Megasquirt Update, with PICS Options

[lapuwali]

The DB-37 is simply the wrong part for the job. The newer Microsquirt uses the AMPSeal connectors, which are a far better part for this use.

I haven't decided exactly on the course I'll follow, yet. What's settled is I won't be using the MS case, but another, larger case. The through connector for that case will be a circular plastic multi-pin connector from DigiKey. This is available with a variety of pin sizes, including 18g - 20g, and each pin can carry 19A. I'll either connect this to the MS board by direct soldering of the wires to the board (no DB-37 connector at all), or by using a short cable with a DB37 on one end. This cable would be clamped into place, so vibration shouldn't be too big an issue, so a solder cup connector may be used (mostly because I have them already).

I intend to be lazy and just buy a v3.0 board pre-assembled from DIY or RS-Autosport, in which case I'll probably go the DB-37 route, only to save myself the hassle of de-soldering the DB-37 from the board first.

I'm also NOT going to be using the zillion ground wires run through the DB-37 all the way out to the main connector, but will be joining them to one screw post inside the big case, which will then run to a single 14g ground wire to be run to the big case (metal, so it should provide shielding) and then to the engine case.

The MS board will be mounted via rubber mounts inside the large case.

Right now, the "large case" is most likely to be a gutted D-Jet ECU case. Lots of room in there, and it seems to be adequately water resistant.

[toon1]

QUOTE (lapuwali @ Mar 29 2006, 09:50 AM)

The DB-37 is simply the wrong part for the job. The newer Microsquirt uses the AMPSeal connectors, which are a far better part for this use. I haven't decided exactly on the course I'll follow, yet. What's settled is I won't be using the MS case, but another, larger case. The through connector for that case will be a circular plastic multi-pin connector from DigiKey. This is available with a variety of pin sizes, including 18g - 20g, and each pin can carry 19A. I'll either connect this to the MS board by direct soldering of the wires to the board (no DB-37 connector at all), or by using a short cable with a DB37 on one end. This cable would be clamped into place, so vibration shouldn't be too big an issue, so a solder cup connector may be used (mostly because I have them already). I intend to be lazy and just buy a v3.0 board pre-assembled from DIY or RS-Autosport, in which case I'll probably go the DB-37 route, only to save myself the hassle of de-soldering the DB-37 from the board first. I'm also NOT going to be using the zillion ground wires run through the DB-37 all the way out to the main connector, but will be joining them to one screw post inside the big case, which will then run to a single 14g ground wire to be run to the big case (metal, so it should provide shielding) and then to the engine case. The MS board will be mounted via rubber mounts inside the large case. Right now, the "large case" is most likely to be a gutted D-Jet ECU case. Lots of room in there, and it seems to be adequately water resistant.

While thinking about doing this myself I also had the same idea of putting everything in the d-jet case. Hey why not there is a mount there, there is lots of room for both boards and noextra labor in running the wiring through the F/W.

I was also thinking you could probably insulate the rest of the box with foam from vibration and heat.

[yarin]

In the event that I need to rewire a DB-37 connector or a pin pulls out I dont want to have to cut off the entire connector as would be the case if i filled the shell with epoxy.

I have a nice solid AMP metal shell for the connector which adds a level of strain relief.

My relay box is a Hammond 1554V2GYCL Enclosure with a clear polycarbonate watertight light. Too bad they don't make it in black. However I did punch a hole in the side of the box for the relay cable to pass through. So no it is no longer water tight, but it is water resistant to a certain degree. If I had to do it again I would have purchased a panel mount heavy duty connector from digikey and hard wire it to the relay box. If I have connection problems on the relay box side that is going to be my solution.

I used rubber grommets between the relay box and the mounting brackets. It floats nicely. Same thing goes for the megasquirt brain.

(IMG:http://www.newark.com/productimages/standard/98H1102-20.jpg)

Crimping 20ga wire isn't a problem if you have the right tooling. Its expensive ($100+), but I know its done right. The comforting parts about the use of a DB-37 connector is the parallel injector and ground lines. If one pin breaks you won't know it assuming the remaining pin(s) can handle the current draw.

