Started my MS3X build this week, questions to follow
 

I started my Megasquirt build this week. I have been reading for the last few months and decided the MS3x was what I wanted for my '99 nb. I figured I could start a thread to fill in some of the gaps for anybody else thinking about MS3x DIY.

I am still a little in the dark because it is hard to wrap your head around all of the variables associated with inputs and outputs without a finished MS in front of you. For now I am following Frank's (Westfield MX-5) amalgamated instructions on his blog. I am sure that once I have the built board in front of me and I am trying to make it work, some modification/repair will be necessary.

I ordered the parts listed on Frank's build info page. I didn't realize the price of the MS had been slowly creeping upwards. Still a great deal for the money.

Do I need the STIM? I have fancy multimeters, oscilloscope, and a variable power supply- is that enough to fake the signal for testing?

I also bought an Innovate Motorsports LC-1 w/gauge. I will buy tunerstudio when this thing is together.

Ran into my first problems though. I ordered the closed element IAT, which won't work for me. Also, the MS3 doesn't ship with any wire to make a harness, not a big deal but for those placing orders buy some wire while it is convenient. I am going to copy the Tyco connector in the box PNP. Bought the connector from onlinecomponents.

I am still on the first couple steps of building. I am sure I will run into problems soon enough. I'll post them here. The alternator circuit may be challenging to fit in the smallest space possible on the proto area. I will be working on it a couple hours a night after work so it should be done by the end of next week, hopefully.

One more question, on the instructions it says U5 should be mounted with a plastic screw, but later the MS3x gets mounted to this spot using a metal screw+standoff? Can anyone who has been there before shed some light on this?

https://www.miataturbo.net/attachmen...ine=1333061802

Update
 

I needed to check the power supply voltages without a Stim and it took a little while to piece everything together so here is a step by step for how to test your voltages.

Hook up 9V DC, with lowest amperage possible (<300ma). You only need voltage and too many amps can kill improperly installed components.

On the DB37 use pin 28 for +, use pin 19 for GND.

Check U1 pins for proper voltages, pin 1 is closest to C19, 40 is adjacent to 1.
GNDs 2/19/32
5V 1/20/31

Also check pin 16 against GND for ~9 volts to check jumper.

Don't worry about the plastic screw. You end up using the metal one for mounting the daughterboard.

If you're using the Expander for idle control, DO install the 1N4001 diode across the idle valve. IMO that shouldn't be optional.

Another update.
 

Thanks for the heads up on the daughterboard standoff. I ended up finding the directions for this but it was hard to tell from the pictures I found if the screw needed to be plastic or metal.

Here is a gratuitous photo of current progress, I couldn't find that many pictures of MS3/x boards in progress so I figured I would post what I have. Sorry for the skewed image. I used the optical magnifier to get a better focus for the lettering.

https://www.miataturbo.net/attachmen...ine=1333128601

First question
 

Can the ms3x control the a/c or do I need to build another circuit? I have searched but found mostly things about the ms1 & 2.


Progress so far:
https://www.miataturbo.net/attachmen...ine=1333237320

yes, it has feed forward code.

What is the difference between these two diagrams? I see the difference in wiring, but what is the effect of one version over the other? Is it just preference, or are they two different setups?

https://lh3.googleusercontent.com/-7...52520frank.jpg

https://lh6.googleusercontent.com/-I...00-ECU-MS3.jpg


Going on to the alternator circuit, here is the board as of now:
https://www.miataturbo.net/attachmen...ine=1333485553

The "Frank" one is sequential fuel and spark. The other one is just sequential fuel. Unless you replaced your coils you want the 2nd one.

Thanks
 

I see now what I was missing, I thought it might have been that but other pins had moved too.

Does anyone have a better photo of the alternator circuit in the Proto area? I will try to make it as small as possible, like Frank's, but for tracing everything out it would be helpful to see a picture. If not I will proceed to prototype board it and take some pictures.

the Frank one is actually not for publication. It is how I wired my own megasquirt. A couple of pins have moved because I have a euro car (clutch switch and fuel pump are wired different).
Use the other diagram. It has no sequential spark. If you want that, just use 2 free pins for the 2 other spark outputs.

Frank

Thanks for clarifying. The walkthrough so far is very easy to follow.

