Megasquirt - a story of injection

[boofhead]

It has been asked to create a thread for a Megasquirt install so this will serve the purpose. It will be a longer term thread as the car is still a project - this is Shaunp's build though generic info will also be included here.

Current state of this project install is (I am involved in this build)
1) Shaun purchased the MS3 diy kit plus generic wiring harness
2) Boof has built and tested the MS3 with a base tuned I think will get it started
3) Boof has built a custom relay pack and fusebox for the wiring of the system
4) Shaun has built the engine (mostly there) - 347 should have 480 to 500 hp
5) Shaun has purchased an 8 stack injection set up - sexy as can be.

TODO
1) Get a few last items done before engine goes into the car, e.g. brake lines and such.
2) Mate the engine to the 4 speed and install engine
3) Install the 8 stack, water pump etc
4) Install BA pump into fuel tank and return fuel line
5) Install EFI wiring
6) Go vroooommmmm
7) Finish the car
Tune it then thrash it

I (and SP) will include further info and background info as we go forward

[Dwayne]

[Husky65]

Great thread. You going to get rid of the dizzy?

[boofhead]

Megasquirt ECU background

Megasquirt ECU started originally as efi332 project (and I built one) then grew into Megasquirt after the direction changed from learned lessons. There have been a number of revisions though what is important now is there is;

1) MS series of ECUs - more hobbist hands on solutions
2) Microsquirt series of ECUs - simple smaller design originally for motor cycles
3) Pro series of ECUs - similar to the fully featured MS3 but packaged just as other commercial ECUs are such as a Motec etc ECU.

Capabilities increase in most cases in the series. The MS3 is fully capable of anything you need, VVT, Nitrous, PWM fuel pumps, Coil on plug, sequential injection, etc etc etc. In most cases you do not require all those features so you can spend less and still have an excellent ECU control.

In the MS series;
1) MS1 - Oldest ECU with slowest CPU is still available though I would not go less than an MS2.
2) MS2 - Most builds need no more than this ECU. Microsquirt is a cut down version of the MS2.

Boofs Mustang, gbx78, Hybrid, Luke (microsquirt), use a MS2 level ECU. I have installed MS2 on other cars as well. The main issue is IO might be limited depending on what extras you may want. For example, my mustang I controlled two fans separately, programmable shift light (led on the dash), idle up taking 4 extra IO pins. Hybrid went coil on plug (waisted spark) 4 extra IO and gbx uses 1 IO for idle up.

3) MS3 - latest ECU fast CPU lots of IO plus most effort in the firmware that runs the ECU. Can be operated as full sequential injection for V8 cars.

SP and I discussed the option and I felt a MS3 (without the sequential - good for later upgrade if desired) was the best option. He agreed. A little more expensive but future proofs it a little.

[boofhead]

Great thread. You going to get rid of the dizzy?

We could though the plan at the moment is to run a MSD Pro that is locked out. We still need to check it fits with the 8 stack injection. Lukes 8 stack uses a distributor so should be ok as long as it is not a TFI distributor which is to large.

Hybrid is no longer running a distributor - one of his recent changes - he is using LS2 (edit: LS1) coils. I was going to go down that road as well but he beat me to it. It is detailed in his build thread.

[hybrid]

I used LS1 coils - $50-$70 for a second hand set from ebay.

[Husky65]

Yeah cool. I've been researching MS the last few weeks every night. Really want to pull the trigger soon

[boofhead]

If going the MS route you have a choice to build the board or purchase pre-built.
The v3.0 board is the DYI board while the 3.57 is the per-built. If you have little experience soldering small components get the 3.57 per-built version. gbx78 is using a pre-built board that I modified to suite the TFI distributor he is running. The extra cost is approximately $100 extra for the per-built version.

In this efforts case I built it myself (as I have 4 times previously).

This is an image of a MS2 (I built an MS3 for SP though the board below the CPU board that you solder is the same).



I did add a Mapdaddy - dual air pressure sensor so it has MAP for air fuel calcs and can also provide dynamic fueling changes due to height changes. It also has the advantage of handling far higher boost levels - not important for this build.

The final box is taller for the MS3 than the MS2 - here is an idea how it looks.

[boofhead]

Yeah cool. I've been researching MS the last few weeks every night. Really want to pull the trigger soon

Ask away as there is alot of knowledge here.

[Husky65]

I was looking at the ms3 pro, which has now been outdated the the ms3 pro evo. It will give me options to play with things in the future when I get bored with stuff. One of the things I have been looking into it, is the MAF vs MAP side. I know the downsides to a MAF in regards to air leaks and a potentional restriction(depending on the setup of course), but for tuning is it superior to the MAP, in regards to setting up VE tables etc? I understand 8 stack etc is different, but if you have the ability to run a maf, is this the way to go?
I know heaps of people who used to convert from factory map setups to maf for ease of tuning and the ability of the maf to adapt to mild modifications.

Enlighten me

[Shaunp]

I don't plan running coil packs, I don't like the LS engine look. I've got one of the small diameter MSD pro billets so I think we should be ok, it really should be not much different to running Webers for disy fitment I dont think. I've actually been giving the pump set up some thought. I've been thinking about not using the BA in tank pump. but rather I've been thinking about fabricating a fuel sump, and tig welding this to the lower rear of the tank, this would keep all the fuel lines out side the car. I figure if I just drill a few holes in the tank prior to welding the sump on, this would take care of surge, Perhaps feed the spill return back via the original pick up, or conversely back to the sump.

[Husky65]

I like this setup that one of the guys in the states did. He ended up making a plate that bolted on top I believe. Nice and neat, yet easily accessible. You may still be able to use the BA fuel pump with a higher 70 tank.

[Shaunp]

Like this

http://www.vpw.com.au/Category/Index/170238

[boofhead]

I've been thinking about not using the BA in tank pump. but rather I've been thinking about fabricating a fuel sump, and tig welding this to the lower rear of the tank, this would keep all the fuel lines out side the car. I figure if I just drill a few holes in the tank prior to welding the sump on, this would take care of surge, Perhaps feed the spill return back via the original pick up, or conversely back to the sump.

Very good idea - I like it. Simple and neat. Ticks all the boxes.

My approach was great but complex so yep do the fuel sump idea. Good stuff.

[boofhead]

I was looking at the ms3 pro, which has now been outdated the the ms3 pro evo. It will give me options to play with things in the future when I get bored with stuff. One of the things I have been looking into it, is the MAF vs MAP side. I know the downsides to a MAF in regards to air leaks and a potentional restriction(depending on the setup of course), but for tuning is it superior to the MAP, in regards to setting up VE tables etc? I understand 8 stack etc is different, but if you have the ability to run a maf, is this the way to go?
I know heaps of people who used to convert from factory map setups to maf for ease of tuning and the ability of the maf to adapt to mild modifications.

Enlighten me

The MS3 Pro (1st gen, Evo, Ultimate) are top class systems - lots of room for future fun and games. Good stuff.

I have never bothered as MAP (SD) using VE tables and such works well for me.
MAF means your tuning a MAF transition table instead of a VE table. I do not see it changing much in the effort though it may be faster to tune - not sure - you could argue it is more accurate (as long as no leaks) way to measure engine requirements.

Go for it - good to see the write up.