Imagining over the phone debugging. Must be tons of fun.

is that not your nasty zener? i must be looking at the wrong die-ode.

and yeah, colored wires.

That's not a zener, it looks like a 1N4001 to me.

The one I left installed isn't pictured.

Basically when I first installed MS3, I was having a horrible problem with misses and erratic idle. Turns out my timing belt was too loose and since MS3 is faster and more accurate with timing, it was causing shit to be out of sync and missing. It was fine with MS-I and MS-II.

Before I realized this, I thought the extra board I built the MS3 around might have some issue. So I modded my old MS-I box, that wasn't running idle mod, for the MS3. When I moved the TIP120 over, I forgot to pull d8.

Sounds more like a sign of making power.

I don't think we tweaked your idle settings at all. In the first idle screen, raise MinDC and ClosedDC to 21 and see if that helps/


Here's the diode he's talking about:

http://img40.imagefra.me/img/img40/8...jm_7a5d489.jpg

The diagnonal one is going from the TIP120 that he installed as part of the IAC mod to S12, which says to me that it's a flyback diode.



Scott- I don't see any relay drivers on that board, yet you've done the idle mod. Standalone or not?

I don't have a shot of my current board in it's current state. but yes I'm fully standalone on MS-III and have no complaints...just still finding new features daily.

for example: the 4x4 idle VE table is awesome. I have my rpm/load at: 750,875,950,1000 & 32, 36, 43, 47 kPa. Able to keep afrs within a .5 point spread at idle with ease.

overrun has a "return fuel when X rpm" setting, default it was at 1100RPM, which was just too late to kick the injectors back on...finding that one made idle recovery significantly better.

yesterday I had to run: a/c, headlights, wipers, and my subwoofer. Noticed AFRs went hella rich at idle. Was able to adjust my injector dead time correction % curve since the voltage in that situation dropped from 13.8v to 11.8v...so I had to reduce the default curve a little.

It looks kind of wierd without all the circuits in there but look at all the space on that heat sink. Hubba hubba! Sorry about the shitty close up. Bad Riggy! Hopefully the sexy bootloader micro switch will make up for it!:giggle:

http://img705.imageshack.us/img705/1413/im000594f.jpg

Uploaded with ImageShack.us

http://img9.imageshack.us/img9/1163/im000597.jpg

Uploaded with ImageShack.us

Not bringing those drivers out to a separate ground line makes baby Jesus cry.

Looks nice, though. It really is amazing how little it takes to make an engine run.

I know, I was thinking the same thing. All the ground pins on the DB37 have 18 ga wire ran to the block so the ground plane is pretty solid. All the signals are pretty clean. So far there doesn't seem to be any extra signal noise but I have been pondering routing the driver ground out anyway, just for safety's sake. I bet it's sinking a lot of current. 4 amp peak, 2 amp hold, right? Alternating 4 injectors, would it be safe to say it's seeing 8-10 amps with residual voltage or is it cumlative and it's seeing 16 amps?

Thank you! Coming from you, it's quite a compliment.

I still can't believe that it runs on it either. At the same time, I can remember a time when it needed way less than that.

However, user interaction was severely limited and we were at the mercy of how well the mechanical parts behaved especially after they had aged a bit. If I had a choice, I think I would stick with the latter version anyway. No sense in going backward at this point. On that note, press on and sally forth all!

I want that stuff bad, but also it's more stuff I'll never tune properly :laugh:

Seriously though the Idle VE table is a big want for me

here's my MS3:

http://www.boostedmiata.com/gallery2...2/DSC_0112.JPG

http://www.boostedmiata.com/gallery2...2/DSC_0118.JPG

http://www.boostedmiata.com/gallery2...2/DSC_0120.JPG

Putting an MS3 on a Rev 3 board makes baby Jesus cry. Looks like an elephant fucking an octopus...

There's no difference in mods between MS-II and MS-III. just swap daugtherboards. But I have added an a/c input/output (spare 2&3), Fan control (spare 1), High side driver to run fuel pump (Spare 4). Then I have the opto circuit for my second trigger. And the typical Idle, boost, baro.

I didn't make this pretty cause it was my Ms-I board that I didn't want to mod to MS-II/III but I was trying to track down and issue and decided to try the other mainboard.

It sure runs well...so who cares. it'll look worse when i add the ms3x board and run full sequential.

There's just no sense keeping the 3.0 / 3.57 design alive any longer. Pretty much every single circuit on it has been obsoleted and replaced with a mod or a daughterboard. The only active part left on it at this point is the power supply regulator.

And yet we keep on using them.

Reminds me of the PC market eons ago, when you could buy, for instance, a little daughterboard that would allow you to plug a 486-SX processor into a motherboard designed for a 386-DX. Or ISA cards to add ATAPI controllers to machines that weren't born with them. Or expansion boards to hold more RAM chips (yes, chips) in addition to what the motherboard was socketed for. Or sidecar modules (now I'm really dating myself) for machines like the PCjr or the TI-99/4A, to add printer ports and whatnot.

Hell, expanding the RAM on the Amiga 1000 (from 512k to 1 meg) and upgrading to the FatAgnus controller involved a soldering iron and an X-acto knife. I think I still have the instructions for that mod (printed on greenbar tractor-fed fanfold paper) somewhere in a box, complete with ASCII-art illustrations.

Remember when shit like that was considered normal? Nowadays, we just chuck the motherboard (or the whole computer) and buy one that's actually correct for the application.

I'm slightly bitter and opinionated about this, in case you can't tell.

Building a Megasquirt for a "normal" car (one with crank and cam triggers, fully sequential injectors and ignition, ECU-controlled fans, a PWM idle valve, variable geometry intake, variable cam timing, etc) should require precisely zero jumpers run across the back of the board, zero transistors in the proto area, zero daughterboards, etc.

working with what i got. $50 spare mainboard + $190 CPU :)

I like the diypnp myself, i think theyll have a ms3 version soon.

I'm glad I went the mechanical route and not the computer engr. or electrical route. More than half of what you just said was Chinese to me.

Now I really feel old.

We had a lot of people asking for an upgrade path from existing MS1s and MS2s to MS3, and Bowling & Grippo wanted a quick proof of concept, which is why there's a setup like this with three PCBs in one box. The MS3 + MS3X is meant to give an option that works for most cars with a few jumper wires but no other components added. However, there are a few post-V3.57 designs out there (DIYPNP, eMS-Pro, arguably the Microsquirt) and more on the drawing board. Like I said, the current MS3 design is a proof of concept; it's not going to be the end of the line.

Still, I don't see the V3.0 and V3.57 going away anytime soon. Generally, MegaSquirt products may get updates (like the jump from V1.01 to V2.2, or the DIYPNP V1.1 to V1.5 jump) but if people still buy them, we'll keep making them.

I just want in on what the "x-connect" is going to be used for!