Megasquirt EFI conversion
- Thread starter dpstark2
- Start date
dpstark2
Well-Known Member
ONE STEP CLOSER!
As of today, I have a functioning crank position sensor. I drove around a good 20 minutes and didn't lose sync once! Compare that with the previous setup which lost sync about once every 30 seconds or so... Finally! Here's a picture of the new sensor and the wonderful mount made my Meep-meep.
The next step is to plumb the electric fuel pump in and run the carb off the regulated high pressure line. That will be a few weeks, but hey- I'm making progress for once...
As of today, I have a functioning crank position sensor. I drove around a good 20 minutes and didn't lose sync once! Compare that with the previous setup which lost sync about once every 30 seconds or so... Finally! Here's a picture of the new sensor and the wonderful mount made my Meep-meep.
The next step is to plumb the electric fuel pump in and run the carb off the regulated high pressure line. That will be a few weeks, but hey- I'm making progress for once...
dpstark2
Well-Known Member
I haven't plumbed the electric one yet. It is in the tank, but it's not hooked up. That's the next task. I will be happy to see the mechanical one go and make more room for the alternator, though...
dpstark2
Well-Known Member
Woo! Finally an update!
I spent the weekend (with help) plumbing up the modified gas tank Meep-meep helped me make. Well, he mostly made it. I helped. Anyway- here's a picture of the modified tank with the added in-tank pump and sump:
I ran a second factory-type hardline along the fuel rail, spliced in a filter, and ran braided AN lines up to a regulator. The regulator goes off to the carburetor for now.
Running power to the pump was kind of a pain. I had a lot of other stuff to do while the interior was apart to run the lines, so some sound deadener when it, among other things. But--- key on, and it ran. And leaked. Yes, I forgot to tighten a few fittings and get the AN stuff to seat properly. But after that, I set the pressure at 5 psi for now and it runs. Hot starts are better, fueling seems more consistent, and the hum of the pump is inaudible with the car running. I am also tied to the Megasquirt unit now, because it turns the pump on.
Assuming this set up works, it's only a large check for parts and a lot of work to get MPFI going.
I spent the weekend (with help) plumbing up the modified gas tank Meep-meep helped me make. Well, he mostly made it. I helped. Anyway- here's a picture of the modified tank with the added in-tank pump and sump:
I ran a second factory-type hardline along the fuel rail, spliced in a filter, and ran braided AN lines up to a regulator. The regulator goes off to the carburetor for now.
Running power to the pump was kind of a pain. I had a lot of other stuff to do while the interior was apart to run the lines, so some sound deadener when it, among other things. But--- key on, and it ran. And leaked. Yes, I forgot to tighten a few fittings and get the AN stuff to seat properly. But after that, I set the pressure at 5 psi for now and it runs. Hot starts are better, fueling seems more consistent, and the hum of the pump is inaudible with the car running. I am also tied to the Megasquirt unit now, because it turns the pump on.
Assuming this set up works, it's only a large check for parts and a lot of work to get MPFI going.
Nice work! Can't wait to see it in person. Are you also still running the mechanical fuel pump or is that now gone? The tank I bought has an in-tank pump and I have been wondering about that.
Gpuller
Well-Known Member
Have been thinking about the LS EFI set up. Found a place somewhat close that tunes LS dirt cars, need to call them and see if they would be interested in doing a Mopar.
Got some free 5.3 injectors and computer laying at work.
Got some free 5.3 injectors and computer laying at work.
dpstark2
Well-Known Member
If you can make the LS stuff work - go for it! I know we're all loyal, but from what I've seen, it's good stuff. I would imagine it's possible, but I think LS motors use a specific crank trigger that might be hard to fab up. Of course, I'm sure it can be done with enough skill and money.
