Fred's firmware development diary

[Fred]

Last night Sean (EssEss) and I had a good long chat about ECU design and software methods and process etc etc. It's good to know there are some really talented people out there willing to help - Sean is definitely one of them! :-)

I was working with Aaron tonight to try to nail a few minor issues in the tuner, but no joy yet. Hopefully we can debug those issues in the near future and get something working.

I still need to sit down and put some time into my firmware in the next few days until it behaves as anticipated. Once I do there should be more posts from me here related to that progress.

It was really good to meet Sean and talk over the project with him, I look forward to working with him remotely on it in future.

Fred.

[Fred]

Last night Aaron finally fixed the bad packet data issues with the Tuner, however introduced a big speed hit and some instability in the process. I reckon another few hours and he'll have it licked, once he does I have zero excuses left :-)

Tonight I measured the internal bias resistor for the IAT sensor on the Volvo and created tables with FreeTherm for that configuration - it finally reads temps accurately. Yippee.



It's already 3am now that I've done that and tested a few other things and driven the car 10 miles or so to ensure it was fully warmed up, so off to bed for me, tomorrow I'll get stuck into some serious work and hopefully get some good work out of young Aaron too.

I really need to tidy up my GIT situation and see what is going on with it and get the latest stuff up onto the github repo for others to see. I'm not sure what hacked up stuff is there at this time, but I need to create some new files and maybe directories too and stuff too.

Location: 1hr north of Denver in Colorado with a guy who has built and tuned about 30 ms units and will be very happy to never build another :-)

Fred.

[Fred]

My USA travels are happening at a suitably high pace that it's difficult to achieve anything, however I spent an hour on the code earlier and found what I think is the issue, so have patched that and will read through it again soon to make sure and then try it.

In other news the tuner is sweet, it just needs a little more to be useful to actually tune something. Hopefully Aaron comes through with that soon. Datalogs would be nice too, hint hint, Sean :-)

More soon.

Fred.

[Fred]

Late late late last night here in Toronto I added some diagnostic code and will load it this morning and give it a try today. The thing I need is a logger (Sean!) and then we can see if it was legitimately running or just by luck. I'll get a feel for that today just driving it and watching the numbers. Perhaps I can stop, tweak the values around and give it a more serious try later? More soon.

Fred.

[Fred]

Code works, so next step is to adjust vars to make sense and add some more stuff. I'm pretty much relying on/waiting on Sean and Aaron now to make more progress. Will be away from the car for a few days while in NYC so you have a day to get me test code and then a gap to refine it before I have more of a play.

Fred.

[Fred]

Watching the tuner real time display tonight and found a bug...

http://issues.freeems.org/view.php?id=105

It's a wonder it ran at all! Shouldn't have... hmmmm...

Working on that asap.

Fred.

[Fred]

Code: Select all

./fuelAndIgnitionCalcs.c:189:	masterPulseWidth = (ADCArrays->EGO << 6) + (ADCArrays->MAP >> 4);

This explains why it works at WOT and not otherwise.... (not even using the fuel calcs.... old debug in there... woops...)

Now I need to fix the RPM readout properly and the table issue and it should be good. at the moment it will probably run ok when i get back to the car, but we'll see...

Another thing fixed! Pre-breakfast coding in a high rise in new york city for the win :-)

Fred.

[Fred]

Actually, I found the table bug last night and fixed it and promptly forgot, just the RPM values to fix now, yippee :-)

[Fred]

Table bug tested by Cam in Auckland NZ while I coached him through it from NYC USA :-)

Confirmed fixed!

http://issues.freeems.org/view.php?id=105

Aaron has the data he needs in his inbox right now :-)

Fred.

[Fred]

I did some debugging and coding at the top of the Empire State Building a day or two ago, helped Sean with the parser a little and tracked down some of my own issues too. Still need to solve the RPM bug, started on that, and now need to make the build not be broken again. Once I figure out what I was doing before and what I'm doing wrong now, we should be golden for that too.

I'll see my car again for the first time in a week tomorrow morning - yay! :-)

I even have fresh code to test on it ASAP, both tuner and firmware :-)

Fred.