Opinions on the "Mega-Squirt" EMS? : Ongoing MegaSquirt Discussion

[Enthalpy]

Not really. They don't have a very large dynamic range of airflow values. I think it's a 16X16 or perhaps 12X16.

Larger maps come in handy when you are making more power and want streetability and good manners. Plus, you can use the resolution to throw in things like spark timing to help idle control if you don't have an IAC valve, or cam timing control to help pass emissions NOx-wise (which is definitely not MS territory).

Accel DFI shouldn't take that long if you are starting from a known-good config. An experienced tuner starts with a map he has already done a lot of R&D on, and customizes it to the car in front of him. That way, things like injection phasing and throttle pump are almost-bang-on from the get-go.

[MuMan]

I think there's 20 load points per map, but I could be wrong, probably am. Ricky would know.
Interesting to note there's 2 locked maps in there, Gen 2 that is, one for 14psi and one for 16psi. Used to be a company in Ca I think it was, that would unlock the maps and install a user programable chip.

[mr220v]

Originally Posted by Enthalpy

I think the definition of "a little skill" is the debate here.

Have you ever setup MS on a brand new, never-been-done-before platform? I have. Twice. It's a ROYAL pain! I have my engineering degree, I'm proficient with electronics and I can tune very well. It took a LOT of tinkering with the thing to get it to so much as correctly interpret the triggers, and that was only after we were forced to shave some of the trigger teeth and rework the reluctor gap plus add a signal conditioner. After mapping it, we could only run a handful of the stock accessories due to the limited GPO setup it has.

I'm inclined to agree with Ricky.

Where you went wrong is trying to use the stock triggers. With the MS, you're supposed to weld the escort edis wheel to the end of the crankpulley and fab up a bracket for the pickup. It might be sloppy, prone to failure, and even kind of dangerous, but this seems like the way 90% of MS setups have been done. Who cares if it makes changing a drivebelt intresting.

I'd much rather pay a few hundred more dollars and pick up a used haltech which will actually support the multiwheel triggers with just a little bit of software configuration. I have yet to hear of an MS running the standard 3sgte type cam/crank position sensor without milling out a few teeth.

On the otherhand, the MS seems to work well with the older 4 tooth trigger on the original 4age's.

I guess the first question to ask if you're thinking about an MS is does it support my cam/crank sensor semi easily? The answer for the 3sgte is no.

If I was planning on turbocharging my escort, I wouldn't consider anything but an MS.

[Enthalpy]

The early 5Ses have a 4-wheel, but I have yet to hear of any success on MS. Perhaps nobody has tried.

I heard of the EDIS stuff, but the customer was dead-set on using the stock triggers. We gave up after playing with the antequated trigger interface on the MS and started grinding teeth.

[mr220v]

I only mentioned welding on the edis wheel because I consider it one of the "costs" of using an MS on certain engines. I would NEVER consider taking this approach myself.

[RickyB]

Originally Posted by TomsMR2

anyone happen to know how many cells the stock ecus use? i wouldnt imagine they have incredibly high resolution..

Gen2 ECUs have two main map sets. 16x15 for the high octane map and 16x13 for the low octane map. This is for both fuel and timing advance.

[RickyB]

Originally Posted by MuMan

Interesting to note there's 2 locked maps in there, Gen 2 that is, one for 14psi and one for 16psi. Used to be a company in Ca I think it was, that would unlock the maps and install a user programable chip.

This bit of misinformation was most likely started to sell ROM tunes and is still quoted as a "fact" by MR2 "literati." There are two map set available to the ECU at all times. One is only intended for use when the ECU believes that the fuel is of sufficient octane to run the more aggressive of the two maps. The high octane map set has 2 additional load columns for higher resolution at the top end. The ROM tunes effectively lock out the low octane map which probably doesn't sound as cool as saying that they unlock the high boost map which was always available as long as you ran what Toyota considered to be sufficiently high octane to use these maps safely.

Given that the load sensor on the gen2 ECU is the AFM, there is no notion of boost acting as the y index into the maps. What is being calculated from the AFM voltage signal is what we can best call "flow." It is basically the flapper voltage divided by RPMs. Flow is proportional to boost only if you keep VE fixed. When you start messing around with exhaust systems and turbos, VE changes, so the point on the maps which the ECU uses under a particular boost changes. To further complicate matters, the AFM flapper door completely opens once 210-220rwhp is reached. Beyond this point, the ECU is always referencing the same top row of the map regardless of how much more air flows into the motor. If you adjust the ROM maps to bring fuel and timing in line at this point, you also end up adjusting fuel and timing at lower part throttle settings. ROM tunes will not be all that they can be in the MR2 world until the AFM is replaced with a device that measures flow to at least the level or power that the motor is expenced to develop with said tune.

So how did we end up discussing the gen2 ECU in a thread about the MS? I've spent months pulling apart the stock ECU code and I can tell you that it is a case study into how tight you can write EFI code. Most of the MS code I've seen is pretty slow and inefficient compared to the stock ECU code.

[TomsMR2]

mildly off topic.. its good to have you here rickyb

[mr2greasemonkey]

It's pretty hard to beat the factory systems when it comes to engineering. More so the newer the vehicle. I like the fact that we are discussing all EMS' cause it gives us a comparison of more sophisticated aftermarket units vs. the standard MS vs. stock.

