASE HELP

Never got out of first gear or above 7mph


And here's the datalog for megalog viewer

[Edited on 17/1/09 by omega 24 v6]

From a quick look. Your req fuel loks very low especially as the 750s use the same injectors as the 600s and mine is around 14.

The advance table also looks like it has a lot of low numbers. Not sure what the xe needs, but I would expect around 12-14 at idle, and mid 30s from 4krpm upwards.

It does sound like you have some interference on your spark signal though especially with the erratic rpm readings.... I can probably get you some sheilded 4 core cable if you need some? I would say you need some between the coil and MS, and from your VR to MS as a minimum.

David

I have on the vr already
Not on the coil though but I do have some 6 core shielded (same as vr) so I could twist them in 3 pairs if required.
The reqd fuel is low. The initial calcs came out at 13.9 but the readings in the ve table at idle were around 6 and I'd read somewhere none of the readings should be much lower than 30 for some reason. At a reqd fuel of 13.9 there was masses of black smoke and soot out of the exhaust.
It's getting me down now

vr sensor mouting has to be rock solid; you should not be able to feel it flex if you push on it.

stupid question but the trigger wheel is cnetred? i.e. when it is spinning, the gap between the sensor and tooth is same for all teeth?

are you able to put a oscilloscope to see the VR signal? (I could have helped you there but you are other end of country )


>>>> rev counter reads 4K and the megatune one reads 2.5K revs


just saw that.

which do you think is correct?

try runing but disconnect rev counter totally as see what difference there is , if any.


also on your board, what are all those colouored wires for? is it possible becasue they are crossingn one another you are getting crosstalk?

[Edited on 17/1/09 by 02GF74]

I'd say megatune was reading wrong. These coloured wires are for the 2 vb921's top left of pic on pinboard. and there are also the cables from the vr sensor feeding the vr conditioner components on the pinboard. Wires go from there to the main board to led#s 17 and 19 tfor the spark outputs to the vb921's.
I do wonder if that's where the interfrence is coming from but as you can't see interference then WTF can I do. I don't have a scope. What size of scope would i need????

The on board rev counter is acewell and the cable is wound round an HT wire to give the output. I'd say the acewell was reading right. The datalog does show higher rpm's than megatune gauge as well.


[Edited on 17/1/09 by omega 24 v6]

quote:

---

Originally posted by 02GF74
vr sensor mounting has to be rock solid; you should not be able to feel it flex if you push on it.

stupid question but the trigger wheel is centred? i.e. when it is spinning, the gap between the sensor and tooth is same for all teeth?

---

sorry but that is MS goobbledegook to me - I am not knowledgeable in megasquirt to know what all that means.

Those transistors (vb921) are for fuel pump? Are they running to switch something - how many amps are they switching?

I see they are mounted on a heat sink - are you able to mount them outside of the box, again on a heatsink and with an
earth strap back to the box (if required, see comment below) to see what difference there is?

Depending on the device, the tag (and heatsink) is either ground or live - do you know what it is for these?

If it is meant to be 0 V, is the heatsink earthed?

Are you saying the veroboard is the VR conditioning? Why is this needed? I would have thought MS has VR input - why do you need this extra circuit?

I assume you have tried a different VR sensor?

Like wise with the vero board; again can you mount it away from the main box - I would say that circuit is far too big; are you able to redo it but make it more compact and run the screened cable from VR driectly to the board rather than via the multi-way connector?

note - without knowing what is the cause, ^^^ none of the above is guaranteed to work but what I am suggesting is safe and not too time consuming. A 'scope would help - nooen you know or on forum cannot help?

^^^ in case you have not worked it out, I am hinting that those long wires, running parallel to each other are not the best layout, and my suggestions
are to move them physically apart. Distance and crossing at 90 degree reduces crosstalk between them.

re: veroboard - I assume you have cut the tracks to the side of the connections so that the copper strip is the minimum length?

see my poxy attempt at a picture below.

The thick black wires represent components.

the red blobs are cut track.

The construction is such that the copper strip makes the connections but is the shortest length rather than the full width of the board - if that makes sense to ya.

Also on that board - is that really a 10 W 4.7 K resistor?!?!? for signal condtioning VR sensor????
Rescued attachment vero.JPG

board is fine and the vr is std vauxhall setup so its gotta beat my end i suppose.
circuit is as ms drawing and tied and tested

The wires from the coil drivers are very close to the input circuit wiring and could be causing crosstalk. Can you seperate them a bit more and maybe sheild the input circuit with some alloy sheet? I recall something about the VR conditioner needing to be adjusted to work correctly using a preset pot?
Otherwise I would fit a 36-1 wheel to the crank and convert to megasquirt and Edis set up.
Paul.

