EFI Learning Curve

[tmsmini]

I am hitting a steep spot on my EFI learning curve.

I have the engine running but it is not running the way I would like it to, as it is almost not drivable.
There seem to be some variables not under control. I certainly expected a good period of adapting and learning. After all we are taking  an engine with variable specifications and adapting a system to it. There are a number of variables, all of which impact running and interact with each other, which the FI vendor cannot control.

It seems there are a number of people who have used other systems for other cars and they may be able to suggest something simple I am missing.

This is an MPi engine rebuilt by Seven Enterprises running a Specialist Components injection kit and adapted to a MK I Traveller. Part of the problem is the long distance support issues with a vendor in the UK. But Simon has been helpful and offered to connect remotely to analyze the problem.

I did originally have it set up on a dyno, but that dyno is also 100 miles away. There may have been an issue with the fuel pump and it was swapped to a lower output pump, so that requires some changes to what was configured on the dyno session. There was some backfire with the original configuration on cold starts, so adjustments needed to be made anyway. I am also using an Innovate meter with a tailpipe sensor which may not give the most consistent results. SC provided an alternate MAP which should have been close to the specifications.

It starts and idles just fine after some throttle stop and pulse width adjustments.

Not sure what I expect from anyone in long distance troubleshooting, but maybe someone will see something I am overlooking. This is new to me.

This is what I sent Simon at SC.
I went back to the MAP you provided to start over again. I tried a few more adjustments to raise the idle with the idle stop.
I was able to get a consistent warm idle at  around 1050 and adjusted the map to get a reasonable AFR.

With these settings the car would die at almost all stop signs.  I turned the closed loop control off as it appeared to be fighting the settings, but the problem still continued.

I then increased the idle to around 1150 and adjusted the AFR. I was able to drive the car with no dying, but it was fully warmed up with no air or coolant temp compensation interaction.
Although the idle was set high, it seemed to have stopped the dying. Unfortunately, later in the day, I went out to run some errands and it died at a stop light again. It has no trouble restarting.

I realize I need to get it on a rolling road to consistently test at a specific load, but it seems like there is either something wrong/malfunctioning or there is a major setting that is off.

I am concerned that the way the breather is connected could be causing the problem. There is a smaller connector on the breather pipe that is connected to the manifold on a Rover MPi setup. The end of the pipe has a fixed orifice or port in it. I am not sure how that works with the charcoal canister/purge valve on the Rover configuration, but this car does not have the purge valve. This is currently connected to the manifold port as shown in this photo:
http://www.tmsmini.com/cooper/images/sc_hose_routing.JPG

Could this be the cause of the problem?
Update:
It was easy enough to disconnect the breathers from the inlet manifold and although the mixture became richer as expected, the dying at stop signs did not go away.

I need some help with a logical troubleshooting process.

[94touring]

Do you have maps you can post?  I can try to decifer and compare to maps and correction maps on my tunable system, albiet non mini related. 

[tmsmini]

let's see if this works:
Updated image same as below

there is no idle air control valve and this is a TPS only system, no manifold absolute pressure sensor either

[94touring]

That's the stock base map?

[tmsmini]

That is the MAP provided by Specialist Components.
Although I may have increased the pulse width at 0/1000 to get it to idle.
I may need to check that.
here is the unmodified MAP:

[94touring]

What afr's are you seeing from a stop to your max rpm?  You're still running a dizzy and we're just adjusting fuel right?

[tmsmini]

No this is a full computer.
It looks the the AFR jumps past 20 at the stops.
With the tail pipe sniffer that could be an artifact.
I will have to look at the date to be sure.
Unfortunately I have an Innovate LM-2, perhaps npt the most reliable meter.
I was tempted to try to get a vacuum gauge on it tonight, but it was dark by the time I got home.
Terry
Ignition map

[94touring]

The stops?  You should be 14.5 at cruise and wot slightly rich on na typically provides best power, something more like 13:1.  On decel you'll see 20-22.  Ignition at first glance looks good. 

[94touring]

Idle you'll be 12:1 to 13:1, whatever it likes. 

