thanks, i'll look into that
found the problem. i typo'd and wrote "map" instead of "maf" in the wideband analyzer for that axis.also, on the analysis cruising AFR, I'm not sure what's going on, but I'm getting no data in the MAF AFR below 31 AFGS. Not sure if it's not lining up, or the data just isn't making it there.
the degree symbol doesn't have any trouble with windows, but i'll drop it in the definition file and just write 'deg' instead, it's easier.on the datalog main tab for Spark adv and KR, I'm assuming that there should bed a degree symbol, but I'm seeing � instead.
thanks for that log. very helpful. i can test the hell out of it. especially nice to have something with wideband data to test the analyzer.I've attached the log
can you tell me your wideband settings?
Originally Posted by steveo
lowpoint 9.00:1
highpoint 5v 19.00:1
trim 0.00:1
1994 LT1/4L60E Formula
i have some decent bug fixes now but i wont release another beta until the definition file is good to go. speed logging should work again in the next version as should e-side mode1
Hiya
Just been out with 4.5 Beta2 and noticed a couple of things:
AFR display on Datalog window sits at 0.4:1
Injector BPW only show ms units. there are no fractional parts shown.
My car is idling around 1.8ms and the display is showing a static 2 until i tickle the throttle
CL Performance : Analyzer: BLM averages are showing 192 where the read BLM data cells shows 128
All cells showing showing 180-192
Q: Can you show an average of both sides ?
Graphing: Select element & select overlay don't show the selected item ??? Not sure it should but feels wrong ( Don't know )
On the graphs, selecting Injector bpw sets the scale 0-900. single digit ms values hardly makes a full dot on the graph.
Other than that it's working excelently.
Thankyou.
Mitch
The graphing options are so much better in the latest version!
I was pretty excited to see I could graph afr target, and then also wideband and see the big picture. It's interesting.
Also, here's a couple more logs for you if you want some more to work with. They're my daily commute from going to and coming from work today. about an hour total. I fixed the maf analysis in my local source so I can try to dial my afr in a little better. still pretty rich at low airflow.
anyway, great progress.
1994 LT1/4L60E Formula
hi!
you need to be patched up for that to work. i will be fixing it so it will display blank lines if not patched. work in progressJust been out with 4.5 Beta2 and noticed a couple of things:
AFR display on Datalog window sits at 0.4:1
thanks i'll fix that. the definition is in early stages. i plan to spend a lot more time on it. display precision is also user configurable if you edit the definition file yourself (for guys that insist they need three decimal places of MAF AFGS or whtaever)Injector BPW only show ms units. there are no fractional parts shown.
My car is idling around 1.8ms and the display is showing a static 2 until i tickle the throttle
read blm cell memory shows the stored blm values in the ecm. the analyzer actually analyzes the running data. the analyzer is using an average of the left and right cells. edit: can you post a log? i have no closed loop log for testing.CL Performance : Analyzer: BLM averages are showing 192 where the read BLM data cells shows 128
All cells showing showing 180-192
Q: Can you show an average of both sides ?
the selected element is displayed in the graph (look at the text by the graph) but not the overlay. i plan to fix that soonGraphing: Select element & select overlay don't show the selected item ??? Not sure it should but feels wrong ( Don't know )
the graph uses theoretical maximum and minimum values now, which are calculated... perhaps i should use a bit of logic there or allow zooming that axis. i'll look at itOn the graphs, selecting Injector bpw sets the scale 0-900. single digit ms values hardly makes a full dot on the graph.
thanks for the feedback!Other than that it's working excelently.
Thankyou.
Mitch
thanks man more data the better. the analyzer will be getting a bit more work done to it soon too.
kur4o's e-side mode1 patch works flawlessly. im working on constructing the ideal message right now, and working out some other bugs related to dealing with independent devices in the datalog. should be ready soon.
new beta: http://fbodytech.com/eehack-2/beta/
biggest thing is that multiple device comms work now. this is a big deal
i rolled in patch 'version 3' with eside modes 1-3 enabled thanks to kur4o, and a short custom e-side message and definition to play with.
please flash to try it (i doubt it'll hurt anything), but also let me know how it works with your 'old patch version'
i have the graphing ui working a bit better but still haven't solved the range problem with some variables, i'll do that soon
definition file is improving, but feel free to load it into a spreadsheet and suggest further improvements. there are still two messages not done yet.
one thing to note about the definition format which is a cool design feature is it actually loads and displays parameters 'in order', so you can change the order that things appear in the parameter lists in eehack by simply moving lines around in the def file (for example if 'wideband' is too far down the list for you, simply move its line to the top).
when I launch i get these messages in the console:
Can't find element: FAN1 in definition 5
Can't find element: FAN2 in definition 5
Can't find element: AIR in definition 5
Can't find element: AC in definition 5
Can't find element: PRNDLFLG in definition 5
Can't find element: PARKNEUTRAL in definition 5
Can't find element: TCCENABLE in definition 5
Do I need a newer definition file?
1994 LT1/4L60E Formula
I`ve started working on more patches. Found much better way to upload tables and execute small code without mode 5 and 6 be needed and it will be very easy to use.
For now it will be only on eside due to unlimited free space and memory.
I cracked the way aldl works. When you send any mode, any lenght command even not existent mode [ff for example] up to 168 bytes long it gets stored on memory. On eside the storing address starts at 188a on tside at 1993.
If you send existent mode there is address pointer for code execution and if you put code there with some commands to transfer the memory location stored at 188a to another free memory location first, then execute code from there gives unlimited hack possibilities, like direct memory write, load custom tables and switch to enable them. Transfering whole tables from rom to ram with switch to enable them and then using direct memory write to change single cells of the table. That way real time tuning for some of the tables is very possible.
Good news is on eside there are seven aldl modes that are not used and will be repurposed with custom code to execute.
You can change mode 2 to upload custom lenght message upto 112 bytes long in 16 bytes increment.
Change $41 to $71 for 112 bytes long mode2 message.
You can have any lenght message change the first digit only.
I also dig in the mode 1 msg 5 on tside. 44-49 are defenitely derived from byte 1-6.
Some looks like timers for errors set at ignition cycles. Still needs more work.
its just bitching because those are in the dashboard but not my hacked up mode5 'speed logging' definition. windows people dont see that output. i'll see if i can shut it up
that's brilliantly simple. im not suprised it works. so that memory is never initialized at the tail during subsequent commands, it's just filled up to length? we could freely add a small table to the end. maybe we can even figure out a way to initialize it first. that'd be a good way to temporarily test individual cylinder fuel trims or something.
kur4o i think a good goal now is real-time modification of the individual cylinder fuel trims. its something people really want.
we talked about this before, but do you have any idea of the best approach? i was always thinking a mode-4-like command on the eside that is 8 bytes, one for each cylinder.. but your new idea of kinda 'buffer overrun' is really smart.
the nice thing about the individual trims code is it's already kinda set up for multiple tables with a subroutine and an output value so switching should be easy
i will integrate anything you want into eehack.
Posting Permissions
Reply With Quote
Bookmarks