NA6, MS1 + Stock ECU - AFM.....How?
 

Hi all

Firstly, my apologies for asking what I'm about to ask as I'm sure it was asked many times before, unfortunately I couldn't find the answer I was looking for, maybe I was using the wrong term, so here goes....

I've been building a MS1 (V3.0) to run in parallel with the Stock ECU on my Jap delivered (import) 1989 NA 1.6, the idea is the MS will take care of fuel and Ignition while the stock ECU looks after everything else (e.g Idle, A/C, P/S etc).

http://img8.imageshack.us/img8/9591/img5599v.jpg
http://img85.imageshack.us/img85/4033/img5600a.jpg
http://img21.imageshack.us/img21/5424/img5601w.jpg

Q1. What are the steps to remove the AFM in a parallel setup where the stock ECU is still connected?
What I know so far is that I need to wire up a seperate IAT onto pin 1&6 like so...
http://img44.imageshack.us/img44/8797/image022fy.jpg
(MegaSquirtPNP by DIYAutoTune.com)
Also run a tap from the light green fuel pump relay wire to pin#37 on DB37 connector. Are these mods enough to fool the stock ECU into thinking it's still got signal or will it go on limp mode.

Q2. I'm planing on sharing the stock CLT sensor, I've been told to simply remove R7 on the 3.0 PCB and leave R4 the as is, are there calibration settings I need to make in Megatune? if so what are they?

Q3. Can the TPS signal be shared with the stock ECU? Does the MS even need it to function as it's just a switch. Eventually when I do upgrade to a variable TPS, will there be issue with sharing the sensor? as the MS has taken control of fuel and ignition, will the stock ECU still needs TPS to sing along?

Thanks all for helping out, the forum has been great a source particularly the sticky write-ups.

Cheers
M

A1. You should remove R4 along with R7 and flash it with the correct easytherm values (seen in my how to thread or the wikidot page.)

A2. see above.

A3. The stock ecu does not need to see it, in it's current state, the MS cannot use it. Once you fit the vtps, just let the MS see it. I bring it in through the stock wiring and cut the signal to the oem ecu.


everything else one the build looks good, however it looks like you used a 2490ohm resistor from on he D1 jumper...it needs to be 470ohm

PS. make sure under spark settings, spark inverted is set to YES.

Thanks Braineack, awesome writeup on Miata Turbo FAQ: Installing The Software btw:bigtu:. Using the info there I removed the PWM circuit for low-Z injectors in order to run HiRes code later down the track.

Last night (before any of the above mods were done), I feed MS with Innovate LC-1 Wide-band signal, MegaTune suggest that sensor output be programmed to 0V=10:1AFR, 5V=20.0:1AFR to give better resolution, I did that but soon notice that Megatune Configurator seems to say otherwise...

00057 # set INNOVATE_LC1_DEFAULT "Innovate LC-1 default, 0-5v = 0.5-1.5 lambda"

Do I need to let MegaTune know somewhere that I've reprogrammed the sensor so that it doesn't think my sensor output "0-5v = 0.5-1.5", currently during idle the AFR gauge in MegaTune swings from 10-2 o'clock it should be abit more stable than that.

Ignore all that, I've found what I should set the sensor to...
00055 # set INNOVATE_0_5_LINEAR "Innovate, PLX 0-5V 10-20:1 AFR"


Q1. "1. Create your Firmware file (only necessary if you're running parallel)." Is this a different step or simply saying what the below steps are essentially, if it's a totally different step then how?

Q2. Link (http://www.megamanual.com/files/soft...ll_Install.zip) to EasyTherm is broken, I found and downloaded EasyTherm V5.0.5 this version ok?

Q3. "3. Change the 'Code Version' (towards the bottom on the RHS of the screen) to "Custom"" Can't find this reference in the copy of EasyTherm above.

Q4. "Piggyback without R7 or R4 (recommended for piggyback sharing the sensors): Use 2625 for CLT and 3475 for AIT for the bias."

