Megasquirt/EDIS Project

[kmead]

This is great, please keep posting more details as you get the time. There are many of us who are watching with rapt attention.

I really like the idea of MS as I want to go to a hotter cam which doesn't play well with the existing Bosch system and the manner it uses to measure air into the engine.

Do you have an opinion on how important it is to have MS control spark? Does MS use the ignition timing wheel to judge injection time like some modern injection systems do?

 

[///Mike]

You can use a MAF but there doesn't seem to be much point in it. I read quite a bit about it and it looks like a huge headache with no significant advantage. However, I do kinda wish I'd gone with MS3 so I could run sequential injection. There could be a worthwhile advantage in performance and fuel economy there.

It's interesting that you say that. I wanted sequential for the longest time but was finally talked into the concept that it doesn't really help power, only emissions. I gather your research indicates there are power gains as well?

On the other hand, I got the idea that a properly set up MAF would probably improve driveability across the powerband. It does seem like a lot of hassle and it probably wouldn't be worth messing with if I were going to stick with the stock plenum. But for real power I think ITBs are a requirement, and in that case I think MAF is the best way to measure air. So my plan was to first set up MAF on the original plenum since it's a much less documented path in the MS world, and then move to ITBs once I'd figured out the whole MAF thing.

I'd welcome any comments, as you've clearly done your research.

Thanks,

 

[TonyK]

Fuel only.

MegSquirt uses a pick up wire from the ignition coil to fire the injectors, so the toothed wheel is only required for engine timing for spark control. When Fiat changed the X to electronic control it used stand alone components. Fuel was independant of Spark, so either could be changed, upgraded or plain messed with, without effecting the other. That being said having MegaSquirt control both spark and fuel now leaves one engine control unit to do both.

TonyK.

Grimsby Ontario Canada.

 

[racerx]

Wow

This is grate. I too have always wanted to do this .as soon as you get more time can you post a list of all the components necessary to do the job and the price and if possible the source that would be grate. Thanks

 

[Kevin B (Asheville NC)]

easy source:

http://www.diyautotune.com/

 

[racerx]

yes

Thank you I know about how to get the CPU but what I'm talking about Is a comprehensive list of parts on the whole project Like when I posted my VW radiator project It give folks a chance to see if they can make a budget to do it them selves.

 

[Matthew]

It's interesting that you say that. I wanted sequential for the longest time but was finally talked into the concept that it doesn't really help power, only emissions. I gather your research indicates there are power gains as well?

On the other hand, I got the idea that a properly set up MAF would probably improve driveability across the powerband. It does seem like a lot of hassle and it probably wouldn't be worth messing with if I were going to stick with the stock plenum. But for real power I think ITBs are a requirement, and in that case I think MAF is the best way to measure air. So my plan was to first set up MAF on the original plenum since it's a much less documented path in the MS world, and then move to ITBs once I'd figured out the whole MAF thing.

I'd welcome any comments, as you've clearly done your research.

Thanks,

I can certainly see where a MAF would be the way to go with ITBs. I know that with ITBs often MAP doesn't work well for idle and speed density/alpha-n blending isn't an ideal setup either. The MAF would solve that problem once you got the calibration right.

The Megamanual had me thinking that sequential might be a waste of time but I think it downplays it because MS doesn't (until very recently) do it. Talking with experienced tuners they all say sequential improves performance and fuel economy in almost every case. Since it doesn't add significantly to the complexity of the system it's really the way to go.

 

[Matthew]

This is great, please keep posting more details as you get the time. There are many of us who are watching with rapt attention.

I really like the idea of MS as I want to go to a hotter cam which doesn't play well with the existing Bosch system and the manner it uses to measure air into the engine.

Do you have an opinion on how important it is to have MS control spark? Does MS use the ignition timing wheel to judge injection time like some modern injection systems do?

I could have left the stock distributor and it would have worked fine. However, having an infinitely tunable spark table is really the key to drivabilty, economy and maximum power at all RPM and throttle positions. In order to tune the ignition properly you really need to spend several hours on a dyno tuning each individual MAP bins for maximum power. Doing this on the street is extremely difficult and will not achieve the best results unless you are lucky. Some my find the increased adjustably of MS controlled spark to be more trouble than it is worth.

