MGExp

MG Midget Forum

EFI conversion running lean

Moss Motors
AutoShrine Sponsor
AutoShrine Sponsor
AutoShrine Sponsor
AutoShrine Sponsor

nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
My first post here after browsing for a few months. I have a mk3 Sprite with a blueprinted 1275 set up with a Specialist Components EFI conversion kit. After a long road of many twists and turns, I'm nearly at the finish line of actually driving it on the road, but I'm stumbling on the fuel mapping. Wondering if anyone here is running a similar setup, and if so what did you need to do to get it running smoothly? It's theoretically a plug-n-play kit, but as I've learned, nothing about this car is plug and play.spinning smiley sticking its tongue out

The closest map I could find was the "fast road 1275" but it still runs super-lean and won't hold an idle. Since there's no choke, I tweaked the fuel ratios below 60 degrees C coolant temp to help with the cold startup. That works fine (though super rich) for startup, but once the engine heats up out of that range it starts to struggle again. Tried to work with the guys at SC over email, but that's been sketchy at best. Fuel pressure's good (3.5 bar), timing seems fine (at least the sensor is), TPS is fully functional. Curious if anyone here has any suggestions or experience to share.

. Hide banner ads & support this website by becoming a > Supporting Member <
trevorwj Avatar
trevorwj Trevor Jessie
Louisville, KY, USA   USA
Is this a closed loop system? If so, where is your wideband O2 sensor located?



Midget 50th - A Spridget Event
It was a Blast!
http://sprite-midgetclub.org/

nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
Not closed loop yet. Was considering getting a wideband for precisely this reason. I'm guessing common wisdom is that's a good idea? I'm just a little flabbergasted that the program theoretically made for my engine doesn't work with my engine. Also need to get it running stably to bring it back to my exhaust guy and get a bung installed. Only inputs to the ECU right now are TPS, crank position, coolant temp, inlet air temp.

Another factor: I'm currently venting the PCV out to atmosphere (using a tube going to the vicinity of the air filter, no valve). Is common wisdom to send this into the air filter or to a port in the intake manifold (there's a plugged 5/8" port that seems made for a MAP sensor)? Regardless, would this have a drastic impact on performance?

Full disclosure: This engine is still in the break-in phase. Only just got through initial break-in with heating it up, torquing head bolts and gapping the valves. So I don't even really know how it's supposed to run...

. Hide banner ads & support this website by becoming a > Supporting Member <
trevorwj Avatar
trevorwj Trevor Jessie
Louisville, KY, USA   USA
What brand ecu are you using?



Midget 50th - A Spridget Event
It was a Blast!
http://sprite-midgetclub.org/

nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
Specialist Components "Typhoon". Several years old - doesn't look like the current one. Inherited this project after most parts were purchased, but very few installed.

trevorwj Avatar
trevorwj Trevor Jessie
Louisville, KY, USA   USA
So what type of fuel tables and modifiers do you have available?



Midget 50th - A Spridget Event
It was a Blast!
http://sprite-midgetclub.org/

Yankeedriver Avatar
Yankeedriver Platinum AdvertiserAdvertiser Joel Young
Albuquerque, New Mexico, USA   USA
In reply to # 3553776 by nkaufman79 Not closed loop yet. Was considering getting a wideband for precisely this reason. I'm guessing common wisdom is that's a good idea? I'm just a little flabbergasted that the program theoretically made for my engine doesn't work with my engine. Also need to get it running stably to bring it back to my exhaust guy and get a bung installed. Only inputs to the ECU right now are TPS, crank position, coolant temp, inlet air temp.

Another factor: I'm currently venting the PCV out to atmosphere (using a tube going to the vicinity of the air filter, no valve). Is common wisdom to send this into the air filter or to a port in the intake manifold (there's a plugged 5/8" port that seems made for a MAP sensor)? Regardless, would this have a drastic impact on performance?

Full disclosure: This engine is still in the break-in phase. Only just got through initial break-in with heating it up, torquing head bolts and gapping the valves. So I don't even really know how it's supposed to run...

Nathan,

First - I'm only passingly familiar with the SC system, but can't imagine a TBI EFI system running properly without a MAP sensor; the sensors you list can't give complete data to run the car unless atmospheric pressure is presumed constant. Could be their (older) system is unique and I'm wrong but you might ask them that specific question. If it's designed with MAP input it's a wonder it's running off idle at all, and you've found the trouble. Stop fiddling until you correct this, as your maps could've been serviceable.

Second - sorry if this is obvious but you should not run your car with a non-metered draw through the crankcase. Depending on what kind of valve cover cap you have, that would produce dramatically lean running. So, yes you're right metered draw through the crankcase is the standard method--but the cap and PCV valve need to be matched.