[jsteele22]

QUOTE (lapuwali @ Mar 29 2006, 10:50 AM)

I'm also NOT going to be using the zillion ground wires run through the DB-37 all the way out to the main connector, but will be joining them to one screw post inside the big case, which will then run to a single 14g ground wire to be run to the big case (metal, so it should provide shielding) and then to the engine case.

A zillion might be a few too many, but you definitely don't want to skimp on ground wires in the main cable. For purely DC currents, it really doesn't matter where the ground current runs, as long as it makes it where it needs to go. But for anything with high frequency AC components (including on/off pulses, like the ones to the injectors)) you need to have a ground path really close to the signal path.

Lets say the current in one of the injector lines changes abrubtly. This is going to cause an abrubt change in the magnetic field circling around that wire. Consider anonther wire right next to it : Lenz's Law tells us that an EMF is going to be induced in that nearby wire to try and negate the change in the local magnetic field. If that nearby wire is a ground wire, it's no big deal - a little current flows along it, and at either end of the wire the current spreads out into the chassis and becomes insignificant. But if the nearby wire is a signal (as are most of the other wires in that cable), then the induced EMF is going to be superimposed on that signal.

It might be that it won't matter too much, since the other signals are things that should vary slowly compared to the injector firing frequency, but I don't know if MS does any low-pass filtering on them or not. Safest approach, IMO, is to keep the ground wires.

[lapuwali]

Signal wires and injector wires won't be in the same cable. I'll be using four connectors on the big box. One for signals, one for injectors, one for the serial cable, and one for "power" (both B+ and switched +12). I'll shield the signals cable.

[brant]

I want to see a pic of the LC-1

Nice Job!

brant

[yarin]

QUOTE (brant @ Mar 29 2006, 12:33 PM)

I want to see a pic of the LC-1 Nice Job! brant

I could take a pic of my LC-1 but that won't help you very much. Right now its sitting on the floor under the car with the wires running through the engine compartment. I just had a bung welded on before the muffler (entry-exit left to right) so over the next few days i'll bolt it back on and figure out where to mount the LC-1.

(IMG:http://www.914world.com/bbs2/html/emoticons/beer.gif)

[Hydra]

Sorry for the (IMG:http://www.914world.com/bbs2/html/emoticons/hijacked.gif)
but last time i checked (4 months ago), everyone on the MS forums agreed that if i wanted FI and Ignition from MS, MS1 was the way to go, thanks to megasquirt'n spark (sp?) and that even though MS2 offered much more options and was much more powerful, the MS1 setup for ignition was well honed, was more widely used, and MS2 needed an extra "couple" of month before it could deploy its full potential...

My question to you megasquirters is: Is that still the case right now? cuz i'm planning on placing my order but i don't want to get my time invested in MS1, assembling it and tuning it, and finding out that it would have been much easier and much more efficient with MS2...
Thanks
Nick

[brant]

QUOTE (yarin @ Mar 29 2006, 02:11 PM)

QUOTE (brant @ Mar 29 2006, 12:33 PM) I want to see a pic of the LC-1 Nice Job! brant I could take a pic of my LC-1 but that won't help you very much. Right now its sitting on the floor under the car with the wires running through the engine compartment. I just had a bung welded on before the muffler (entry-exit left to right) so over the next few days i'll bolt it back on and figure out where to mount the LC-1. (IMG:http://www.914world.com/bbs2/html/emoticons/beer.gif)

I'll be anxiously watching...

I finally found the longer MTS cable I need for my LC-1 yesterday (after going to 6 stores without luck, they had it at the 7th)

brant

[lapuwali]

Switching to MS2 from a fully working MS1 setup involves just swapping the processors, flipping a switch in Megatune, then using the new options you get from MS2. The maps you make for MS1 should work. So, learning how to do all of the tuning with MS1 is hardly wasted effort. The MS1 software is still more stable than the MS2 software, and it's still likely to be awhile before the MS2 code has all of the options and stability of the MS1 code.