 

[bikesandcars]

What does a stock F.I. 1500 running correctly really put down at the rear wheels? Are there any dyno examples out there? The only dyno sheet I found in a quick search was from a lemons car that 'bikesandcars' posted last year. They started at 65 hp, and finally tweaked and fine tuned to 83 hp.. it had a cam in it as well (from elgin) and presumably a very free flow exhaust. (1300 or 1500 in the Lemons car? Of course that would make a difference!)

http://xwebforums.com/forum/index.php?threads/8783/

The 65 wasn't actually the stock hp, it was the un-tuned horsepower. As we were new to Fiat's we just gave it a general setup before we visited the dyno. We fiddled with several items including valve and ignition timing to get the peak up to 83 rwhp. Later on (at the race) we switched ignition wires and cleaned the carburetor and the car pulled even better... I'd like to dyno it again

 

[mx5rush]

I'd assumed so... it can be frustratingly hard to 'butt tune' a car since you only have that car to compare it to. My MGB buddy was stunned when his 'good running' car was a total dog compared to a better tuned stocker. Of course over the years he'd played with this and that, trying to get it to idle smooth in the cold, not overheat in the summer, etc. And eventually he'd succeded with a high idling, timing retarded, constantly requiring the choke mess of an engine. And this is even more common when you change camshafts and stuff. I swapped the camshaft in my old 78 Z/28 for a 'mild' performance grind and the stock carb ran so lean, my mufflers had my carpet smoking! Throwing in richer primary jets made the car driveable... but it wasn't making near the power it could have with a few dyno pulls showing what was wrong.

 

[Matthew]

I drove it 32 miles today and things are going pretty well. It's a bit rich under load and I didn't realize until the last drive that I had the rev limiter set a 6K so when I tried to check high RPM mixture it was violently shutting down. It happened a few times before I realized what was wrong.

Later this week I'm going to post a detailed description of how I built the harness and wired everything. I'll also get into some of the settings that work with the fuel system. I'm not going to go into detail on tuning because it's too complicated and quite frankly I'm still a bit of an amateur in that department anyway. Also, anyone who is going to attempt this needs to read the Megamanual and the MS Extra manual multiple times. That will answer most of your questions about tuning. The Megamanual will teach you 90% of what you will learn in a EFI tuning class. It is really very good.

 

[TonyK]

Watch Auto Tune.

Matt if you are now running and logging just watch the Auto Tune function for re-scaling your VE Map. I am not saying don't use it, but before you run the car take a look at the values. The Auto tune function when I used it breifly will put some wild values in there, either too high or too low and when the engine falls into that part of the grid the computer tells the injectors to do exactly what is in that grid value. I made this mistake and on accelleration the map value went from 155 to 5, well the engine just about shut down at 5000 rpm. I normally let my son drive the car and I look at the lap top and adjust accordingly.

You have gone great steps in short order. Well done. It took me weeks of reading, tuning and trying to be comfortable driving the car in traffic without surpises.


TonyK.

Grimsby Ontario Canada.

 

[lookforjoe]

Very cool.

Good to see this laid out. As was said, gives ideas for other directions, also.

I was thinking of converting my X to LH2.4 w/EZK ignition from a early 90's Volvo, but the remapping for the smaller displacement/output has put me off. I already have enough going on with my laptop programmable ECU in the Volvo.

What I may steal from your thread, is using a newer TPS, and adding CIS (from an '80's Volvo) to get rid of the aux air slide & improve idle control, especially w/ AC!

 

[Matthew]

Project Details

When I first began researching this project I started by reading a lot. Here is a list of Megasquirt and tuning resources that I used:

Megamanual

MS Extra Manual

MSExtra Forums

Megasquirt Forums

A write up about installing Megasquirt on an X1/9 specifically.