[tmsmini]

In one run where it dies at stop signs, the AFR seems to spike at over 36, close to 40, but that could be an issue with the Innovate and the tailpipe. But as a relative reading, it is going very lean at that point no matter what the Innovate says. I checked for a manifold leak, but could find nothing.

[94touring]

Aside from idle, what's it doing? 

[tmsmini]

It does well with the small MAP adjustments I have made. It needs fine tuning, but cruise is about 14.X and it goes rich as you step on it.
I probably need to get the sensor back in the middle branch of the LCB to remove the tailpipe sensor variation. The wiring harness is currently configured for the MPi style narrow band sensor.

[tmsmini]

If the values at 0 TPS are not out of scale then I do not believe it is the MAP that is the issue. It really seems like an inlet leak which is why I capped the breathers in the last test.

Using the Innovate as a relative measure and not an exact instrument, the AFR values are within scale. Ideally it would get on a dyno so that load can be held at certain values and the AFR adjusted. At this point dyno time might be wasted as there appears to be either a component malfunction or installation issue.

[94touring]

Time to spray around with some starter fluid perhaps. 

[tmsmini]

I have done that with no result. That is why I thought it was an internal system issue.
I had originally planned to sort this all out on a engine test stand before installing in a car. But with the SC system we went ahead and installed it anyway.
I have a Microsquirt based system that I plan to use as well as it will work with all the Rover FI components, but takes more adapting and fabrication of wiring harnesses, etc.

I originally purchased a used SC kit, but when I found they no longer supplied parts for the older system(2012), I foolishly bought the newer throttle body.

Interestingly enough the original machined SC throttle body had a fitting for using the "choke" cable as as fast idle adjustment. The newer cast bodies do not.
I am committed to making it work though. It does idle wonderfully and the aux fan is ECU controlled.
I do need to get to know the system better.
Terry

[94touring]

If it makes you feel any better my rx7 has a fully programmable system, I use a laptop and all that to change anything and everything.  But my idle is inconsistent.  It jumps all over and hunts.  I did have it adjusted pretty good but now with the weather change it's being screwy again after it warms up.  I'm about where you are in I can't figure out if it's something mechanical or something map related.  It's an old car so it's a guessing game. 

[tmsmini]

It is like the SU manuals start out, make sure all other components are working properly.
And then 90% of carburetor problems are electrical.

I did find this on air bleed screws on a Ducati forum of all places, but it explains the use of air bleed without IACV.

There is an interesting difference between systems with IACs (Idle Air Control) and those without.
The IAC system is a single air bypassscrew operated by a stepper motor. This 'T's into what amounts to a balance tube between the two throttle bodies.
In this type of system the amount of air at idle is always the same whether the bypass screws are open or shut. On the MTS1200 the idle air/fuel ratio is adjusted by the O2 sensors.
The exception is when the IAC is either open to max or closed to min, at which point the ECU can't control engine speed and it will display a check engine light. This can occur when the throttle stop is either open too much or closed too much for the IAC to control the idle speed. I kind of doubt that you could close the throttles enough that the IAC (which is a fairly large port) couldn't suupply enough air to idle, but I haven't tried that out.

In a system without an IAC, the air-bypass screws adjust the amount of air at idle. Without an O2 sensor, this also has a large effect on AFR for each cylinder, but it also affects the idle speed.

Ultimately, the correct way to set one up is to close the air bypass screws and disconnect the IAC hoses and synch the throttles by vacuum (I don't recommend disconnecting the IAC hoses because they can tear and that causes more problems.) The second problem is that you have to hold the throttle open to get it to idle because you have shut off the air needed to idle.
The second problem is that you have made the idle mixture rich by cutting off the air and now you have hold the throttle off idle to synch the throttle blades.
With the MTS1200 the sensors will adjust the AFR to 14.6:1 so that's not an issue.
After synchronizing the throttle blades, both of the air bypass screws should be opened to the same number of turns. And, of course the IAC should be reconnected.
At this point, in a system without O2 sensors in closed loop, the air bypass screws have a large effect on the AFR of each cylinder. At high vacuum, a small difference in vacuum makes a large difference in air-flow. A large difference of air-flow makes a large difference in AFR.
Therefore, the bypass screws are the devices used to equalize the cyl-to-cyl AFR at idle.