However, if I understand correctly, you're venting crankcase vapors (you used the acronym for positive crankcase ventilation, which means subject to vacuum, so confused me a little) into the atmosphere with no draw from the intake, metered or otherwise. If so, that won't impact mixture--unless there's an open port or other leak in your intake tract. My guess is the plugged port in the manifold is for a PCV valve. Here are photos of my EFI manifold, set up to feed vacuum to both the OEM PCV valve and standard GM MAP sensor (those these come in differing ranges/sensitivities - make sure you ask SC what bar rating their MAP sensor needs to be). If you need/like the setup, I have all the part numbers--super cheap brand new.

Finally, the break-in period won't affect mixture, absent a poorly sealing valve, but running lean can burn a valve or in extreme cases hole a piston. I'd stop running it until at minimum you eliminate the MAP sensor question.

Joel



Edited 1 time(s). Last edit at 2017-07-17 08:18 PM by Yankeedriver.


Member Services:
Innovative DIY performance and reliability upgrades.

Attachments:
IAC and fuel psi regulator.jpg    53.2 KB
IAC and fuel psi regulator.jpg

MAP sensor hose(2).jpg    55.3 KB
Sign In or Register to view this photo
MAP sensor location final.jpg    34.7 KB
Sign In or Register to view this photo
reducer.JPG    8.2 KB
Sign In or Register to view this photo
1st77mgb Avatar
1st77mgb Scott Mervine
Cleveland, OH, USA   USA
I am confused, you do not have an oxygen sensor in this set up? How is the ecu going to determine fuel trim?

Yankeedriver Avatar
Yankeedriver Platinum AdvertiserAdvertiser Joel Young
Albuquerque, New Mexico, USA   USA
In reply to # 3553889 by 1st77mgb I am confused, you do not have an oxygen sensor in this set up? How is the ecu going to determine fuel trim?

Yes - forgot that obvious omission - notwithstanding Trevor's question about it. Long day...

Nathan - some systems will work with a wide-band providing input to run the car as well as inform a gauge, some won't. Another question for SC.

Joel


Member Services:
Innovative DIY performance and reliability upgrades.
nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
Ok, clearly I needed to post here earlier. Been flying a bit blind til now, but here's my setup as it exists today (image below). Forgive me as all of this is new to me. Engineer by training, but haven't gone very deep with the practical application. Per my communication with SC, they claim not to need a MAP without forced induction and specifically claim that TPS-based mapping provides a crisper feel. I believe the theory in not shipping with an O2 sensor is that the out of the box maps are already tuned for the 1275 setup, and a wideband is only required for further tweaking. To this point, I've just been trying to get it up and running, with the intent of leaving the fine-tuning for the future.

My ECU is setup on two primary maps - one for fuel injector pulse width (in MS) and the other for spark advance (in deg), both mapping inputs into a TPS vs. RPM table. My understanding is going to a MAP-based mapping would actually be replacing the TPS-based mapping. The ECU configuration has 3 different load input types - Throttle Position (AN), Manifold Pressure (SD), or MAP compensated TPS. I'm currently set to use Throttle Position (AN). As for other modifiers, I'm currently using the coolant temp multiplier, but have a variety of other options as well that I haven't explored. There is also the aforementioned option of adding a wideband O2 sensor for closed loop control.

As for Joel's other questions, I have vented valve cover cap (stock?) and the canister on the timing cover which I was referring to as the PCV (currently has a red hose bringing it up near the air filter). No intentional metering, but I believe the cap and the vent are matched (not sure how to confirm this). Would definitely be interested in bettering the system, but I was under the impression that the PCV valve wasn't even included until the mk4 Sprite (mk3 Midget). Again, forgive the ignorance. And thanks for the input!


Attachments:
EFI_intake.jpg    49 KB
EFI_intake.jpg

Yankeedriver Avatar
Yankeedriver Platinum AdvertiserAdvertiser Joel Young
Albuquerque, New Mexico, USA   USA
In reply to # 3553954 by nkaufman79 Ok, clearly I needed to post here earlier. Been flying a bit blind til now, but here's my setup as it exists today (image below). Forgive me as all of this is new to me. Engineer by training, but haven't gone very deep with the practical application. Per my communication with SC, they claim not to need a MAP without forced induction and specifically claim that TPS-based mapping provides a crisper feel. I believe the theory in not shipping with an O2 sensor is that the out of the box maps are already tuned for the 1275 setup, and a wideband is only required for further tweaking. To this point, I've just been trying to get it up and running, with the intent of leaving the fine-tuning for the future.

My ECU is setup on two primary maps - one for fuel injector pulse width (in MS) and the other for spark advance (in deg), both mapping inputs into a TPS vs. RPM table. My understanding is going to a MAP-based mapping would actually be replacing the TPS-based mapping. The ECU configuration has 3 different load input types - Throttle Position (AN), Manifold Pressure (SD), or MAP compensated TPS. I'm currently set to use Throttle Position (AN). As for other modifiers, I'm currently using the coolant temp multiplier, but have a variety of other options as well that I haven't explored. There is also the aforementioned option of adding a wideband O2 sensor for closed loop control.

