Fuel injected 'A' Series 998 with MS and sequential siamese code
Turbo Mini Forums

Just a quick update.
Unfortunately I have not been able to get to our workshop to work on tuning this for the last month, but I have tidied the wiring loom, made my fuse box/ ecu and relay board.














It’s not quite finished yet because I have run out of solder. Also I need it in situ to finalise wire lengths.

I also have some basic adaptors/ flanges being made up to move the Throttle body from on top of the MPI inlet manifold to the side with a 90deg elbow.

It Should all be done by Saturday, when I hope to get the engine started again and get moving with configuring the Megasquirt.
That's if my wiring is up to scratch..

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

sickeningly tidy


show off


lol

"Turbo's make torque, and torque makes fun"

"did you know you can toast potato waffles?"

That looks the business!
I need to sort ours out, was waiting till we had everything to plug in but can get all the basics finished.

Last night’s log. Or at least one of them.
This was the best with injection1 at 60deg and Injection2 at 0deg. Req_fuel was set at 1.6



After a small incident with plumbing the fuelling system backward, I managed to get back on track. When I put the intake back together after some "refinements" to my initial set up, (read fixing air leaks) I got it totally wrong.
I had Put the regulator in the inlet side of the fuel rail pipe and to make matters worse, I even put the fuel filter on backwards (flow arrow going towards the tank..)
So, that sorted and a bit of tuning advice from Rod earlier in the day and I managed to move forward.
However, I’m not sure what the “spikes” that can be seen in the PW and duty cycle graphs are.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Looks good, AFRs look nice and close but there's no scale on the left for absolute units.

If by the spikes on PW and duty cycle you mean the ones at the very right of the graph, they co-incide with a fast openning of the throttle so will probably just be acceleration enrichment which you probably need to adjust as the AFRs rose at that point instead of falling slightly.

What's more interesting is the dip in both AFRs 3/4 of the way along - I can't see what they relate to ???

There's also quite a few other small "noise" type spikes on the traces, you may need to improve your grounding.

But you're making more progress than me

Schrödinger's cat - so which one am I ???

How do I set the scale in megalogviewer?

The dip in AFR's are due to me playing with the settings. I think I upped the req_fuel at that point, then lowered it because the engine sounded rough.
The little spikes were my concern. I hope I was not going to get earthing problems. But, I can reroute the signal wires and check the earths.
The cam trigger wiring is a bit close to the starter solenoid, hence the main incoming power..
I'm guessing because the spikes appear to be regular, it could be cam signal as opposed to general noise?

Still the cold start problem, but I'm adjusting the crank pulse settings little by little.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Here is something I've just noticed. At each spike of PW the battery voltage drops to 10.5V

Could this be significant?

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Don't know about setting scales in MegalogViewer as I use Winlog (TechEdge based software), I'm sure one of the MS experts will answer.

Without a scale the little spikes may be "normal" but the PW to battery voltage suggests otherwise, ie, PW is responding to a battery voltage drop - the code increases pulse width to compensate for low voltage at the injectors.....

Schrödinger's cat - so which one am I ???

You can't set a scale in MLV, just min and max.

As Rod said, the spikes are battery voltage compensation. It seems to be working well in fact.

I think you either need to look at your grounds or your power supply. Are you running this of an alternator or by some other means?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I've got a battery connected, but it is a spare, which is suspect. It may not be charging. What I have also noted is that as time the time goes on in the log, the average battery voltage is dropping, so I'll swap the battery for tonights work and check my grounding and perhaps swap out the alternator.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

This definitely looks like a grounding issue. Make sure that all you sensors ground at the MS and that the MS grounds at a very good car ground and then this ground goes to the battery with a sizable wire. You also want to make sure you alternator an battery are in good working order because both will easily create noise issues.

Also, you have to be careful how you ground your WBO2 controllers. Power ground has to go to the car ground but the signal ground has to go to the MS.

And, the AFR dip can't be related a req fuel change because there's no pulse width change. There doesn't seem to be anything other than AFR changing which is strange.

But besides the grounding issue, you're doing well.

Jean

http://www.jbperf.com/

A useless piece of information for anyone interested: The latest release of tunerstudio appears not to work particularly well with windows 7 64 bit. Just my own experience over the last couple of days.

I downloaded and installed V0.991.2 on Saturday and had intermittent problems on sunday with comms dropping out.
Today I downloaded V0.991.4 and since then I get a comms drop out every 30 seconds or so.

Unless there is something else updated causing the problem, like the USB serial lead driver again.

Hence I have spent all evening fighting with tunerstudio and got no where with tuning.
There's always Saturday...

Edited by Graham T on 16th Feb, 2010.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

There has been quite a few reports with comm problem with the last few TS releases (0.991.x). Phil is aware and is on the case but it may be good to report it on the msextra forum so he has more data points to work from to correct this.

Jean

http://www.jbperf.com/

Will do Jean.
That's actually a relief because another option was something I did not want to think about: something wrong with my megasquirt.

I think I have got to the bottom of the grounding issue. Nothing major, just the temporary battery leads. Earlier and later logs I took on Sunday show no spikes, so it has to be something between runs. which would have been disconnecting the battery for charging (due to the difficult starting).
Once I get sorted with tunerstudio, I'll be able to prove or disprove this.
Either way, its forced me to make good some of the temporary connections I made and re route the VR and cam sensor wires.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

It looks like the problem may have already been fixed. Tunerstudio V0.992 is now avaliable. I've downloaded it and will install and test with the jimstim tonight.

http://www.msextra.com/forums/viewtopic.ph...rstudio#p238065

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Interesting.....
For the last few days I've been testing on the JimStim on my desktop which TS auto-updated to 0.991.2 and every time I turned the MS off (to alter something) and back on again there would be no comms.
To get comms up again I was unplugging the USB/serial convertor from the PC and plugging it back in again, tedious but it worked.
Because Windows had also just downloaded a load of updates I assumed it was another Microsoft bug that was messing up my USB/serial drivers.
However, I've just installed TS 0.992 and the comms are back there every time like before !!!