I have been banging my head against the wall with the alternator circuit. I found it only works up to <9 volts. I had to go through the whole circuit and found that I was given a 4.87k resistor instead of a 48.7k. At least it works, now I need to make another trip to the store. Anchor electronics is great if you need a source for parts around Santa Clara.

whoops!

Alternator Circuit
 

I made the switch to the 48.7k and it works! However, I get a shutoff at ~14.15V (Using lab bench power supply and Fluke MM).

Is this too low? I read the spec as <14.4 but I think that must be what the OEM setting?

I have more resistors to try and see which gets me closest to 14.4V. I think the 1% range is working against me right now, I'm off by 1.7%.

Here is my roughed out proto-board.

https://www.miataturbo.net/attachmen...ine=1334013414

That's good. I use all 1% error and it switches at 14.2v

Any helpful suggestions for cramming the circuit in the proto area? Is this: http://lh4.ggpht.com/-8IMGSlUU_5I/S4...jpg?imgmax=912
a pin by pin guide?

Yup, that's how you build it in the proto area (if you can figure it out ;)).
Red are connections. Blue and black are parts.
All resistors are standing up (not enough room otherwise).

I hate building it that way. I much prefer the pinout i use for the diypnp proto area.

Do you have a link to your version? I am open to other options, I even considered using a separate prototyping board.

Like this:http://www.robotshop.com/prototyping...m_campaign=jos

Also, any known problems using a lab power supply to power the unit while I update the firmware? In theory it should be the same as the Stim, minus the inputs. Just use the MS3 DB37 power pin and grounds?

Thin gray lines are copper in the board on DIYPNP boards? The MS PBC I have doesn't have continuity between pins.

Thick gray is jumpered wire? Mustard lines are tails of resistors (confused by the line between resistors 30k and 3k?)?

Am I just reading into it, and this would be a breadboard style layout and it still needs to be adapted to the proto area?

gray lines are traces between pins.

thick gray are jumper wires.

orange are tails.

and then the orange line in question is just another jumper with a piece of wire.

I managed to get the circuit soldered up, but didn't get a chance to run the wires to finish it. I will have to wait until tomorrow.

The circuit was easier to solder up on the proto area than to layout on the Breadboard. I really figured out the circuit through working on it, though.

Another question, 1N4001 needs to terminate in Pin 9 on the MS3X, with the other end to a 12v source. What 12V source? So it goes 12V source (on MS3X?) to pin 9 and then pin 9 to the Tyco connector?

Also, Frank is pretty adamant about grounding wires, do most people run their grounding wires outside of the harness to the engine or just use the harness? I know his application is not a Miata, so he may be doing it a bit differently.

I am not looking forward to soldering the wires to the TYCO connector next. It is a pain to hold the wire/solder/iron in just the right place.

Pictures:
https://www.miataturbo.net/attachmen...ine=1334190873
https://www.miataturbo.net/attachmen...ine=1334190873

Yeah, once you figure it out, it's really not that hard. Any 12V source will do actually, but as it's a flyback diode (high currents possible) I felt it safer to solder it on the Tyco connector, not on the board itself. I soldered the diode directly between Tyco pin 1B and Tyco pin 30 (see the 'Frank' diagram). I used the stock harness. The seperate sensor ground is already there on a miata and there are multiple seperate ground wires to the engine, so a stock miata harness is good enough as is.

Thanks for the answer. I kind of researched it further and got my answer, but you made it totally clear. The DB37 on my MS3x card was a pain to get out. I think some of the solder had the flux baked out of it and would not reflow. I had to add solder back on and try to suck it out, which took forever.

I made some progress, here are some photos. I keep posting photos to hopefully help others have fewer questions when they decide to build a(n) MS3.

https://www.miataturbo.net/attachmen...ine=1334506692
https://www.miataturbo.net/attachmen...ine=1334506692

looking good so far.
A tip, when you start soldering wires to the tyco connector, start with the bottom row, not the top row (bottom row will be blocked off). And another tip. Don't bother building the low batt warning light. Use a spare output to activate the low batt light, like when Vbatt is lower then 12V or higher than 15.5V or so.

tin the connector pin then tin the wire. hold the iron on the pin, get the solder wet, then touch the wire onto it and have it all melt together....release iron, hold wire in place.

that will make sure you get the best bond, those pins dont like to be soldered poorly.