As for Joel's other questions, I have vented valve cover cap (stock?) and the canister on the timing cover which I was referring to as the PCV (currently has a red hose bringing it up near the air filter). No intentional metering, but I believe the cap and the vent are matched (not sure how to confirm this). Would definitely be interested in bettering the system, but I was under the impression that the PCV valve wasn't even included until the mk4 Sprite (mk3 Midget). Again, forgive the ignorance. And thanks for the input!

Thanks for the additional info. Two things and I'll let others wade in

First - if that's the water temp sender to inform the ECU in the thermostat housing, that's got to move to the head. The ECU must have 'real time' data as the engine warms up to adjust the mixture--though it isn't clear to me whether this early SC setup enabled an idle air control (IAC) valve. Even if not, seems like you'd want the ECU not to be reacting to the sudden change in temp when the thermo opened--but just IMO. Again, I'm not familiar with this early SC system.

Second - sorry if I confused you. The 'matching' I was referring to is between in input end of the crankcase ventilation system--the cap--and the output end--the PCV valve, which of course is fed vacuum from the intake manifold. The tank on the timing cover just contains some steel mesh or something similar to catch suspended oil so it'll drip back down into the oil pan. But, again, none of that would affect your mixture unless somehow connected to the intake tract.

Oh - and finally, please forgive my ignorance if I didn't understand the different theories behind the respective systems vis-a-vis MAP sensor function.

Joel



Edited 1 time(s). Last edit at 2017-07-18 08:50 AM by Yankeedriver.


Member Services:
Innovative DIY performance and reliability upgrades.
trevorwj Avatar
trevorwj Trevor Jessie
Louisville, KY, USA   USA
If you are increasing the pulse width at a given rpm range and throttle position but the car is staying lean ... then either several things come to mind
1. the injector is not capable of the requested pulse due to inadequate voltage
2. the fuel pressure is not high enough (is the regulator on your return line? it appears to be in the picture, but just double checking)
3. There is a intake air leak.

Did you calibrate your TPS? Does the software high light the active cell in your table while running?



Midget 50th - A Spridget Event
It was a Blast!
http://sprite-midgetclub.org/

nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
My head hurts. First Joel:
1. Understood that the water temp sender would ideally be on the head (where the stock temp gauge sender is), but that should only be an issue on startup. Once the engine is warm, the coolant is flowing and discrepancy between the two should be negligible, no? Also, this aspect has not changed in the modern SC system.
2. There is no IAC, only a manual idle set screw that holds the TB open a set amount. The ECU has capabilities to control idle, but this kit did not come with such a valve.
3. Crankcase venting is apparently a total mystery to me. I understand that the cap is the inlet, but I thought the timing cover canister was the outlet. Is there another vent somewhere that I don't know about? I really need to find a flow diagram for how this is supposed to work.

Trevor:
1. I don't know how the voltage could be off, though I also don't know how to confirm that it's good. I had thought that possibly the injectors were somehow mismatched to the system, but I can't find any programming way to configure the injectors
2. Fuel pressure regulator is on the return line and is set to 3.5 bar which is spec for the system. Reference is atmosphere rather than vacuum, but I'm not sure if that matters.
3. I thought that as well. The plug on the manifold port is super easy to turn on it's own making it seem like a prime candidate for leakage. I put a lock nut on top to lock in the threads, but I'm not sure how to confirm whether there's a leak. If I don't end up installing a PCV valve, then I'll get a proper plug. Curious how others have configured this manifold.

Seems link crankcase venting may be my biggest culprit, though I'm not totally clear on the solution. Also would greatly help to have an O2 sensor to give some idea of the actual mixture being burned.

trevorwj Avatar
trevorwj Trevor Jessie
Louisville, KY, USA   USA
You can install a clamp on O2 sensor bung then later have a shop weld in a permanent bung.
I assumed you at least had a narrow band sensor telling you it was lean.

I think I'd spend the time/money on installing a wide band sensor and maybe an IAC valve.



Midget 50th - A Spridget Event
It was a Blast!
http://sprite-midgetclub.org/

nkaufman79 Nathan Kaufman
Reading, MA, USA   USA
Sorry, I deduced it was lean based on startup backfires and inability to idle under 1500 RPM. I had started this post to tease out whether anyone had the SC setup and what they did to make it work (since it is advertised to work out of the box). I suppose I shouldn't be surprised that it isn't a common system given the kit cost - I wouldn't have gone with it had it not come (in a box) with the car. You guys have at least bettered my understanding of what all the components are and what to look for, so thanks. I'll definitely be purchasing a wideband O2 sensor. One other question: are they pretty universal or should I go with the one offered by SC to ensure compatibility? Not sure I really care about having a gauge as I can just plug the laptop in for tuning, and don't need to clutter things once it's setup.

. Hide banner ads & support this website by becoming a > Supporting Member <

To add your reply, or post your own questions

Members Sign In   or   Create an Account

Registration is FREE and takes less than a minute!


Having trouble posting or changing forum settings?
Read the Forum Help (FAQ) or contact the webmaster