DIY AutoTune Tech Articles

EFI University

I read the Megamanual a number of times and planned out how I was going to do the install. After a couple of months I realized that I had MS2 Extra firmware which has a whole other manual :doh:

Once I started collecting the parts I spent a lot of time tracking down all the supplies I was going to need to build the harness and do the install in a way that I would be happy with. Here are some of the places I used for the supplies and parts I needed:

DIYAutoTune Pretty much all the major MS components and wiring.

Waytek Wire, connectors, terminals, relays, crimpers etc.

Ballanger Motorsports Connectors and terminals

Eastern Beaver OEM type terminals, connectors and crimpers.

Digikey Electrical components of all types.

Vetco Electronics Miscellaneous electrical supplies (Local to me)

Fry’s Electronics Miscellaneous electrical supplies (Local to me)


I wanted to build a harness that was as close to OEM quality as I could reasonably get. It would have been easier to just use regular and water proof butt connectors and electrical tape and it would have worked fine. However, that wasn’t the look I was after. With the exception of the EDIS harness all the connectors were crimped directly on the harness rather than spliced on pigtails. Were I needed to splice I used small open barrel connectors and adhesive lined (dual wall) heat shrink.

There were three sizes that I used. From left to right Digikey part numbers A29929CT-ND, A100600CT-ND and A100590CT-ND

They are a little tedious to work with but they make really nice compact splices that won’t leave a big lump in your harness. For tinned wire or sensors where excellent continuity was critical I also soldered the splice.

Here is a finished splice with adhesive lined shrink tube over it.

When working with connector terminals the tools required can be expensive. I have a bunch of crimpers but only use three or four of them most of the time. Unfortunately there are dozens of open barrel crimpers and it was rather hard to figure out which ones I needed. The crimper that worked best for the splices turned out to be a $20 one I bought at Vetco.

For the wire harness I started with a DIY AutoTune 12’ Megasquirt harness. The wires are high quality TXL automotive crosslink wire labeled ever 4”. This made building the harness almost fool proof.

To shield the wires to the crank position sensor and the two wires from the MS to the EDIS module I used ¼’ tinned copper tubular braid. I put the wires need to be shielded through the braid and soldered a wire at one end that went to ground. One of the wires for the EDIS from MS on the DIY AutoTune harness is already shielded but since the other one wasn’t I just put both through. It doesn’t hurt to have too much shielding. I finished the shielded wires by covering the whole thing with shrink tubing. The finished product was too big to fit in the split loom so I wrapped it on the outside. This worked okay because the EDIS wiring leaves the main harness to go over to the right side of the engine bay immediately after it exits the spare tire well.

Here you can see the harness inside the car with the shielded wiring running on outside of the split loom. Also you can see how I ran the vacuum tube for the MAP sensor inside the harness using Cohline 3010 vacuum tube.

The EDIS module lives under the right plastic cap next to the engine lid. I attached it with double back tape. Once it’s installed on the car you can’t even tell it’s there.

The crank sensor is on a custom mount bolted to the front seal plate. The trigger wheel is a stock ’91 Ford Escort part that has been bolted to the front of a slightly modified stock Fiat crank pulley. This was fairly simple to make but required some machining and welding. I would have rather had the trigger wheel behind the pulley but there was too much fabrication involved and the only source I had for one like that was unable to make it for me.

The sensor is a little close to the plastic splash shield but it works.

The Coil I mounted using a coil mount from a ’91 Ford Escort and a piece of ¼” aluminum. I really want to mount the coil in the same place the distributor cap was. I will eventually try again to get the mount I need to do that and move the coil but for now this works fine.

The rest of the harness in the engine compartment was pretty straightforward. I just routed the wires to the sensors and injectors, crimped on the terminals, inserted them in the connectors, put split loom over the wires and wrapped it up with 3M friction tape from Home Depot.

I attached the wire harness to the factory locations under the throttle body using fir tree mounts from Waytek. The vacuum pipe exits the harness near the #1 injector and attaches to manifold vacuum with a “t” fitting in the fuel regulator vacuum hose.

All but one ground wire for MS attaches to a single ring terminal at the factory location on the valve cover. The remaining ground wire is sliced to several more wire inside the harness and provide a ground return for each of the sensors.