Good point. I will go left to right (right handed) on the bottom row and then same for the top.

Thanks for letting me know about using a spare output for Vbatt (is pin 20 okay, or are there problems with using injector pins?). I also need to setup another input/output for my A/C. Do you know what most people use? I was thinking Nitrous? Pins 25 (output) and 29 (input).

Just have to make sure the amperage is adequate and change the software to have that pin affect the MS? This is a basic question that isn't exactly answered in one place- All of the inputs are either switched 12V or a setup to monitor a sensor (ADC?)?

bumping this thread

I am am at the same spot as you in trying to figure out the a/c inputs and outputs. I understand which pins on the ms3/x i need to be using I just do not know which pins on the tyco ecu connector I am supposed to be connecting those to. Braineack said that the a/c switch switched to ground when turned on, so would it be wrong to assume that I can just heck continuity of all the tyco pins with ground and find the one that matches when the a/c switch is pressed?

Other than that my car is currently running with Frank's build and with the alternator circuit, I just need a/c because it is getting hot QUICK here in AZ!

1P is your a/c switch into, that you'll bring into the ms3x.

then youll connect 1S, 1I to a spare INJ output on the ms3x to trigger the a/c compressor and a/c fan relays.

okay so i am really confused now. going off of what you said and what other people have said in this thread this is what i did. I jumpered the a/c fan 1I with the normal fan and then i connected the nitrous in to 1p and my nitrous 2 output into 1s. With this configuration the ac turns on at idle only when the a/c switch is off or when the fans are set to 0. When driving however the a/c stay permanently on. I am not sure where I went wrong here

there's no reason to jump the fans together. I dont like it.

but sounds like you failed at configuring the parameters correctly.

Hmm well i guess I will unjumper that then. What settings would you use assuming that I used the pins you specified? Sorry if this is a stupid problem, this has really been the only part of the build that has gotten me stumped though.

I had to take a break while on a trip in the UK. I tried to finish before I went, but had to stop with 10 wires on the harness left to go (at least they are the easy ones). I will post an update later.

Any updates on the above question about A/C? A tiny bit of spoon feeding may help having to take the megasquirt back out one more time when I am getting it fitted. I understand it all in theory, but reality seems much simpler.

nitrous in is probably not the best input, as it's looking for a 12v input.... I use launch in for my a/c trigger.

Long Overdue Update
 

I have finished the harness, the picture below is before soldering the MS3x. No real roadblocks, just a couple long nights with the car's wiring diagram. I know I can just follow the guide step by step, yes- it is that good for getting a car running, but I wanted to know what I was actually doing. The case will be a tight fit with the amount of wire I used.

I wired the flyback diode (3O) and IAC (3M) to pin 2 on Q12, that should not be a problem, or am I missing something?

I labeled everything. I would recommend it to anyone, it is a pain and it makes everything take longer, but the result is a harness that is easily modified and double checked.

A/C setup: 1S to MS3x pin 4 (INJ F), 1I to MS3x pin 7 (INJ E), and 1P to MS3x pin 11 (Launch-in). It seems that there are very few Switched Ground inputs, so I will go without launch control.

Now, I try to get the case modified and plugged into the car as soon as possible.

https://www.miataturbo.net/attachmen...ine=1337212283

I was reading in another thread about the flyback diode.

Is my placement okay, in experience? In theory it should be fine. The diode is in the thicker portion of heatshrink wrap on the TYCO connector.

I have been too busy to get this thing in the car. Hopefully I get some free time this weekend. I tried to update it with a bench power supply but gave up. I will do it in the car. Still need to unplug the coils, right?

5/32" vacuum hose for the MAP sensor? Is silicon hose worth it for <16PSI max boost? About 10' (3m) enough for a '99? Anyone recommend a vacuum tee for coming off the back/top of the intake manifold (or at least the size hose coming off the manifold)?

Great write-up so far!

Since it looks like I will be going that exact same route in a few weeks, I really appreciate not having to make ALL the mistakes myself, but having somebody make some of them for me first.

Which wiring diagram did you end up using, or did you make one of your own?

Just go to your local parts store and get regular old vac tee and 5/32 rubber vac hose, comes in a box of 6ft or so... plenty for a miata.

