2533 type 4 megasquirt build

[supaninja]

I'm about to play with MS3x on my K20 swap, I know I'm excited too

[Piledriver]

There are possible ways to work around the issues you were having, (map oversample//averaging and semi-sequential) but the MS3+MS3X is a VERY nice, essentially plug and play setup, requiring ~zero science experiments in internal jumper wiring to extra modules/boards, I have no doubt you will be happy with it.

There are a few lines to route/jumpers to set if you want to use certain features, but they have pads/jumpers and excellent documentation, as V3x boards and MS3X is the only "supported" configuration, docs//tech support is a lot easier.
(Hint--- Use kynar 30 ga wire wrapping wire and dabs of 5 minute epoxy for strain relief for all jumper wires)

The V1.2 firmware is supposed to be getting the WB lag table integrated in the firmware (+ tools to ease setting it properly) so closed loop WBO2 will actually have a chance to work really well.

Right now the WB lag table is only in VeAL. Set right it makes all the difference in the world... Someday MLV may get it too.

[warp]

@Pile: I considered to modify my MS2 to run semi-sequential but I have a feeling that it will be just another compromise and still not solve my issues 100%. I have wanted to try out MS3 for a while and this seems to be a perfect opportunity where I actual need some of the nice MS3 features.
I know you have been running and experimenting with MS3, do you have a good method for calculating or measuring fuel transit time? It's critical to get right but it seems difficult to find any good methods for calculating it correctly.
Further more I find it difficult to find valid guidelines for setting injection timing right. When injectors are placed close to the valve the general agreement is that the fuel should be injected just after the valve closes. The reason being that the hot backside of the valve gives good vaporization.
In my case the injectors are placed just below the throttle plate in the throttle bodies hence the fuel is not injected right onto the backside of the valve and the fuel transit time is much longer. I'm can't find any good guidelines on how to set the injection time when the injector is placed away from the valve. I will for sure have much more wall wetting so I'm thinking I may benefit from injecting into the open valve airstream instead of injecting when the valve is closed?

[Piledriver]

The only way to know for sure is try different settings and see how the engine likes it.
We don't get a vote--- The engine decides.

Different inj timing has a definite effect on measured AFR, the closer you get to "happy" the richer it will look to the WBO2.
(for the same fuel, gets burned more effectively, unburned fuel has ~no effect on the reading, rich/lean is only measured by oxygen content %, so for example a dead miss reads dead lean)

I'm setup with the stock injector location currently, which is ~75mm from the valve... On a T4 it's just about a straight shot, and the spray pattern ~ only hits the valve/seat area of the port, all of which is quite toasty once warmed up.

The issue with a high overlap cam is most of that lovely vaporised fuel/air mix may get sucked straight out the exhaust on overlap, so you may find that shooting into moving air (start injection after the overlap period intake open) is a win in that situation.

This can require larger than "normal" injectors to get the whole shot in during the intake stroke.

Until we can pick up a DI setup cheap (probably an air-assisted Orbital style would be easiest) it will always be a compromise.

[Steve Arndt]

My experience.

Injector position ~8 inches from valve
0 - 720 degrees for a cylinder cycle
Ignition timing starts at 0 degrees in my reference plane.
I inject fuel at 390 degrees after ignition.
My ECU advances the injection event relative to RPM automatically

When I tested CB end castings with the injectors closer to the valves I had to move the injection event to 450 degrees to get similar air/fuel readings compared to further away (better mixing I surmise with them positioned further away). It also required ~20% more fuel to attain the same idle stability and AFR numbers with the injectors closer in the end castings compared to my throttle bodies.

[Piledriver]

Note: semi sequential requires no mods, just turn it on and play with the fuel timing (You get 2 shots/cycle but you can determine when and it will be consistent unlike std batch)

It may help a little, using averaging for MAP may help a LOT.

Note:I'm NOT trying to change your mind, MS3 rocks.
I'm just trying to help you get it running OK NOW.

[warp]

@Steve: I just read your baja build-thread. You made some good observations and your ITB set-up is very similar to mine. It was very interesting reading.

@Pile: I know it's a no brainer to turn on semi-sequential to try it out and play with it for a while. Had it been summer time at the moment I would probably have done it. But I'm located in Denmark (Scandinavia) where it's already getting dark and cold so my car is already in the garage for the winter. I won't be driving it for the next five months anyway so I have time to do a proper MS3 install and have some dyno time over the winter.
My MS2 may get installed in one of my other cars if I find the time. It certainly have served me well up until now.

[ProctorSilex]

warp, how did you know how deep to cut the notches in the flywheel for the HE sensor? I looked at the data sheet for the Hamlin 55075, but I only saw a specification for the distance from the gear. Thanks

[warp]

warp, how did you know how deep to cut the notches in the flywheel for the HE sensor? I looked at the data sheet for the Hamlin 55075, but I only saw a specification for the distance from the gear. Thanks

Actually I didn't mill the pockets based on the Hamlin specs. Initially I planned to use a Honeywell GT-1 sensor and the pattern on the flywheel is based on the specs for GT-1 sensor. I can't remember the exact measurements now but it's in the data sheet for the sensor.