Fir tree mount

I also made a dummy harness for the cold start injector. I didn’t like that it looked like something was missing. I have plans to build a blanking plate/fuel pressure test port combo to replace the cold start injector eventually.

The wide band controller is an Innovate LC-1 and is mounted to a bracket I made and attached to the carbon canister bracket. This puts it in the perfect location to connect it to the oxygen sensor. The wiring for the controller I routed along the factory harness behind the coolant bottle and air filter then along the firewall and into the spare tire well.

The LC-1 requires occasional calibration. For that purpose it comes with a button and LED that connects to a wire from the controller. The button is wired to ground the wire when pressed. They don’t provide any neat way to set this up so I installed the LED and switch in a small project box that I bought at Fry’s Electronics.

Wiring the power supply for MS, the injectors, fuel pump, WB controller and idle valve required removing the Bosch double relay. The double relay is used for several reasons three of which are no longer needed with Megasquirt and one that will not work. The double relay turns on the fuel pump circuit when the ignition is turned on but only if the switch in the airflow meter is closed. Since MS grounds a relay to turn on the fuel pump is can’t be used with the double relay without adding yet another relay to control the double relay. That would be too messy in my opinion. The double relay also serves as a link between the body harness and FI harness to allow the starter circuit to control the cold start injector. That feature is no longer needed. It’s my belief that the main reason they used the double relay instead of two less expensive relays is because it was the only way to keep the FI and body harness completely separate which simplifies installation on the assembly line. Anyway, do yourself a favor and ditch the double relay.

The double relay has three power supplies. One is a large red wire that comes directly from the battery and is an unprotected circuit. This kind of bothers me but it’s worked for 30 years so I kept it as is. There is also the fuel pump circuit that comes from the fuse box and is hot all the time. Last but not least there is the power from the fuse box for the double relay. This one is hot only with the ignition on.

These three power supplies are all you need to make MS work. Two relays take the place of the double relay. One is for the fuel pump and the other is the “main” relay which will power MS, the injectors, the idle valve and the WB controller. The switched power wire goes to terminal 86 on the main relay and the large red wire goes to terminal 30. Since the large red wire is an unprotected circuit the wire from terminal 87/87a on the main relay needs to go directly to a fuse box that will distribute power to all the devices. Terminal 85 on the main relay goes to ground.

The B+ wire from the fuel pump fuse goes to terminals 86 and 30 on the fuel pump relay. It supplies power to the relay winding and the fuel pump once MS grounds the relay turning it on. The wire to the double relay that goes to the fuel pump can be connected directly to terminal 87 on the fuel pump relay. Since the circuit going to terminal 30 comes from the fuel pump fuse no additional protection is need. Terminal 85 connects to pin 37 of the MS harness.

I ran the power from terminal 87 of the main relay to a Blue Sea Systems 5025 6 position fuse panel with a ground bus. The Blue Sea Systems fuse panels are the nicest I could find and are much better than anything the auto parts stores sell. These can be found at marine supply store. I got mine from West Marine.

Here is the relay setup. I used interlocking panel mounted relay sockets from Waytek. These accept the factory relay terminals which Waytek also sells.

Here is the fuse box. It supplies power to the injectors, idle valve, Megasquirt and the WB controller.

Once everything is installed the software needs to be setup. This really should not be attempted without many hours studying the Megamanual and MS Extra manual. However, I will share some of the settings that are known to work with the X1/9 fuel system and what I am currently running. Much of this information I found here on Xweb or various places around the internet.

Required fuel is the setting that tells MS how big your injectors are and how big your engine is so it has some idea how much fuel they inject and how much they need to inject. 15 is the number most people suggested and it seems to work well

The Injector Characteristics setting tells MS what kind of injectors you have. The X1/9 injectors are low impedance “peak and hold” injectors. This means they need a certain current to open them and a lower current to keep them open. If you run in high impedance mode it will blow the injector fuses. These settings work well for me.

The GM IAT sensor is plug and play with MS but the Bosch coolant sensor requires some calibration. Under Calibrate Thermistor Tables select “Coolant Temperature Sensor” from the upper drop down box and “Saab (Bosch)” from the lower box. Once you write to controller the CTS will read properly.