Big update
 

Alright, I finally got some time to get this thing in order. I had a little home improvement project get in the way.

I had to install an LC-1, new header, and midpipe. I also extended my narrowband rear wiring. I have to leave the car able to run on its own ECU for CA smog.

Tips on grounding the LC-1:

When you pull your vacuum line through the firewall, tape 6 wires to it. Leave 3 long enough to bolt to the intake manifold for wideband grounding, and the other 3 so you don't have to do it again when you do sequential spark.

Buy good electrical tape and the plastic wire shielding to make your own protective loom. Take the time to solder and shrink wrap. Do a good enough job so that it looks better than factory.

After all of this, I finally was ready to install the MS3. First I had to find how to unplug the coils. I searched using every combination of words I could think of and came up with nothing. I knew about the plugs on the coil, but there had to be another plug to the harness. I looked, followed, and checked the wiring diagram. I needed the brn/yellow,yellow,black/white,black wired connector is what I discovered and followed it to in front of the throttle body next to the valve cover. Right of the injector plug. Hopefully, someone with an NB will be able to use this info and save themselves 30 minutes to an hour.

I also discovered that my plug wires had corroded to the coil plug (any solution to this?) Time for new wires.

I finally plugged in the MS3x opened tuner studio and it found it! I loaded the Beta 24 firmware. Changed some conflicts and everything was reading as it should with the car off. TPS/CLT/IAT/MAP all reading what they should. I loaded Franks base map, too.

Tonight I hope to go through the settings and check everything, with the goal to start it tonight.
Are there any changes to the settings between MS2 and MS3 as far as setup on an NB?

Anyone have a base tune that is better than Frank's?

Do I need to verify base timing still, even using the stock cam sensor???

Redo check the trigger angle since the input methods are different. What year car I have a base tune u could work from.

I have a '99 with gm iat and wideband. No boost yet.

So tunerstudio does not let you have 2 a/c idleup outputs. I will need to go back and jumper a/c fan and a/c clutch, unless there is another way. I would like to have the option to control the a/c fan separate from the a/c clutch. I looked through the generic settings for pins and couldn't find a condition that would turn on the a/c fan when the a/c ground was pressed. Any ideas?

Also, what is the best way to control vics? Use generic pin control like the diypnp does or mess around with the vvt map?

Should I use the diypnp settings for spark/ignition or are there new settings that work better?

Use the fan control output for the a/c fan; have it activate with the a/c activation and have it turn on with temperature about 10*F above that of the main fan.

Then use a generic output (INJ E-F-G-H) as the main fan control, giving it a simple method of: turn on above 200*F, turn off after 190*.

For VICS, use another generic output (INJ E-F-G-H). Have it activate below 5250RPM and off above it.

What coils do you have?

I will look further into the fan control settings. The regular fan is already setup as above. I couldn't find a condition for a/c. Should I use the pin description instead?

I think you meant the opposite of what you posted on Vics. It is always on but you want to activate the solenoid above 5200 but I know what you mean.

I have stock coils running in wasted spark.

okay, so you have the main and a/c fan outputs running off a spare INJ output right? so setup the software like I suggested.

just make it so the output for the main fan is controlled using the generic ouputs screen, and the a/c fan is controlled using the fan control screen. that way the a/c fan can activate with a/c and also by temp. the main fan only needs to run with temp...it's possible you can add a condition to allow it to activate with a/c as well, haven't really searched through the options in the dropdown. I have a 91 so the a/c fan runs off hte a/c relay.

no, for vics the solenoid should be powered below 5250, and off above.

stock diypnp spark/ignition settings should be fine, they are the same outputs on the ms3x.

I got the A/C fan setup properly. The method above worked great.

I read some more about VICS. It seems like DIYPNP doesn't really know what the solenoid is looking for in their description of settings. I set it up brain's way and I will check for vacuum at idle to verify.

I setup my spark settings like the DIYPNP instructions specify.

I attempted to start last night and discovered I had a low battery from leaving the key on while changing settings. It cranked but it would start to fire, vibrate and die, and then the starter would stop turning. I took a log and I will try to post it here. All of my sensors/inputs are reading correctly.

I do not know if my ignition settings are right based on anecdotal evidence I have read. Are these settings right? : http://www.diyautotune.com/diypnp/ap...0-18bp-mt.html

Am I missing anything?

