Toymods Car Club Inc: Tech & Conversions => MS2, UMS and PWC update (From Bruce)

Toymods Car Club
www.toymods.org.au

F.A.Q.

Home

Members

Toymods » Tech & Conversions » MS2, UMS and PWC update (From Bruce) Show: Today's Posts :: Show Polls Email to friend

Author

Topic

oldcorollas
Forums Junkie

Location:
Kita-Ku, Sapporo, Japan
Registered:
January 2003

MS2, UMS and PWC update (From Bruce)

Thu, 07 October 2004 13:39

BRUCE wrote on Sat Oct 02, 2004

I posted this in the PWC group a few days ago re. update on our activity on PWC, MS-II, MS board upgrade, and UMS:

Lately, my time has been split among getting PWC to a final shape, working on the first pass of UMS, helping Al get MS-II out the door, and doing a board spin on MS. Just a typical workload for any working individual in today's work environment [Laughing] But I like having lots of things to do - it keeps my mind from wandering.....

The PWC has been progressing with the available working time slices I have been applying to it. It is more bench measurements/verifications than anything else, in particular the pump current measurement. Things look good, but some more refinement in heater temperature feedback may be needed. I have access to a dyno which I can use as an external heat source (i.e. sensor head mounted in the header). The nice thing about the dyno is that drastic temperature variations can be applied pretty quickly (i.e. a dyno pull) and the transient response can then be monitored. Once the heater temperature feedback is nailed down things should move pretty quickly.

I also have a simple CAN communication setup running on this part, lots more to implement but at least I can talk CAN.

But, for the next few weeks, I am really focusing on helping Al get MS-II out. We have really been tesing the code and checking everything, especially the ignition side. There have been a few instances where Codewarrior has introduced something that was not exactly what we had intended, so we have been delving into the generated assembler to verify critical sections of code. Programming in C is excellent - but there are times where one really needs to see what the compiler has generated. Part of this exercise is to satisfy my own paranoia on what the compiler is really producing - I can say that Codewarrior has been doing an excellent job, and the generated code is extremely tight.

Right now, I am making up a quick missing pulse detector circuit to verify that MS-II never misses ignition or FI pulses - there are a lot of timer gymnastics going on in the code (overflows, etc) and it makes sense to verify the pulse operation over an extended period of time (i.e. many hours). But all of the ignition functions (like charging the coil before the next input pulse, adjusting dwell and spark discharge time for really fast RPMs, etc) are operating, as is extreme low RPM operation (in the tens of RPM, where timer overflows need to be accounted for). Al is also adding filtering (user optional) for the MAP and RPM signals.

The only things left for MS-II is the bootloader and the final MS-II board spin. Then more testing, in real vehicles.

In parallel to this, the MS board upgrade is also progressing. A lot of this is to emcompass the MS-II new features, but general cleanup is also being implemented. More on this shortly when I have something final to present.

For UMS - I have to get the next rev PCB made up and populated. I have a MCF5235 eval board running a bunch of TPU code - this device is awesome to program, there is so much horsepower available. For MS and MS-II, we always have to try to implement code in an efficient way and then verify this later, often changing things around to improve efficiency. For the Coldfire processor, there is so much processing power available that code efficiency is not as important - the part will blast thru any code woth gobs of time to spare. Its nice to have the extra power, and code development is much, much quicker because we do not have to spend our time trolling thru the generated assembler to make things more efficient. But I probably still will do so, just for satisfy my paranoia.

- Bruce

code and boards for MS2 seem to be nov/dec. UMS is aimed at 1st quarter 05.

i'm happy to wait for a DIY ECU that runs around 100MHz, has dual onboard wideband support (the LSU, not the L1H1) and has fully configurable outputs (inj, coil, whatever, just reconfigure it)

Cya, Stewart

PS,
PWC = precision wide-band controller. a new approach to fast response wideband with higher accuracy than has been seen previously.
MS2, Megasquirt upgrade board (plugs into processor socket) to give 12x12 tables for both injection AND ignition. initially use EDIS units to drive wasted spark.
UMS = Ultramegasquirt.

www.msefi.com

[Report message to a moderator]

[Message index]

-=] Back to Top [=-

Current Time: Fri Jan 31 12:37:48 UTC 2025

Total time taken to generate the page: 0.0086519718170166 seconds

.:: Contact :: Home ::.