I've remove R7 as well as R4 as per your earlier suggestion; sharing sensor.
Currently all the sensors are stock; stock CLT temp, stock IAT inside the AFM, eventually what I'll do is remove the AFM but reuse this sensor or should I use the GM open element IAT (which I've already bought) instead?

Have I interpret the instructions correctly with the below settings...
http://img21.imageshack.us/img21/719...90924at115.png
http://img21.imageshack.us/img21/611...0924at115n.png

Q5. What's the function of the "ST SIG" fuse? MSPNP requires this remove, do I need to do the same for my parallel MS1 install?
http://img17.imageshack.us/img17/2728/image002fy.jpg
MegaSquirtPNP by DIYAutoTune.com


Here's how it's looking atm...
http://img35.imageshack.us/img35/1917/img5612ma.jpg
http://img35.imageshack.us/img35/7360/img5619i.jpg

many hours after...and the MS is installed and running, currently taking control over Injectors (Squirt) and Ignition (Spark). On the road it's ok but at time quite rough at idle, especially after the cooling fan had turn off after a blow idle will surge to 2k rpm then dips to almost stall (at times it did stall). At it's best it's still dips and surge by a 100 rpm either side of idle.

I'm starting to think that this parallel setup might not be as good as it looks on paper, this surging idle is making me think the MS should also control IAC (idle) but if it controls idle then it'll need to cater for A/C as well...wonder if Power Steering needs any ECU control, I recall seeing a P/S wire going to the ECU.

Maybe I'm missing something or not doing something right...ideas?

I can't personally speak to A/C (don't have it) but you're correct in that both A/C and P/S influence the idle control in the stock config.

P/S doesn't really need to. I can turn the wheel from lock to lock at a standstill without causing more than a 100 RPM or so deviation in idle.

A/C, OTOH, seems to be a problem for some folks. I still firmly believe that this can be solved with nothing more than an adequately sized bypass valve (such as the valve used by FM for EBC, or even the stock canister purge valve) slaved to the A/C clutch circuit, though I have no means of testing this myself.


Normally, this line delivers +12v to the fuel pump relay (circuit opening relay) to ensure that the fuel pump runs during cranking even if the AFM isn't registering airflow. It also goes to the stock ECU to signal to it that the starter is turning, so the ECU will be in the correct mode for starting (retard timing, inject fuel according to starting table).

In the MSPnP, they hijack this latter wire (the one going to the ECU) and use it to send power to the fuel pump relay so that it will operate even with the AFM removed. The fuse is removed to protect the MSPnP from seeing +12v coming in on this output.

In a regular MS configuration, you do not need to remove the fuse, since you are (presumably) not driving the fuel pump relay in the same fashion.

I wonder how the MSPnP control A/C, why can't similar technique be replicated? MSPnP makes no engine bay modification to control A/C and it controls it quite well.

So this means when I remove the AFM, I can also hijack this line to drive the FP without needing to run a seperate tap off the FP relay...Noice!!! I'd remove ST_SIGN fuse of course. I think I stumble across a guide for this somewhere.

Does the MSPnP compensate for the fuel pump safety feature when the AFM is removed? if so where do find the details for the mod?

The MSPnP does not do anything special for A/C in terms of idle. All they do is pass along the signal to engage the clutch and fans when the A/C control switch is depressed. Functionally, the MSPnP's A/C control is identical to the modification that I posted. This is one reason why I find it difficult to accept that people have such trouble with getting idle to work properly with non-PnP Megasquirts.



If you want to build a circuit to send +12 down that wire, controlled by the MS's "FP-1" circuit, then yes, that will work just fine.

The Megasquirt has a built-in fuel pump relay driver, which is active during cranking and running, but inactive when no cam/crank trigger inputs are detected for a period of time. On page 5 of the Rev 3.0 schematic, this is the circuit controlled by Q2.

