I've decided to take the leap.
 

So I ordered my Megasquirt hardware this morning from DIY. The complete bill of materials is:

1 x MegaSquirt-I Programmable EFI System PCB3.0 - UnAssembled
1 x JimStim
1 x ModKit, PWM IAC
2 x Modkit, Boost Control (one for EBC, one for future WI)
1 x GM Boost Control Solenoid
1 x GM Boost Control Connector
2 x Modkit, Relay Control (one for fans, one for future WI)
1 x Modkit VB921 High Current Ignition (future DCC)
1 x 3/8" NPT Steel Weld-On Bung
1 x GM Open Element IAT Sensor
1 x MPX4250 2.5 Bar MAP Sensor (for realtime baro)

A few things are being provisioned for in the build that won't be immediately exploited, in the name of not having to pull and modify the unit later. I'm planning to configure the ignition drivers for both igniter-based operation as well as direct coil control. IOW, Q6 and Q8 will be wired to the DB37, and the two VB921s will be wired out on a separate Molex connector.

Ditto the WI. For now, I'm going to keep my DevilsOwn progressive controller, but I'm going to configure a relay driver and a high-current PWM driver, and bring those out on a separate connector as well for the day when I decide to buy a high-speed valve.

My KnockSenseMS came in last week, and I think I'll probably hold off installing it until the MS comes together. No sense making it work with the EMU only to change it later.

I'm planning to re-use the factory CLT sensor- from what I understand, the
configuration for this is available.

Unless I find a compelling reason to do otherwise, this will be a standalone setup. Between idle and cold-start this should be challenging. Wiring will be direct from the DB-37 into the vehicle's harness. It's so cut up anyway (thanks to my current EMU setup) that trying to piece it all back together than then build a plug-n-play harness would be a waste of time. I'll just finish cutting it and direct-wire the DB-37 into the factory wires.

Firmware will be MSnS-E, looks like 029y4 is the current build. I've also just downloaded a copy of MegaTune-Extra, 225p1 it says. Not sure yet what other software I'll absolutely need- I'm still digesting Brainey's ever-evolving howto thread.

Guess I should start looking (again) for an analog TPS. Probably easiest to just grab a whole throttle body if I can find one...

Can't believe i forgot to hit up the junkyard. My next final is at 1 i'll be done around 2. Should know by the end of the business day if I can get you that TPS... sorry

Glad you're going ms. Did you get that extra map sensor for realtime baro to save money compared to just getting the 4bar mapdaddy that has the extra sensor integrated?

You will want to run 029V for MSNSE.

True but a 4 bar capable setup would be sweet. GL Joe

Dork. I totally bought like half that stuff, you don't need your own stim, etc. Ah, well.

It's be interesting to see the differences between your MS-I and my MS-II install. I think I just like doing things the hard way.

How does the Devil's Own work - how does it control water delivery? Personally I'd say, with as pricey as they are, you could just sell it. :-)

I need to look into the VB chips, I have my stock coils, but I have some GM ones, as well, which are pretty hot. I have some sort of ignitors for it, but I don't know what they are just yet, I put them in a box to "deal with later".


I totally don't see why you'd want to half your MAP resolution. 2.5 bar is the way to go. What I don't know is which (4bar or 2.5) is higher quality. I know folks with a 3bar GM who said going to ANY other sensor helped because those are low quality.

Anyway, I did the same thing with the 2.5 bar.

Well, I've since learned that the auto TPS apparently will not bolt up directly to a manual throttle body. So if you can score the whole TB for a reasonable price I'll still take that off your hands, otherwise I'll keep searching.
Yes, I researched the MapDaddy and didn't really see any advantage to using it as opposed to simply getting a second MPX4250. I don't plan on running any more than 21 PSI of boost. And from a purely technical standpoint, the 2.5 bar sensor should outperform the 4 bar sensor in terms of useful resolution- they all use the same output voltage range, but with the 2.5 bar sensor it's scaled in such a way that nearly all of that range is "useful". Put another way, it has a better S/N ratio.

Interesting. I'm going solely by what's posted at the MS-Extra site here: http://megasquirt.sourceforge.net/extra/index.html

I haven't been able to find a release history page (release notes on the various versions) and unfortunately a lot of the links that seemed promising point to nowhere.

Well, I kinda wanted to have my own stim available as I plen to do some fairly significant bench testing. My current project (other than fixing this damn oil leak) is to gather scope captures of all of the significant engine events. I did coil current vs. igniter trigger over the weekend. For that I brought home the 2 channel color scope with the ultra-fancy current probe (a clamp-on inductive probe capable of reading steady-state DC with 10ma accuracy).

Next I need to get access to the 4-channel scope so I can do IGN and INJ timing relative to CMP/CKP, and I'll probably do IAC while I'm at it. Once I've got all that data gathered up, I'll assemble the hardware and try to get all the funky stuff (wheel decode, INJ and IGN timing, etc) set up and verified on the bench. It'll take a while.