So far everything is working well. A couple things I've learned- cheap liquid filled mechanical fuel gauges aren't accurate under the hood. For some reason the heat really affects them. It's mostly fine because I set the pressure with the motor off and cool... but it's worth noting. Also, braided stainless hose is abrasive. I will be going back and covering some of it to prevent paint and plug wire damage. And, if you order this type of stuff, get extra olives for your fittings (that's the little compression ring in the fitting). They are one use only. Of course...
So far everything is working well. A couple things I've learned- cheap liquid filled mechanical fuel gauges aren't accurate under the hood. For some reason the heat really affects them. It's mostly fine because I set the pressure with the motor off and cool... but it's worth noting. Also, braided stainless hose is abrasive. I will be going back and covering some of it to prevent paint and plug wire damage. And, if you order this type of stuff, get extra olives for your fittings (that's the little compression ring in the fitting). They are one use only. Of course...
Gpuller
Well-Known Member
There is a guy over on Moparts that is putting together an LS based EFI for his Challenger. Yes, the LS uses a special crank signal but the tone ring is available from GM. He mounted it on the pulleys just like you did.
A GM computer can be bought used for $40, he also shows the pin outs of the GM ECU and what to change for a Mopar application.
Biggest expense is a throttle body and manifold, if you didn't modify one for injectors.
http://board.moparts.org/ubbthreads/showflat.php?Cat=0&Board=Racer&Number=7562532&fpart=&PHPSESSID=
I do have some problem getting my head around using GM stuff on a Mopar but the thought of a cheap EFI intrigues me.
A GM computer can be bought used for $40, he also shows the pin outs of the GM ECU and what to change for a Mopar application.
Biggest expense is a throttle body and manifold, if you didn't modify one for injectors.
http://board.moparts.org/ubbthreads/showflat.php?Cat=0&Board=Racer&Number=7562532&fpart=&PHPSESSID=
I do have some problem getting my head around using GM stuff on a Mopar but the thought of a cheap EFI intrigues me.
dpstark2
Well-Known Member
A year and a half later:
I have EFI up and running (more or less)! As these things go, I ended up shorter on funds than expected and picked up a used 4bbl TBI setup. It came complete with fuel rails and injectors. I modified the rails to get rid of some unnecessary stuff and clean it up. Otherwise, it bolted right onto the Edelbrock RPM manifold.
The fuel lines were rerouted from the original carb application (with the modified in-tank pump) and the pressure was set to 43psi. A friend and I did some testing to make sure the TPS and injectors were working, which they were.
And then crank, crank, boom. It fired up and ran, then died. One of the big challenges of tuning EFI is the cold start, and it was 40-some degrees in my garage. Eventually we tracked down a non-working IAC as the culprit. After that, some coaxing and tuning got it up to temperature and idling.
The IAC still doesn't work, so I need to track that down. It could be wired wrong, configured wrong, or just plain not wired up on the PC board. The O2 sensor is also showing noise that will make tuning impossible. No idea what that's about yet, but noise has been my biggest issue, so no surprise there. And there's a lot of clean-up to come under the hood before it drives anywhere.
Here is a link to more pictures as well as a video of it running. Those big injectors are noisy! Big thanks to my friend Yamen (in the video) and his experience with Megasquirt! More to come!
I have EFI up and running (more or less)! As these things go, I ended up shorter on funds than expected and picked up a used 4bbl TBI setup. It came complete with fuel rails and injectors. I modified the rails to get rid of some unnecessary stuff and clean it up. Otherwise, it bolted right onto the Edelbrock RPM manifold.
The fuel lines were rerouted from the original carb application (with the modified in-tank pump) and the pressure was set to 43psi. A friend and I did some testing to make sure the TPS and injectors were working, which they were.
And then crank, crank, boom. It fired up and ran, then died. One of the big challenges of tuning EFI is the cold start, and it was 40-some degrees in my garage. Eventually we tracked down a non-working IAC as the culprit. After that, some coaxing and tuning got it up to temperature and idling.
