MS3 update (August 2009)

[jsmcortina]

Time for an update.

Back in April we announced the MS3 daughtercard concept. A batch of 100 of these cards has been made. A small number of these cards are with the developers and are actively being used.

The code on the MS3 is an evolution from the MS2/Extra 2.1.0 code, so all the features present on that platform are carried over.

On top of the MS2/Extra features we have added:
- Use XGATE co-processor for fuel and spark for increased accuracy
- Sequential fuel and spark for up to 8 cylinders (or semi-sequential and wasted spark for up to 16 cylinders.)
- Individual injector trim
- On-board SD card datalogging
- True PWM for idle
- Spark tables extended to 16x16
- VE tables in 0.1% precision
- New MAP sampling method
- Existing wheel support extended to allow COP/sequential where possible
- boost table switching
- timed boost delay from launch
- variable launch limit
- 3-step limiter
- features extended to use new available I/O
- airdensity and barometric correction changed to 0.1% steps
- simplify tuning menu arrangement

The hardware features a built-in USB-serial adapter and the SDcard slot.

Within the next few weeks MS3 cards will be issued to a small beta test group to expand out testing and ensure quality before the MS3 goes on
general sale.

The first release of the MS3 will be the daughtercard alone which plugs into a V3 or V3.57 mainboard in the same way as an MS2. This allows
existing users a plug-in swap upgrade.

Soon after the MS3 release we should expect to see the MS3X board, this contains the hardware required to benefit from the many new inputs and
outputs of the MS3.

New MS3 homepage at: http://www.ms3efi.com
Start of MS3 specific documentation at: http://www.ms3efi.com/doc/index.html

Code TODO list:
For 1.0 code:
- Test all seq modes with all wheels
- Add semi-seq on crank for wheels that will support it
- Merge X-tau
- Merge MAF
- Add AFM mode
- Add new staging duty mode
- Sequential Staging for rotary
- ITB load mode
- Merge any recent changes from ms2/extra
- 2-output idle speed control
- Regular AE on with EAE.
- Regular AE in % of req_fuel instead of ms.
- Non-linear injector pulsewidth curve
- Non-linear battery voltage correction

For 2.0 code:
- Add angle clock
- Create "Rotary mode" to support up to 4-rotor engines
- Knock windows

That's what Ken and I have been up to code wise, additionally Phil Tobin has been keeping track with TunerStudioMS with a number of fixes and enhancements specifically for MS3.
I know that DIYautotune have been working on the revised case requirements to suit MS3, not sure about any other vendors at the moment.

James

[muythaibxr]

The XGATE is only used for fuel/spark in "ms3x" mode though. In "upgrade" mode fuel and spark are still done on the main CPU.

Also, I'm sure that we'll do more for 2.0. Just getting angle clock working is a huge task though as it involves ripping out huge pieces of code and replacing them with more huge pieces of code (in XGATE instead of on the main CPU), and then regression testing EVERY WHEEL AND SPARK MODE.

Ken

[Black99rt]

- Regular AE in % of req_fuel instead of ms.
- Non-linear injector pulsewidth curve
- Non-linear battery voltage correction

Awesome. Keep up the good work guys.

[Maglin]

I've been gone from the forums for a bit. I've been waiting for MS3 for sometime. I have my MSII/Extra all done and running on a 4g63.

One quick question. With the use of sequential COP ignition and Seq. injection using a 36-1 wheel would the use of a cam sensor with one window be needed? I'm thinking so but I thought I would ask what might be a dumb question. Oh and sign me up for a beta board if it's still possible. I have the injector driver board already for my MSII so I think I would be easy to run sequential COP and injection on my 4 cylinder.

Jeremiah

[UnaClocker]

Any word on making another batch or maybe even entering into production on the MS3 units? ie... How long do the rest of us have till we can buy our way into beta testing?

[Matt Cramer]

I've been gone from the forums for a bit. I've been waiting for MS3 for sometime. I have my MSII/Extra all done and running on a 4g63.

One quick question. With the use of sequential COP ignition and Seq. injection using a 36-1 wheel would the use of a cam sensor with one window be needed? I'm thinking so but I thought I would ask what might be a dumb question. Oh and sign me up for a beta board if it's still possible. I have the injector driver board already for my MSII so I think I would be easy to run sequential COP and injection on my 4 cylinder.

Jeremiah

Yes, you would need some means of cylinder identification to tell if they're on the intake or exhaust stroke. (Theoretically, it could be a cam sensor with a different number of windows. But that might need special code.)

[muythaibxr]

Any word on making another batch or maybe even entering into production on the MS3 units? ie... How long do the rest of us have till we can buy our way into beta testing?