When I first started working with the MS setup I was using the original tuning software called Megatune. I got really frustrated with it and finally gave up. I felt like you need to be a software engineer to understand the setup procedure. I switched to Tunerstudio MS and found it to be much easier to use. Both are free but the registered version of Tunerstudio ($40 I think) has features that you will want.

I still need to install the idle valve and setup closed loop idle and I will let everyone know how that works out. I probably won't make it back to the dyno for a few months so don't expect any exciting numbers soon.

I tried to be as detailed as reasonably possible but if there is anything I left out remind me and I’ll add it.

 

[budgetzagato]

Awesome post made sticky

...just to preserve the info and keep it handy, thanks Matthew.

 

[sdlsaginaw]

What a beautiful install

I'll be going through this with the Uno Turbo motor if/when it finally goes together (knock on wood).

I love the look of the friction tape, I'd like to do the same thing but I think I'll have to locate something more heat resistant because of the turbo (for the injector area at least).

I think I'll also use those splices, they look great.

Thank you for detailing your installation!

 

[Matthew]

The car is running awesome. I'm currently running no EGO correction, no acceleration enrichment and no idle valve. The idle could hardly be better. It runs around 1000 +/-50 RPM and never ever even thinks about stalling once it's at least partially warmed up. Warm up and after start enrichment needs some work but it's acceptable until I get the VE table perfected. The car runs better than it ever has and I'm not even close to done yet.

Today I tried to figure out why the tach doesn't work and I'm kinda lost here. I have a tach signal and every tach I connect to it works fine except the one in the cluster. It worked when I ran the car for about 5 minutes with EDIS and L-Jet but now it almost never works and when it does it's never for more than 20 seconds and it will always quit when reved over 1500 RPM. Those of you that are running EDIS I'd really like to know how you got your tach to work.

I also messed with the idle valve a bit today. It doesn't seem to work properly and I'm not sure why yet. I can scope the PW and verify MS works fine but the valve doesn't move when the PW changes. The documentation for setting up the valve is really bad, especially for closed loop idle. I'm not going to actually install the valve until I can get it to at least appear to be working properly.

 

[ng_randolph]

Tachometer issue

This link looks somewhat relevant. I think this was brought up on Xweb forums a while back as well.

 

[dcioccarelli]

Hi Matt,

Today I tried to figure out why the tach doesn't work and I'm kinda lost here. I have a tach signal and every tach I connect to it works fine except the one in the cluster. It worked when I ran the car for about 5 minutes with EDIS and L-Jet but now it almost never works and when it does it's never for more than 20 seconds and it will always quit when reved over 1500 RPM. Those of you that are running EDIS I'd really like to know how you got your tach to work.

I also messed with the idle valve a bit today. It doesn't seem to work properly and I'm not sure why yet. I can scope the PW and verify MS works fine but the valve doesn't move when the PW changes. The documentation for setting up the valve is really bad, especially for closed loop idle. I'm not going to actually install the valve until I can get it to at least appear to be working properly.

The X1/9 tach needs a high energy spike to run properly (some tachs simply need a 12v square wave). I know that for an MSD ignition, there is a "tach adaptor" which is basically a coil that you connect to the tach output to simulate this spike (you can even use a standard ignition coil). Be careful that the tach output you are using is properly protected though (typically via a zener diode) to prevent any damage caused by the collapsing magnetic field in the coil.

As for the idle valve, I know that there are two types: PWM and stepper. Are you sure you have the correct type?

Cheers,
Dom.

 

[TonyK]

Idle control.

Matt, I found the same thing with my conversion. Idle was way better than the carb ever was. Since I built a custom intake manifold I did not plan on using the Idle control valve. Just a point here that I am not sure of Idle control normally applies to MAP. control and you and I are using Speed density. This is due to the wild camshaft and fluctuating MAP values. So as long as you are on Speed density, the idle valve will not work as it allows more air into the manifold, thus changing the MAP value and the ECU increases the fuel to the engine. Idle control is MAP based.

TonyK.

Grimsby Ontario Canada.