Hi Garry, car is looking great!
Don't know much about this stuff but what makes you think its a spark problem? Just if it is idling okay I can't understand why the sparks would be effected by anything at revs off idle? Unless I guess there is something chaging ie. advance or the fueling being too rich or lean off idle? Have you got a wideband lambda sensor to try?? At least then you could use process of elimination? Otherwise - a long shot but I've got my megasquirt board sat at home. Has been tested on the stim and is okay but never tried in the car (yet), your welcome to borrow it to try out? I think its a V3.2 though so might be different....I wouldn't have thought its worth converting to carbs and dizzy. Would have thought you could buy a secondhand (or perhaps) even new emerald or omex ecu for the same cost as a pair of 45's and well sorted dizzy. Guess if you get really desperate you could always buy yourself a micro-squirt? Let me know if there is anything I can do to help out. My dad's living up in Ellon now so I'm up in your neck of the woods every couple of weeks. Oli.

I hate to say it, and I'm sure it's been mentioned on other threads, but If I were you and worried about spark, I would consider going to an EDIS 4 set up....It takes all the guess work out of the spark set up.....If you later want to run the VB921's then fine, but to get you started EDIS is great......

I had loads of trouble with the VB921's (again, I know people have used them sucessfully), going to EDIS has given me a more reliable set up, less guess work.

Do you have a wide band fitted?.

Wide band is fitted yes. Edis may be an option at a later date but not for now. I'm not going to be defeated by this bloody thing ( I hope)

quote:

---

Originally posted by omega 24 v6
at steady revs the in car rev counter reads 4K and the megatune one reads 2.5K revs
however in megalog viewer you can see the revs peaking and falling and the logviewer says I'm reaching up to 4.5K rev

---

Have a look at this:

http://extraefi.co.uk/vr_setup.htm

Worth a try?!

David

Cheers for that Flak but mines a 2.2 board and doesn't have the pots. However I may have found something tonight that could be the problem.
I'll post tommorrow if it cures it.

Ditch the veroboard, I've tried at least 4 different setups on vero and in each case induced spikes cause problems once the revs rise. Even removing the copper tracks not needed to reduce stray capacitance didn't work.
On my V2 I used an error* daughterboard with dual LM's on board. Totally successful and has other useful hardware onboard.
Out of the six systems I've been involved with all had problems with VR circuits and final tweaking. Once set though they usually stay stable.
The V3 board is easier to setup. Not what you want to hear I suspect.

Well folks a dry soldered joint was found ( very luckily and by accident) on the veroboard in the vr circuit. It basically when moved disrupted the vr signal and gave all the symptoms (including flashing leds 17 and 19) that have been plaguing me for almost 4 months. Have I got it sorted?? Well I'm hopefull and this afternoon it's airfield run number 2 and a post will follow later on this afternoon/evening. There might even be a u tube video, if it goes well, in celebration of the end of my megasquirt woes.

HOPE I HAVE'NT SPOKE TO SOON

[Edited on 24/1/09 by omega 24 v6]

I've got everything crossed for you

Sounds like you have found the root of your woes...

Good luck!

David

well no video it never got to that point. It did however run and boy was it MUCH MUCH better. Did about 8 logged runs with megalog viewer and analized and accepted the new tables etc. Each time it just got better and better.
But after the last run and ve analize it wouldn't run at all again WTF
put the settings back to before and still the same So when it fires and you need to blip the throtttle to keep it going the afr is lean way way lean. I'm thinking either no fuel from injectors ( but why will it run blipping the throttle) or the sparks gone haywire again I'm off to look at the megalogs to see what's happening now. 2 loggs attached for perusal if your interested.

not running so well

must remember to save my msq more often (twat). I had reset my reqd fuel after the ve table higher figures had reached over 300. reset required fuel which updates the ve table right away but never saved the msq. So when turned of again and load new msq there was still a reqd fuel of 6 but at the nuch lesser ve figures resulting in total leaness.
Been out into garage and reloaded with reqd fuel of 11 and it runs again (well it idles without farting and stalling so it's gotta be right aint it????

Sweeet

woo hoo!!! glad you found it




.... did I not say that veroboard circuit was not ideal?

woo hoo!!! glad you found it




.... did I not say that veroboard circuit was not ideal?

quote:

---

.... did I not say that veroboard circuit was not ideal?

---

FYI on the VB921s - last time I saw Bill Shurvinton, I had a chat with him about them.

The upshot of the conversation was they could be temperamental and if possible he would recommend using EDIS.