The IAC still doesn't work, so I need to track that down. It could be wired wrong, configured wrong, or just plain not wired up on the PC board. The O2 sensor is also showing noise that will make tuning impossible. No idea what that's about yet, but noise has been my biggest issue, so no surprise there. And there's a lot of clean-up to come under the hood before it drives anywhere.
Here is a link to more pictures as well as a video of it running. Those big injectors are noisy! Big thanks to my friend Yamen (in the video) and his experience with Megasquirt! More to come!
dpstark2
Well-Known Member
Turns out I had one coil on the IAC wired backwards. No matter how many times I looked at it this weekend, I couldn't see it. The computer can now move the pintle just fine- but I think it's moving it the wrong direction. In that case, the two coils need to be swapped.
It also looks like the cause of noise on the O2 sensor is likely the four massive low-z injectors driven with PWM. This is a known issue in some situations where current eddies and voltage spikes are induced on the 12V supply from the injectors, just like the spikes your ignition coil puts back onto the battery. Only, in this case, it's getting into the digital circuits. There are some easy tests to isolate this. I hope to have it sorted out by Saturday night.
It also looks like the cause of noise on the O2 sensor is likely the four massive low-z injectors driven with PWM. This is a known issue in some situations where current eddies and voltage spikes are induced on the 12V supply from the injectors, just like the spikes your ignition coil puts back onto the battery. Only, in this case, it's getting into the digital circuits. There are some easy tests to isolate this. I hope to have it sorted out by Saturday night.
kb67mopar
Well-Known Member
Have you thought about taking the wires coming from the O2 sensor and twisting them? Making a twisted pair will help alleviate the noise but you know that you are an engineer.
Also to help with noise issues make sure you have the ECM grounded to the engine block and not to the battery. This way the sensor reference ground is the same as the ECM. Sometimes induced interference can move the ground reference to the ecm out of range and cause loss of sensor signal or false readings.
dpstark2
Well-Known Member
You know, the O2 sensor is a different beast. Since it's a stand alone controller (LC-1), and it's entirely digital, I'm not sure how it should be grounded. The official stance is to ground both of its grounds (heater and electronics) to the same place as all the other stuff, which I did. However, some people suggest taking the heater ground directly to the battery. Later models also have a dedicated sensor ground, in which case shielding and twisting the pairs would make sense. In my case, not so much.
Plus, the "noise" is "real" in that the controller thinks it's seeing that noise. It's on both the analog output (to the Megasquirt) as well as the digital serial output- so it's not being induced on the sensor analog output. We did a test and powered the O2 off its own battery, and noise was less, but not by much. We then switched to current limiting resistors for the injectors, instead of PWM mode, and noise went away. So it's definitely the injectors. However, the current limiting resistor method was nearly impossible to get tuned. Maybe after more research, but I think PWM is the better method (and cleaner).
By the way, all the Megasquirt analog sensors have a dedicated ground line. They must all tie to this line, which goes to a dedicated filtered ground on the PC board. All other grounds are direct to the engine block, but (as instructed) not to the same place as the batter ground.
More work to do, but at least I can drive it now! Time to break out the oscilloscope and see what's actually going on with the power lines.
Plus, the "noise" is "real" in that the controller thinks it's seeing that noise. It's on both the analog output (to the Megasquirt) as well as the digital serial output- so it's not being induced on the sensor analog output. We did a test and powered the O2 off its own battery, and noise was less, but not by much. We then switched to current limiting resistors for the injectors, instead of PWM mode, and noise went away. So it's definitely the injectors. However, the current limiting resistor method was nearly impossible to get tuned. Maybe after more research, but I think PWM is the better method (and cleaner).
By the way, all the Megasquirt analog sensors have a dedicated ground line. They must all tie to this line, which goes to a dedicated filtered ground on the PC board. All other grounds are direct to the engine block, but (as instructed) not to the same place as the batter ground.