[Mekkahfire]

If toyota offered a full EMS, id buy it. But unfortunately... They dont.

[MuMan]

[quote=RickyB;168039]This bit of misinformation was most likely started to sell ROM tunes.

The myth goes on. I think the resurrection of the techtom has perpetuated the notion of the locked maps being available. I think it all started back in '97 with the SCCA project. Most people, unlike yourself, not knowing a whole lot about the inner workings of the stock ECU. It's one hell of a shame someone can't come up with a user programmable version of the ECU.
How much work has been done on the AFM..it appears to be logrithmic in operation, as in 75% of it's range appears to be directed towards the lower end of the scale, where it has higher resolution, tapering off to nothing around that 230whp level you mention. Has any attempt been made to re-engineer the variable resistance to a linear output, that you are aware of, to extend the useable range?

[Enthalpy]

I think you mean, "logarithmic."

Extending the usable range like that craps on the resolution at lower flowrates. I have seen the same types of things while playing with Subarus and really large hotwire MAFs. It simply doesn't provide the dynamic range to make the car smooth at all operating points.

[MuMan]

Knew there was a spell check somewhere
Well, yes it does, but you can't have it both ways. If your looking for upper level resolution, you have to get it from somewhere. I'm thinking the output can be 'log' in essence, but to a lesser degree than stock, in as much as resolution can be maintained in the lower scale, but escalating to a more linear scale.
Spell that ok?

[MuMan]

I'm not talking spring tension here, or even modifying the AFM itself for that matter (but it would be necessary). More a circuit (black box if you will) to modify the output voltage. We are talking 0-5v here right..it could be made infinitely variable in rate and amplitude.
Oh, well..back to MS

[RickyB]

Originally Posted by MuMan

I'm not talking spring tension here, or even modifying the AFM itself for that matter (but it would be necessary). More a circuit (black box if you will) to modify the output voltage. We are talking 0-5v here right..it could be made infinitely variable in rate and amplitude.

Anything that can be done with a circuit between the AFM and the ECU I can do by modifying the ECU ROM code. Although a little more power could be properly controlled by modifying the AFM itself, better control over a wider power range requires either a larger flow meter or switching to another method of fuel control (speed-density).

[MuMan]

What's ATS doing to get such high numbers with the stock AFM in place?
Any idea what the max flow of the AFM is, regardless of the flapper.

Could be an interesting place to start:

http://www.globaldensoproducts.com/e...flowmeter.html

[RickyB]

Originally Posted by MuMan

What's ATS doing to get such high numbers with the stock AFM in place?

Folks confuse the ability to make power at a given level with the ability to properly control all part throttle conditions below that level. Once you open the flap completely, you are accessing the ECU map's top row. Any changes you make to that row affect not only WOT, but also all part throttle opening points in which the flap is still completely open. That means that you have a sweet spot right around where the ECU was tuned to run and then overly lean and overly rich conditions above and below it. I've seen this while doing dyno pulls at various boost levels on ROM tuned cars that were brought to me for analysis.

[MuMan]

So with the ROM tune, there's no modification to the AFM, it's just maxing out early, and continuing to access the top row..interesting
Did you take a look at the in-line denso AFM.

[mrturrari]

Originally Posted by Enthalpy

The early 5Ses have a 4-wheel, but I have yet to hear of any success on MS. Perhaps nobody has tried.

That's actually backwards. The newer 5sfe's have the 4 trigger wheel and the older first gen ones have a setup similar to the 3sgte with the 24 trigger wheel. I'm working on a MS2 for a gen 2 5sfe but it's been on the back burner for a while.

I didn't see anyone mention of it but MS2 and MSS&S are both 12x12 maps. It's possible that with future revisions of the MS2 they will get bigger. Also once the new router board is complete you will be able to add on the ability to read multiple triggers, run fully sequential injection and fire multiple coils. It may take a while for them to work out all the bugs though so don't hold your breath just yet.

[toyotaspeed90]

Originally Posted by mr220v

I only mentioned welding on the edis wheel because I consider it one of the "costs" of using an MS on certain engines. I would NEVER consider taking this approach myself.

why? as long as it's properly placed and bracketed.... it's 100% reliable.....

there is 1 downside to welding the trigger wheel onto the end of a 16V pulley in an aw11..... but unless you know what the slight annoyance is, then i'll keep my mouth shut.....

don't get me wrong.... i'm reading the discussion and following along....

but mainly what I hear is people bashing the MS..... when in the end (at least locally) the 4ag based mr2's with MS seem to stand alone here in WA.... the best turbo mk1 in WA is (or at least was) on an MS1... yes, an original very basic ms1.....brentpicasso.... and i'm sure many of you have seen his builds and are awe-inspired by the overall job.....

and not to give myself a big head... but I haven't run across a faster s/c with a stock longblock regardless if it's stock, rebuilt, LSD, higher boost, etc.... and my MS hasn't even been dyno tuned, in which i'm positive I could gain an extra 10+whp and 10+wtq.... and to top it all off, I turn the s/c off and get better gas mileage on the highway than my 100% engine 86 n/a does....

lastly, to make you all cringe..... my sc is on 8x8 maps

I personally recommend MS to those enthusiasts who want to understand the system (yes, in case anything happens because it CAN fail) because it's been proven to perform.