It is MAP-based only. The controller hooks up to a 2-bar sensor, and as MAP increases it PWM's the main pump motor directly. Ideally, I'd like to have a HSV slaved to an INJ channel, so I can simply turn the pump "on" and control flow with the valve. But that's down the road a ways. Once it's done however I'll definitely sell the controller.


So far as I can tell, direct coil control, even for four outputs, is a pretty simple affair. You just need to install some high-current drivers and flip the output polarity in the software. Like I said, that's a "future" project for me.

Whoa, +1 for DIYAutotune already. I put my order in two hours ago, and I just got a tracking number!

too late to suggest getting an MSII chip then. put that on your todo list. someone needs to fully tackle and catalog it.

Ha! Who knows, they might have a tracking number out but not yet have it shipped. :-) Doubtful.

You know, Joe, if you really want to DO a lot with it, having many times the CPU capability will likely be a good thing (hence my MS-II choice). You don't want to be against a wall of limited CPU, or you might as well just buy a LINK. :-)

I'm not sure how LONG I'll use the stim for - it's something I want access to, but I don't think I'll use it all that often, you can get a lot by letting the stock ecu run the motor and monitor the MS through software. Either way, at least I've got it.

Um, I've got a few 4 channel scopes at work, one really fancy I probably shouldn't mess with, and another standard 4 channel that would work fine I'm sure.

Yeah - that sounds like it could be done better. I'm not sure you want water proportional to fuel. It's a good first stab, but I think you'd want a rising multiple with either boost or maybe possibly RPM. I think the AEM does this, proportional to fuel but non-linear.

Well, as I understand it the MSnS-E software doesn't exist for the MSII, correct? Or at least, it's in "very early alpha release". No thanks- I'm good with hardware, but I am *not* a C guru. Besides, I seem to recall asking this very question within the past month (MSI vs MSII) and being told to stick with MSI.

But hey, it's not that expensive should I see a need to upgrade later on. I guess I'll just make sure as I'm assigning pins that I'm not deliberately conflicting with any MSII defaults. As I said, I'm not constraining myself to the DB37 for my I/O, primarily because I want the wiring to be modular- one connector for the factory stuff, one for WI, one for knock, one for direct ignition, etc.

So , uhhhh, my ms2 cpu is on ebay right now. $60 BIN :)

^^shameless plug^^

That's ok, we've got several storage scopes here in the lab. Actually, I just realized that we have more scopes than engineers, if you count the old. I took the TDS3052 first since I needed it for the ignition current measurement (the current probe we have only works with that one scope) and it's in constant use so I can only pilfer it on weekends. The TDS420A (4 channel, but only 200Mhz) hardly ever gets used now that we have a pair of Agilent mixed-signal scopes, so I'm bringing that one home tonight and will probably keep it for a while.


Actually, everything I've read indicates that water proportional to fuel is exactly what you want. If you think about it, that's pretty much what you're getting from the more expensive WI controllers that look at both MAP and RPM. Well, fuel is already proportional to MAP and RPM, so locking water to fuel gets you pretty much the same thing.

Ok, then. Serious show of hands. Apart from saying "It's faster, and you can code in C rather than assembly" (neither of which I intend to do anyway) what would be the reasons for using an MSII as opposed to an MSI with MSnS-E? I'm not averse to attempting to reverse my decision at this point.

that stuff plus 16x16 ve table, 12x12 afr target table, individual tables for each injector bank, map/maf blend for airflow, tps & map blend for enrichments

MSI does a damn good job on my car, and it will on yours too. But with your background and resources, you might want to go for it. You could even do it later, if you get bored with MSI and want a new project.

Interesting point:

One thing that bothered me about the MSnS-E setup was that my VE table was 12x12, but my AFR table is 8x8. Doesn't make sense to me how you could align all the cells...

Looking at the MSII Tuning docs (http://www.megamanual.com/ms2/configure.htm) it looks like they are all 12x12. (I'm not seeing 16x16 anywhere). But this does seem to resolve a conflict, at least in my mind.

However, I'm not seeing anything about boost control. Is this available on MSII?

You mean NIB? And I assume this is you: http://cgi.ebay.com/ebaymotors/MegaS...QQcmdZViewItem

OK, somebody convince me to buy it.

MSnS-E has 12x12 VE and spark, it has AFR autotue, it has EBC. Seriously, sell me on MSII. If I'm gonna do it, I'd rather do it right at the beginning.

Fuel control to 1 µsec (100 times more resolution than MegaSquirt-I) enough reason?
Check your PM's i got your throttlebody today.

That be her alright. BIN = Buy It Now. There is not box :P But it is new in a sense that I've only tested it for 1 week. I couldn't get it to fire my COPs and was running out of time before the semester started. So I guess brand new in the sense that it is exactly the same as what you would get from DIY...

I was able to get all the input mods and fuel working, but spark is part of the triple threat and I couldn't get it within my window. The miata is my DD so you do the math.

The MSII cpu has less current coming from its spark outputs which I believe has something to do with not getting consistent, full spark. I have not messed with using the led transistors to boost the current. I would love to humbly pass the torch to you.