[Hydra]

MS1 it is then (IMG:http://www.914world.com/bbs2/html/emoticons/biggrin.gif)
Thanks for your input james.
highjack ended...

[fiid]

QUOTE (lapuwali @ Mar 30 2006, 10:55 AM)

Switching to MS2 from a fully working MS1 setup involves just swapping the processors, flipping a switch in Megatune, then using the new options you get from MS2. The maps you make for MS1 should work. So, learning how to do all of the tuning with MS1 is hardly wasted effort. The MS1 software is still more stable than the MS2 software, and it's still likely to be awhile before the MS2 code has all of the options and stability of the MS1 code.

Almost true....

If you were running MS-n-S EXTRA on your MS-1 processor and you upgrade to MSII - you might have to reroute some of the extra inputs that you might have been using... for example - I used to use the LED17 output for my EDIS output - and MSII has an explicit output for that - so I had to reroute my Xxx (whatever it was output).

This basically means you can't swap back and forth a lot; but it did take less than 2 hours for me to make all the changes and migrate everything over.

The spark and fuel maps can be moved over, and if you have an 8x8 map in your MS1 - the software will interpolate it up to a 12x12 map for the MSII, so that's a very painless process.

[lapuwali]

QUOTE (fiid @ Mar 30 2006, 02:29 PM)

QUOTE (lapuwali @ Mar 30 2006, 10:55 AM) Switching to MS2 from a fully working MS1 setup involves just swapping the processors, flipping a switch in Megatune, then using the new options you get from MS2.  The maps you make for MS1 should work.  So, learning how to do all of the tuning with MS1 is hardly wasted effort.  The MS1 software is still more stable than the MS2 software, and it's still likely to be awhile before the MS2 code has all of the options and stability of the MS1 code. Almost true.... If you were running MS-n-S EXTRA on your MS-1 processor and you upgrade to MSII - you might have to reroute some of the extra inputs that you might have been using... for example - I used to use the LED17 output for my EDIS output - and MSII has an explicit output for that - so I had to reroute my Xxx (whatever it was output). This basically means you can't swap back and forth a lot; but it did take less than 2 hours for me to make all the changes and migrate everything over. The spark and fuel maps can be moved over, and if you have an 8x8 map in your MS1 - the software will interpolate it up to a 12x12 map for the MSII, so that's a very painless process.

If you're starting with the v3.0 board, though, you should be able to use one of the other new v3.0 outputs for EDIS with MS1, rather than steal one of the LED drivers. I expect MS2 will, eventually, be able to drive the LED outputs for yet more "extra" outputs, too. I'm assuming it can't do it now, but only for software reasons.

The main thing you'll gain with MS2 over MS1 is the CAN support, which will be required for some of the later things coming, like the router board. This will be required if you intend to later move to something like coil-on-plug ignition, or you add so much power you need traction control, or you add a turbo and need to idle well with big injectors (hard to do right now) and don't want to go to the expense of staged injection (can be done now even with MS1).

In any case, by the time you have your MS1 working and tuned, this may all be worked out, and you can switch to MS2 easily if you decide there are features you have to have. MS1 will still be adequate for a long time, though.

[yarin]

made some nice progress last night...

i powered up the system through the relay box and cable i installed. Connect it to my laptop, checked a few sensor inputs, it works!

I adapted a $10 junkyard throttle position sensor to my stock throttle body. To configure it in megasquirt you simply click get current 0% throttle and then get current 100% throttle at full throttle. BAM, it creates tables and everything.

I should be able to finish wiring everything up tonight. The Wideband O2 is waiting to get installed, bnug already welded. That leaves reinstalling the fuel tank, manifold referenced adjustable fuel pressure regulator, tighten the lines and pressure test the system.