With the Mitsubishi closed-loop systems that difference can be seen by looking at the live Adaptive Fuel Value. This shows how much the system has to skew to get to 14.6:1.
A small tweak of the air bypass screws will get the AFVs closer to one another. These AFVs affect the closed-loop operation in the rest of the closed-loop operating range.

Doug

Somewhere else in the thread they talk about managing idle stalls, but I am not sure it applies.

[tmsmini]

Doing a remote session with Simon from SC on Monday.
Maybe he will find something.
It starts fine except on the coldest day so far around here and that is only 39, so really pretty warm.
And idles fine once started.
I am going to sort out the breather hoses and do another leak check before Monday.
Terry

[tmsmini]

Update of sorts...
Last week Simon of SC connected remotely made a few adjustments and had me bump the idle.
Definitely better, another updated map and I felt comfortable driving in traffic.
Wife made it to work today with outside temps in the upper 30s with no issues.

A few refinements to be made as it now idles at about 1100+ when fully warm. I would like to get that down, but with limited idle air control that may not be possible
Terry

[94touring]

Good to hear.  Just some fine tuning required sounds like.

[Jims5543]

My only EFI tuning experience was also with my RX7 I had a Motec M4 and purchased the lambda for it and data logging which were both HUGE when tuning in new changes to the car.

I did an initial dyno tune, then I took a tuner guy in my passenger seat with a laptop on his lap, he had a LM2 sniffer in the tail pipe in addition to my lambda (4 wire heated) and to be honest, we did more substantial work on the road driving around than on the dyno.

If you know of someone that can ride with you and make changes on the fly (not sure if your setup does that) that might be helpful.

I too had my idle set at 1100 rpm, it sounded like my engine was always racing, but, it was the trade off for having a high strung car. I could never go lower or I always stalled at lights or stop signs too. I just had to learn to love it.

I was toying with going Megasquirt when I build my next insane car, but, I know Motec so well, I think I will just spend the money and get that again.

Good luck I hope you get it sorted out. Once you get it dialed in you will love it.

Also, I once took my laptop on a road trip, had it on the seat next to me and kept tweeking my fuel while cruising trying to lean it out a little but not too much. I eventually had that beast tuned to get 21 MPG on the highway at 80 MPH.  Considering it had over 500 HP and 4:10 gears in the back cruising at 4K RPM's I was very satisfied with my work.   ;D

[94touring]

Yeah I've got my cruise dialed in for 20+ mpg too.   4.gif.   I really need to work on my idle though. 

[tmsmini]

I need to spend more time with it and try to understand what the ECU is trying to do. This is not a high performance engine by any means.
My wife had a couple of issues with it while I was gone last week, so I am driving it this week. I did have a slight stumble getting on the freeway, but no data connection to see what was going on.

I believe I can adapt the Innovate LM2 output to be recorded in the SC SXTune software and I may be able to use 2 O₂ sensors for inners and outers and record both of those.

Still not a lot of guidance on the air bleed adjustment though. I could learn to live with the 100 idle, but 1150-1200 seems a bit much. I am not sure what aspect/sensor is pushing it higher after it is warmed up.
Terry

[MtyMous]

Regarding the O2 sensor, I've got mine welded into the collector right after all 3 branches come together. Would that not be a more ideal location for data collection? It doesn't  give you any cylinder bias to tell which 2 cylinders are doing what, but without direct port injection, does that really matter?

Just thinking if my current setup needs to be changed as well.

[tmsmini]

With the SC electronic carb setup, no perhaps it does not really make a lot of difference. It is just more knowledge and perhaps you can choose to bias the settings more favorably to both inners and outers. Theoretically an electronic carb should be no worse that a mechanical carb, once setup properly. I am approaching that goal, but no there yet.

With Megasquirt, it is possible to make injection timing adjustments to modify. It may be possible with the SC setup as well, but would probably require more information - data sensors.
Terry