In the usual configuration, this circuit provides a closure to ground. That is appropriate for the '94 and later cars (as well as most other modern automobiles), and also works on the '90-'93 if you run a wire over to the light green lead at the fuel pump relay. The reason DIY does it the way they do on the MSPnP9093 is because this light green wire (which normally goes to the AFM) is not available at the ECU connector, and they wanted to avoid forcing people to run a wire across the underside of the dashboard.

Cattering for idle at A/C depress is something that should be able do with the code shouldn't it, something like...

"If A/C=On
Fast Idle = On
A/C Fan = On
Delay = 1000ms ' give the engine time to raise it's rpm before the A/C clutch is engaged?
A/C Clutch = On
Else
Fast Idle = Off
A/C Fan = Off
End If
"
Not quite the programing language, but something along those lines?

Sounds like sending a line to the green relay cable is an easier way :).

A/C activation isn't handled in any way by the megasquirt. There might be a neat little circuit in the MS' case, but that's due to a matter of convenience. All you need do is translate the high resistance ground provided when the a/c command button is activated into a low resistance ground to trip the a/c comp clutch and cooling fan relays. This can be accomplished with 2 diodes and 2 resistors.

I really don't understand the big deal about it. Get your fuel table, spark table, and idle settings nailed, and the car will idle great with the a/c activated or not. Add the variable TPS, and the MS will open the idle valve proactively and eliminate idle sagging when coming to a stop.

If you feel incessant about running a/c "through" the MS, perhaps bring a lead parallel with the a/c clutch relay ground into table switch. Create secondary fuel and spark tables with a tad more fuel and a tad less advance at idle.

Sounds like unnecessary work to me, but what do I know. :2cents:

Now if you're running a MS2, and if you have the output to spare, it might make sense to run a/c through the processor as you can set up output activation based on multiple conditions, IE set up an RPM window where the compressor is allowed to run, so you could deactivate the comp clutch if you went above and/or below set rpm set points. I don't have the output to spare on mine, but the 1.6 cars don't have VICS, so there's an extra output right there.

Good points there Ben, I should give it a bit more play before jumping into conclusion, after all the MS was only install yesterday (Aussie time), still plenty of settings here and there to learn about, if all else fail, I might need to post up my mqs for expert critics :)

Actually I wonder if any is running a similar setup to mine and has got it running nicely, wouldn't mind looking at your MSQ; parallel, MS controls Spark + Squirt, stock controls idle and the rest. It might just be a few setting that I haven't got right.

The same functionality can be achieved with MS1 by adding one relay, such that the A/C clutch disengages at WOT (like in the stock configuration) or at a certain RPM. This is documented in my writeup.



Yes, it should. It's been discussed, and nobody has ever bothered to implement it.

Yes I know, but my point was that MS2 can be set up for 2 conditions. You could allow the compressor clutch to engage at all times (assuming the dash button is depressed) except for two programmable conditions such as RPM<700 OR MAP> 140kPa. Makes more sense to me to do it like that.

Since you're already building a system and have room in the case, wouldn't this diagram make more sense?

It's slated for a future version of MS2/Extra now.

Looking forward to whats to come Matt :D
...maybe even include a self learning feature....
https://www.miataturbo.net/forum/t39522/#post459060

That's a cool idea Ben, abit like an AND gate which outputs true when both conditions are met. But being a clever ECU like the MS, it really should be able to control these things in the software rather than hardware.

This morning the car wouldn't start:confused:, it cranks...keeps on cranking but no start, did the fuel pump (FP) to GND jumper at the diagnostic port and confirmed fuel pump working but still no start...either no spark or no injection?...abit strange because it did start the night before (with abit of struggle but it did, plus the car was warm at that time)...maybe I've mucked up my settings somehow :(, trouble is there's so many settings, so much to learn and I barely understand what half of them do :(.

Here's my current MSQ, would be awesome if you could point me the right direction.