Meagsquirt EFI on a 2.2T - Real time data view with iPad

[1972_911t]

Hi Gary

For the idle control do you set a base idle with the bleed screws when the engine is warm or does the ecu control it all via the idle valve?

Steve

[Gary71]

Hi Steve,

As I'm still on carbs at the moment it still done via the bleeds and the throttle stops. I have long term plans to fit idle control, but for now I've gone with Tim's suggestion and keep it simple. If I need extra air during warm up I will tweek the hand throttle to remind me of the good old times

I just got back in (feeling a bit sick from running the car in the garage *cough*) and I find that Tim's loan carb already had it's bleed screws wound all the way in... so not much hope of getting the idle below 1200 as it's already off the throttle stop.

I've downloaded another couple of runs to analyse, but my first thoughts are:

1. Fixed 10 deg advance. Starts and idles consistently, but not suprisingly won't rev.
2. Without the MAP connected, so only RPM based advance: Idle stable, but slow to pick up.
3. Damped (with an in-line fuel filter) MAP. Good response, but wandering idle speed.

This is pointing me towards getting all six bleeds connected up to the MAP and hoping this smooths the idle MAP and hence stops altering the advance all the time.

The misfire is not apparent at idle so I need to look into that more. I've connected up the strobe tonight and the actual timing at idle matches the fixed setting perfectly, so the VR sensor is in the right place relative to TDC.

Next step (for tomorrow) is to drive it without the MAP connected (so effectively backing off the advance by 10 deg) and see how she goes.

[Gary71]

..and to follow up on a previous question, this is how the air filter will go together:

The original throttle body:



Ali adaptors, to be pushed down over ports:



Original Zenith trumpets (they will be cleaned!! ) fit on top



Sandwiching the zenith steel airbox, which will need a new base as the triumph bodies sit at about 5 degrees.



This may also get cleaned!

[impmad2000]

Tim's loan carb already had it's bleed screws wound all the way in....

Oh well..... I did warn you !
Can't you just accept the higher tickover and adjust the tickover threshold up to suit ?
When ticking over, can you see the MAP pulsing on the runtime display ?
Tim

[Gary71]

Latest update,

Tim hopefully you've received the *.msq file I sent you?

MAP jumps all over (between 35and 55kpa) when idling even when damped on fixed advance, so I guess I need to get all six ports connected as discussed.

I drove to work this morning just to get it properly warmed up and it then idled around 2200. I haven't adjusted the carbs so if the advance was correct then the idle should be around 1050ish as it was on the dizzy. The only thing causing this could be over advance, so I wound a -7 offset into the settings and this brought the idle right down to where it was before. Now it's just on the edge of the steep advance curve.

Looking at the log I am getting some spikes in the advance, so I'm guessing there is some noise and this maybe related to the occasional stutter as I'm driving. I discussed this with our electrical boys here and in turns out my grounding of the shielded wiring both ends was misjudged... turns it into an aerial or something... *starts getting lost in electrical engineering minefield*

So I'll correct that tonight and see if things clean up.

This is a plot showing the spikes in the constant throttle section. It must be noise as there is no other influence to make this sudden shift in advance.



This is idling at some lights, raise revs to 2500 due to lack of confidence in it not just dying on me, nail it in first, into second then chill back and cruise in third.

I've also borrowed a scope from work to take a look at the VR sensor outputs, just to see if they are as expected and not a noisy mess.

I'll log on the drive home tonight to see if backing off the advance has improved the running, it certainly accelerated cleaner just blipping

[impmad2000]

Yep ! grounding should be from one end only, preferably at MS end. You'll get the hang of it.
I assume the tach trigger is via the EDIS ? this is at least quite clean.
I have sent email, but basically, if you'r idleing around the steep part of the curve, then the timing will be swinging all over the place. I'd try adjusting the RPM bin spacing of the table, 900, 1200, 1600 and then into the rest of the table.
Entertaining though isn't it Don't forget to watch the road
Tim

[Gary71]

Hi Tim, Thanks for the mail,

I'll reset the bins and try that on the way home together with the revised overall timing

Tach trigger to the MS ECU is from EDIS, the actual tach on the dash is still running off the points. From what I read getting the tach to work can be a challenge, so I'll leave the original ignition set up on for now in parallel and spend the time on getting the car running!

With regards to earthing I'll cut the PIP and SAW shield earths at the EDIS unit. What about the VR sensor? Currently this shield is earthed at the EDIS and not at the sensor, guessing this is OK?

[impmad2000]

VR sensor would be grounded at the EDIS only.
Tim

[Ian Donkin]

Keep it coming guys - there's an interested Yorkshireman here.

I suppose carbs that didn't leak so much air might help?

[Gary71]

carbs that didn't leak so much air might help

Just a bit

I loaded in the revised map half way home to keep the 10 deg advance around 'idle' for longer as Tim suggested, this worked well. I was supposed to do it before I left, but it didn't seem to protest at having a complete new .msq file burnt in whilst running! I have yet more datalogs that I need to review tomorrow.

It still has a definite miss, no question, but is now running as well as any 5 cylinder 911 would

I really need to sort the miss before I do anything else, or I'll just end up tuning the ignition map to make a 5 cylinder car on leaky carbs run well

I hooked up the scope to the VR sensor tonight and the signal looks fine, so that's nice to see. Progress is good. I'm happy so far

[impmad2000]

It still has a definite miss, no question,

A permanent miss, always 5, or or just the odd miss ?
Obvious I'm sure but, Switch the leads round, revert to dizzy
Do you know which one ? Does it switch to the other of the pair if you switch the pairs over at the coil pack.
I have an EDIS and coil pack if you want to try an alternate set.
Cheers
Tim

[Gary71]

Difficult one Tim, I always struggle with misfire identification on anything with more than 4 cylinders. I may pull the plugs to see if this gives a clue.

I do some logical playing and see how it goes. I have another EDIS6 coil pack, so I can try that if all else fails.

[Gary71]

Just analysed the log from just after I reset the idle bins and the advance spiking has gone The log from just before was as the previous one, occasional spiking up and down for no reason



No explanation really, just more confusion to throw in to the pot!

Delaying the onset of the steep advance helped response no end and, misfire allowing, it would even take reasonable throttle in a high gear from low speed.

[impmad2000]

Can you overlay the MAP Gary ?
Cheers
Tim

[Gary71]

Can do, but it's a straight line

It's pulling way too much vacuum from the enrichment ports (below butterflies) than the original ported vac advance take off. With the spark table I have at present it ends up advanced by 10 degrees. So currently it breathing nice fresh air from my engine bay.

Project for this lunchtime is to revise the table so the advance is less extreme. Then with all six ports hooked up, a damper, and no misfire I might start mapping it properly

I'll post one of the old logs so you can see the MAP wobbling.