James and I have been discussing that, and will talk to Bruce about it as soon as we get a chance.

Ken

[jsmcortina]

We've discussed with Bruce.
The current view on beta testing is that we want to test hard with the current test group and also with the MS3X board which is due in about a month.

So I'm afraid patience is called for, the aim is to work the hardware as much as possible during beta testing to avoid future problems after release.

With that, if there's someone out there with a full effort V8 with nitrous, water, staged methanol, boost, idle control, fan, VSSes, full sequential, VVT, tacho, shift lights, dual EGO, realtime baro and a kitchen sink who'll be using SD logging and USB then do get in touch if you haven't already!

My two V8s are a reasonable test, but I won't be using ALL of the outputs concurrently.

James

[UnaClocker]

Ok.. Patience it is..
Thad, I'm moving in with you! hahaha.. ;)

[aarc240]

James might need some major house extensions, there are plenty out here sort-of patiently waiting for MS-3.
I've got most bits in place for our trials 240z but no controller to handle seq inj AND CNP ignition on 6cyl.

[Tjabo]

Thad, I'm moving in with you! hahaha.. ;)

I totally understand!

[UnaClocker]

Do you guys have a target date for rolling 1.0 out? I imagine part of it will be waiting for the hardware to be ready for production as well.. I'm just curious if you guys had a date in mind.. I imagine May 2010 for sure, since that's the next MegaMeet, but I'm hoping it'll be "in stores" sooner, so I can get it into my project for this next race season that starts before May..

[jsmcortina]

I think it is likely we'll wait until we have the MS3X prototypes tested, just in case they expose any issue with the base MS3.

I'd still like to think they might be out this year, but that isn't a promise!

James

[joedokes]

I've been anxiously following the various discussions on the development Megasquirt III and from what I understand is that testing is proceeding quite well. It appears that a number of people have successfully installed Megasquirt III in the upgrade mode. (I particularly liked when one beta tester posted just to let everyone know that everything was working very well.)

My question, beyond the development and mechanical installation of the MS3X board and pin outs, is there going to be a lot of new computer code written or 'ported' to make the sequential portion of the injection work?

The reason I ask is that I am well aware that the fabrication issues can often be solved pretty easily while anytime you change or modify code, what appears to be simple might take months, while what at first blush appears to be hard might be easily solved.

Thanks

Regards

Joe Dokes

[muythaibxr]

My question, beyond the development and mechanical installation of the MS3X board and pin outs, is there going to be a lot of new computer code written or 'ported' to make the sequential portion of the injection work?

We've already written a ton of new code... As far as sequential goes, it was a completely from-scratch effort on the MS3. However, fully sequential mode is done, and working fine, and has been tested in a car with no issues.

The reason I ask is that I am well aware that the fabrication issues can often be solved pretty easily while anytime you change or modify code, what appears to be simple might take months, while what at first blush appears to be hard might be easily solved.

Actually, this is backwards... Most of the time, if hardware is bad, it costs a lot of money and time to get it fixed. Basically a respin of a board can cost thousands and take a huge amount of time to test and vet, often causing a full regression of the hardware to be necessary.

Software on the other hand easier to fix. If there's a problem, we reproduce it, find the piece of code that's causing it, fix it, and you reflash a new firmware. It costs nothing, and usually doesn't take us much time either.

Ken

[jsmcortina]

Sequential 4cyl has been tested in-car by Keith with his own injector driver setup. I have also tested it on my V8 with a mock up of the MS3X board outputs, my car isn't ready for the road yet so it isn't a comprehensive test.

As said though, testing appears to be going fairly well.

James

[joedokes]

Thanks for both the prompt reply and good explanation of the development process.

As a newbie who hopes to use this on a 5.7 Hemi I eagerly await both the release and every developmental update.

Regards

Joe Dokes

[phj420]

As someone who has been working with the megasquirt II for a while I am eagerly awaiting the release of the MS3. I feel confident in my understanding of how the ms works and how to install/test/tune it. I feel like I totally missed out of the beta testing for both MSIII and the GPIO board I started a new job and dropped off of the face of the planet for a little while but I am back now. If there is any chance that another batch of beta boards are to be sent out please put me on the list. I gaurentee that I will provide good and constructive feedback, and I will do my best to help with the product development as best I can.

thank you
Anthony Jacobs

[jsmcortina]

Hopefully we'll be into a wider beta phase at some point soon.

James

[gurov]

so this might have been mentioned somewhere already, but will the entire ms3+ms3x rig provide relay drivers for say electric fan/alky/generic outputs or will i still have to hack in a resistor + transistor + diode and hook it up to the right pin to sink a relay to turn a fan on ?