More work to do, but at least I can drive it now! Time to break out the oscilloscope and see what's actually going on with the power lines.
dpstark2
Well-Known Member
The Megasquirt forums suggested adding some bypass capacitors to the injector flyback circuit to quell the noise. Thankfully, it worked! The O2 sensor is now quiet and working very well. I was able to drive the car in many different conditions and let the computer tune the fuel map, which got me back to nearly the level of tune I had on the carburetor. Starts are a little hard still- that's another table that needs tweaking. And idle control needs some tweaking too. But at least it smokes the tires again!
Mr.SpeedFreak
Well-Known Member
Damn that's awesome to see someone on here getting their feet wet with megasquirt! Glad to hear you got it running and into the tuning phase. I'm hoping to go full sequential with an ms3x like Redmist over on Dodgecharger.com, except probably using an Accel throttle body and isolated runner intake modified for injectors, like an M1 or something. Another guy over there did that and it looked like a nice setup, although the edelbrock pro-flo's pretty nice too
Here's Redmist's thread, awesome stuff:
http://www.dodgecharger.com/forum/index.php/topic,99432.0.html
Here's Redmist's thread, awesome stuff:
http://www.dodgecharger.com/forum/index.php/topic,99432.0.html
dpstark2
Well-Known Member
Wow thanks for the reading material! That jeep cam sensor may be the ticket for going full sequential. I had planned to modify a points distributor, but we'll see if this works better. I'm a little limited right now for funding, but hopefully in a few years it will be full sequential without a distributor. I have the MS3X already, but I just had to get this thing done... and it was cheapest to do it this way for now.
dpstark2
Well-Known Member
For any of you EFI-knowledgeable people following this thread:
Does anyone know if a stepper motor-type IAC valve should completely seal when it is completely closed according to the computer?
I have verified that it is fully extended and the pintle is up against the machined surface in the throttle body, but there is still vacuum there. The pintle and bore are clean. It just looks like they don't entirely match up. It's a very small leak- idle doesn't change at all if you plug it with a finger, but it does suck your finger in. At this point, it doesn't seem to be an issue, but I've Googled myself half to death over this and haven't found an answer.
For anyone else interested, the car has been quite reliable under EFI. Tuning is progressing, but I think the base map was pretty far from ideal for this motor. The initial map was so rich, I think it ruined some plugs. I need to get a new set in there. Cold starts are still an issue, but it runs great at temperature. The logging feature on the computer has been somewhat useful, but the auto-tune feature has really been the best bet. Hopefully I can get the tune sorted a bit better and move on to computer controlled ignition.
Does anyone know if a stepper motor-type IAC valve should completely seal when it is completely closed according to the computer?
I have verified that it is fully extended and the pintle is up against the machined surface in the throttle body, but there is still vacuum there. The pintle and bore are clean. It just looks like they don't entirely match up. It's a very small leak- idle doesn't change at all if you plug it with a finger, but it does suck your finger in. At this point, it doesn't seem to be an issue, but I've Googled myself half to death over this and haven't found an answer.
For anyone else interested, the car has been quite reliable under EFI. Tuning is progressing, but I think the base map was pretty far from ideal for this motor. The initial map was so rich, I think it ruined some plugs. I need to get a new set in there. Cold starts are still an issue, but it runs great at temperature. The logging feature on the computer has been somewhat useful, but the auto-tune feature has really been the best bet. Hopefully I can get the tune sorted a bit better and move on to computer controlled ignition.
mentor70
Well-Known Member
you will need an o-ring of some kind to seal the sensor off.
dpstark2
Well-Known Member
That's not what I'm referring to. This sensor uses a flat gasket to seal to the throttle body. That part is sealed. What I am referring to is whether the passage from inside the air cleaner to inside the manifold should seal when the pintle is fully extended. Meaning, when it is fully extended, there is currently enough vacuum at the port inside the air cleaner to suck in my finger. I wonder if that is correct or not.
Similar threads