Aside from that I should have my race seats delivered saturday. I need to order a custom air filter from the auto parts store, should take two days. Slap the exhaust back on, make some config changes in software and BAM... crank and see what happens.

[DNHunt]

Yarin

Sounds like you are close.

When consideriing MS I vs MS II look at what each will do. Among other things, MS II has better resolution (a biggy in my book). It will drive a stepper motor for idle air control during warmup only at this point. It has relatively few extra outputs and will rely on other boards to do this and these will be slow to get in place and up complexity.

MS I is maxed out and likely will not change too much more. It does an awful lot but in most cases you must give up a function to get a new one.

I'm going to run MS I through this season and switch to MS II next winter. It will be better than MS I but, it's going to take some time

[mightyohm]

I am going down a similar path. I will get things working with MSI, then switch to MSII soon thereafter (I already have the processor). The 12x12 tables and increased pulse resolution are the big selling points, plus builtin support for the IAC as Dave mentioned, and also more options for ignition control.

I am going to start building my harness this weekend, so things are getting exciting for me too. (IMG:http://www.914world.com/bbs2/html/emoticons/biggrin.gif)

[fiid]

Not sure about the 3.0 board migration. I'm not sure what the options are for spark output in the extra code - I thought it was just FIDLE and LED17, but there might be something more suitable.


You can actually use the LED outputs on the MSII for other stuff - I'm planning on stealing one of mine for fan control.

[lapuwali]

QUOTE (jkeyzer @ Mar 31 2006, 09:39 AM)

I am going down a similar path. I will get things working with MSI, then switch to MSII soon thereafter (I already have the processor). The 12x12 tables and increased pulse resolution are the big selling points, plus builtin support for the IAC as Dave mentioned, and also more options for ignition control. I am going to start building my harness this weekend, so things are getting exciting for me too. (IMG:http://www.914world.com/bbs2/html/emoticons/biggrin.gif)

MS1 with extra has 12x12 tables now...

[lapuwali]

QUOTE (fiid @ Mar 31 2006, 10:23 AM)

Not sure about the 3.0 board migration. I'm not sure what the options are for spark output in the extra code - I thought it was just FIDLE and LED17, but there might be something more suitable. You can actually use the LED outputs on the MSII for other stuff - I'm planning on stealing one of mine for fan control.

The extra code, I'm pretty sure, will drive the on-board VB921 on the v3.0 board for full dwell control of one coil (for distributor-based ignition), and advance (with a locked dizzy).

The v2.2 board requires an external coil driver that does dwell, like a GM HEI module.

With EDIS, of course, it's a different thing, and there's no real advantage of the v3.0 board over the v2.2 board, so far as I know.

I'm planning on firing a Mallory with the high-current coil driver, not using EDIS.

[yarin]

QUOTE (lapuwali @ Mar 31 2006, 10:38 AM)

QUOTE (fiid @ Mar 31 2006, 10:23 AM) Not sure about the 3.0 board migration.  I'm not sure what the options are for spark output in the extra code - I thought it was just FIDLE and LED17, but there might be something more suitable. You can actually use the LED outputs on the MSII for other stuff - I'm planning on stealing one of mine for fan control. The extra code, I'm pretty sure, will drive the on-board VB921 on the v3.0 board for full dwell control of one coil (for distributor-based ignition), and advance (with a locked dizzy). The v2.2 board requires an external coil driver that does dwell, like a GM HEI module. With EDIS, of course, it's a different thing, and there's no real advantage of the v3.0 board over the v2.2 board, so far as I know. I'm planning on firing a Mallory with the high-current coil driver, not using EDIS.

I feel much more comfortable using the latest and greatest proven technology, especially when it comes hard wiring.

The V3.0 PCB has been upgraded to the following, specs here:
http://www.megasquirt.info/ms2/pcb.htm

Is it time to go home yet? The garage is calling my name. (IMG:http://www.914world.com/bbs2/html/emoticons/boldblue.gif)