I put the stock ecu back in and it started as soon as I turned the key. Not a coil/plug problem. Prob not fuel if the cylinders were already primed enough to fire as soon as the key turned.

When I try to start I get very little exhaust that exits the tail pipe smelling very rich.

I am sure it is a setting but I can't seem to find complete settings for a 99 that haven't been modified to run the older boards or the car isn't running and that's why the settings are posted.

Brain do you have a working msq I could start from, or at least look at?

I posted a msq in Omni's thread. You can look at the DIYPNP 99-00 msq for settings for a 99. It's all pretty much the same though.

for VICS do:

enabled 1 | 0
RPM > 5250 100



did you hook the injector order hooked up correctly?

Otherwise only thing I notice on your msq is the need to change skip pulses to 3 from 1. I'd also put lag factors all to 90, but that's trivel right now.

I am not sure what you mean by hooked up correctly, but I have my injection order in tunerstudio set up properly. Do you think that maybe my settings point injectors to the wrong pins?

I'll look for Omni's thread.

My next step is to verify spark and check timing during cranking.

Do I use rising or falling edge?

but are they wired in the correct order? is Inj A wired to cylinder 1, b - 3, c - 4, d - 2?

should be rising IIRC, you wouldn't sync otherwise. crank for 10 secs while watching the compostie logger and you can easily see how it's reading hte sensors.

did you adjust the 4 pots correctly?

My injectors are wired up like this: https://www.miataturbo.net/attachmen...ine=1339781822

I will make sure it is rising.

The 4 pots have been adjusted, they fell in the range. I used Frank's instructions.

The car wants to start but it gives up before it gets steady combustion. The starter turns for 3-5 seconds, it coughs one to three combustion events, shudders badly- almost like it backfires and stops.

spark A and B flipped?

are you watching the sync indicator? took a composite log? setting those pots is a PITA, I assume you used a voltmeter to set them and not just the turn counts. Also, I found that the points he wanted you to measure in the write up weren't quite right. I spent some time googling figuring out where to measure to set the zero crossing points.

I've never had an issue just turning them like suggested in the msextra.com/docs/

I used a multimeter to set the voltages.

I followed the wiring diagram for all of the important stuff and double checked my wires before I closed everything up. I need to open it again to check where I wired my lowbatt light, so I will check everything again. It would be nice if it were so easy.

I have not done a composite log yet. I only took a regular datalog, I will though.

What is the best way to check my spark output? I was going to check for spark at the plug, but is there a smarter way to determine which spark output is firing (to eliminate a wiring problem).

You can use the output test mode to check spark, you can specify one or several or all of the spark outputs to run. Edit, yeah you do it at the plugs, but its easy enough to take them out then start the loop and make sure you've got the outputs going to the right plug.

swap the two outputs (or plugs 1/4 with 2/3) and see if it runs?

It is running now. Needed an adjustment to 5.5* degrees timing and rising edge.

I checked a/c and fans, everything works perfectly.

Now I need to tune my idle, I will also compare Brain's map to what I have. Idles between 10.5-11.2* timing, 14.4-15.0 afr, mostly around 14.7. Idle rpm is 1100-1200. Needs work but it runs.

Thanks for the help, any recommendations for idle tuning on the beta firmware?

So I tried to use some of the settings in the map brain posted in another thread. It wouldn't run/idle. I changed it back, but my idle is still lumpy between 1000-1100. Afr of 14.4. Does this sound like a fuel problem?

It has trouble starting still, requiring throttle and a lot of cranking. Starts great when it is warm.

Can anyone share their idle settings for the beta 24 firmware? Too many options and dead end threads/articles online.

Okay, the difference a day makes, I read up on how to control the rpm at idle and got it to smooth out a little and drop to 880, varies between 860 and 920. I'm aiming for 750 ish.

Edit: my idle target based on warmup is 750.

I adjusted the PID settings according to what they should do but I have my min PID duty set to like 15.6% and p=6 and I = 65. D = 0. Does this sound right? I found that the idle didn't vary with changing the setting and required a full power cycle to settle. Any recommendations on how to get that sorted?

Went for a drive and did some auto tuning. Car is so much smoother and stronger than before, the stock ecu always felt a little lumpy. No excess vibrations through the shifter. I am more impressed than I thought I'd be.