Schrödinger's cat - so which one am I ???

Great news. I started getting concerned on Sunday when tunerstudio was reporting my engine revving at 26,500 rpm...

Removing the USB lead is what I was doing come to finish, but the dropouts were so often last night, by the time I had made config changes and started to burn, the comms had dropped, so I could not be sure whether changes were written to the MS or not.


Now to get the firmware to update with 64 bit O/S. The .Bat file does not work with 64 bit, so I have picked out the command line to use. Just not got the MS near me to try it on.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

The last week has been a bit frustrating...

I finally got Tunerstudio up and going with the latest build (I’ve had to reduce the frame rate to get it stable with no weird stuff going on) and also managed to flash the megasquirt with the latest code revision, but I have not been able to get the engine started since 2 Sundays ago.
Finally though, I think I’m getting back on track.
I decided to have a sort out and re route some of the wiring on Saturday, namely the cam trigger and VR wiring which is not in my main looms.
However, after making changes to the wiring of the crank and trigger sensors and some other wiring, I started getting VR signal dropouts on Saturday whilst cranking, which got progressively worse as the day went on, to the point that I lost my crank signal altogether. Some more testing on Sunday revealed that by unplugging the cam trigger, my VR signal has returned, so it’s fair to assume that I need to again rethink/ revise and renew the cam trigger wires.

Something that did come to light on Saturday is that I MAY have got my cam trigger timing out.
I know I have read a thread on here talking about cam trigger setting, but I cannot find it at the moment, So just as a double check, Can someone confirm that the cam trigger point should be somewhere more than 90 degrees before TDC number one?
I believe in my initial install I may have thought the dizzy rotates clockwise (not a clue why I would have assumed this or why I did not check it during the install), hence I actually had the cam trigger set after TDC number one. I'm guessing this would reset the injection sequence before number 3 cylinder.
Lots of “if’s”, but perhaps if that is the case, it may explain starting problems and why I could not seem to get the AFR’s as lean as I would thought I should??

Typing this has just made me think: If the cam trigger is pulsing at the same time as the VR missing tooth, what sort of problems with signal would be expected?

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Not sure what the exact effect is of having the cam pulse at the same time as the missing tooth but I know Jean has said make sure you don't.....

I have mine nominally 40 degrees before the missing tooth so 130 degrees BTDC on the No1 compression stroke.

Re the loss of VR, have you tried re-adjusting the pots ??? Don't assume that a good working setting from the JimStim will also be good on the engine - on mine one of them has to be 7 turns different.

Also, I can't see why there should be any relationship between the VR wiring and cam sensor wiring. The VR pickup isn't a powered device, it generates its own (small) voltage without any power feed from the MS. The Hall switch has to be powered but I can't see why disconnecting it should affect VR unless its a shielding issue on the VR cable and the spike from the Hall switch operating is affecting the VR signal. Did it go completely and where were you looking at it, ie, just in terms of loss of synch in TS or were you actually looking at the waveform on a scope ???

Schrödinger's cat - so which one am I ???

The lost sync was on TS. I have not bought the scope yet..

I had No sync in tunerstudio and no ignition diodes flashing on the MS during cranking.
It went completely, I was getting 0 rpm on the rev gauge as opposed to the Usual 1 or just a little higher rpm flitting about if the pots are not exactly right, or the 248(ish) that I get when cranking and sync’d correctly.
I adjusted the pots but got no change.
Once I had put the pots back to their original setting (which is the same as the setting for the Jimstim), then disconnected the hall sensor, the sync came straight back and my ignition LEDs started their alternate flashing again. (It did not start because the plugs were out.)

When I adjust the cam trigger angle to its correct position, I’ll make sure I’m not near the VR miss tooth pulse.
If that does not change things, I’ll check the cam and VR cable screens and earths.
It was running fine until I removed bits to tidy the wiring. When I first had it running, I was using ordinary thin wall wire for both VR and hall sensors, which worked fine. Now I have proper screened cable, I’m getting these issues. There must be something simple there somewhere that I have changed.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

It seems like you have created a wiring fault on the Hall switch circuit when you re-wired it.
Check it carefully as, if it stops everything from working when you plug it in, it may have a short somewhere and be dragging the MS voltages down.
There is no need for screened cable on the Hall switch, just the VR and make sure the screen is grounded at the MS end only.

PS - I see you've had your user ID changed too :)

Schrödinger's cat - so which one am I ???

Blimey that was quick. I only just saw the PM saying my user ID was changed. Thanks Alex

Could a short be why I was seeing the spikes (voltages dropping to 10.5) the last time I had it running?

First job tonight, re wire the hall sensor.
Second job - sort out the timing.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Schrödinger's cat - so which one am I ???

The hall sensor is powered with 12v from a separate fuse in my fusebox. I have a 2amp fuse in it at the moment.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

In which case, less likely. A short, even a very fleeting one should take out a 2A fuse easily unless it's a delayed fuse.

Still sounds like a wiring issue of some sort though if it only started happening when you re-wired it and there is such a direct relationship between the VR synching with the Hall disconnected and not synching when the Hall is connected. Unless during the changes you have managed to get the cam pulse at exactly the same point as the missing tooth and it has an effect that we haven't experienced yet.....

Schrödinger's cat - so which one am I ???