'Save as' bin not initially auto re-load, crash, usb reseat

[bachig24u]

Hi Matt,
Noticed bug with 0.10.3

load bin file 123.bin,
tune car,
goto menu and save-as bin as 456.bin
update VQ map and resyncing,
Nistune will often crash (VLT image with HPX05+MAF VQ, The crash when syncing VQ map changes is more common with WB-LC1 connected).
Reload Nistune
bin 123.bin is loaded instead of 456.bin

Had to reseat the USB connector (laptop end of cable) many times as it would loose the real-time board data connection. note: belkin usb-serial dongle has LC-1 connected did not require reseating. Nistune was able to still see live Lambda data stream when it lost Type 1 data stream.

Kind regards
Simon

[bachig24u]

OK, Upgraded to 0.10.4

still has file reload/save issue after crash or using "save as" with type 1 and type 2 board
type 1 using VLT image and HPXMAF - can't use WB at all (LM-2 -OBDII), nistune crashes. (fatal disconnect error or NO nistune data received).

Type 2 board using HCR32 RB20DET 11U00 image, nistune crashes everytime click WB. same messages, fatal disconnect or NO nistune data received). issue doesn't occur if I don't click WB.

I ran over the top with 9.14.22 but now same issue as with 0.10.x seems like the 0.10.1+ upgrades have broken ability to use WB.
Some residual software cause there.
Both type 1 or 2 are working with nistune collecting from logplayer and logworks3.07 collecting WB data but getting many USB error on two differnet laptops with last stable version and latest beta of nistune ever since upgrading into 0.10.x. and rolling back.

after 5 hours at WSID last night trying to sort. gave up and used logworks.
next course of action, Will remove all software, run ftdi uninstall, clean reg keys and reinstall. keep posted.

[Matt]

Same problem as here:
viewtopic.php?f=12&t=1926&start=105
On external disconnect (ie power loss or comms dropout) the Innovate MTS crashes inside Nistune when it disconnects. Basically once the disconnect occurs it deletes some important memory locations which then shuts down the software

I've repeated the issue in the debugger but not sure why yet the memory is being wiped out. Dont see the same issue in 0.9.14.22 version. I've got an interim workaround which will be available in 0.10.5 this week when available

Note: I've updated that 0.9.14.22 version last night so it will work again with Innovate. Redownload this and install and you should be able to connect and run without crashing (there was a translation file update picked up by the installer in last upload which caused crash on connect even though re-released software was the same)

load bin file 123.bin,
tune car,
goto menu and save-as bin as 456.bin
update VQ map and resyncing,
Nistune will often crash (VLT image with HPX05+MAF VQ, The crash when syncing VQ map changes is more common with WB-LC1 connected).
Reload Nistune
bin 123.bin is loaded instead of 456.bin

Will try this out tomorrow night

Had to reseat the USB connector (laptop end of cable) many times as it would loose the real-time board data connection

This happens when the software doesnt let go of the USB handle when the software crashes. I've added some exception handling to force letting go of the USB handle in the event this occurs so there wont be the need to disconnect/reconnect to force release of USB handle

[bachig24u]

installed the latest 0.9.14.22 after a throughout clean up.. nistune registry key deletions, ftdi driver removal.

gave me same LM2 WB crashing and couldn't close via taskman until usb was physically unplugged.

put 0.10.5 over the top which fixed the crashing issue but found a few more bugs.

WB AFR logging is missing from excel logs. >>> fast play is not working or tracing correctly on fuel or timing maps either.

Sime

[Matt]

fast play is not working or tracing correctly on fuel or timing maps either

raised in released thread and now fixed (waiting next version)

0.9.14.22 rerelease gave me same LM2 WB crashing

unfortunately I cant test LM2 at the moment since ground connection unclipping took out the unit and USB port on my laptop in smoke waiting for replacement from innovate. Stick with 0.10.5 in the interim until next release

WB AFR logging is missing from excel logs

interesting have to look at this again. Wondering if only with LM-2 since had it working with LC-1 okay???

[Matt]

LC-1 testing attached.

Please try playing back the attached file. If successful then please attached your LM-2 log file for me to look at why playback isnt working

[bachig24u]

just checked again now...

I ran debug mode.
opened 0.10.5
soft connected consult to type 1 VLT image.
soft connected WB, it couldn't find belkin usb-serial to LC-1.
changed config setting to com7
connected WB all ok, live data ok.
started log recording
chose options, RPM-red, MAF-blue, TPload-Green, LAFR-purple.
logged data for a minute during cold engine run.
stopped log rec. chose keep file
started log rec again to see if new file was created with good timestamp.
logged a further minute of warm up to about 40c temp.
stopped log. chose keep file
soft disconnected WB
soft disconnected consult.
shutdown laptop.

brought laptop to warmer environment (in the office)
started laptop and nistune 0.10.5 to review logs
opened log player
automatically last saturdays log file was loaded by default even though I just created 2 new logs today.
no LAFR in dropdon list and nistune automatically fills the field with TP instead.
then browsed to the next 2 new log files, and AFR is included in them.

I'm either going stir crazy or is there some code being triggered by the way I use nistune thats causing this.

on saturday night, i stopped and started log recorder 4 or more times to make identifying the runs easier. no debugging logs though sorry.
each time nistune asked if I wanted to keep the file and i chose 'yes'.
the log file didn't stop, it just appended, creating a log file almost 10 minutes long instead of smaller single 10 second versions. The weird part is, the log file doesn't have LAFR in it but I can confirm that I saw the display on screen as i was elated when the WB was able to be used once again after the 0.10.5 install.

The only thing I didn't do on saturday that I did today was choose what I was veiwing in the log recorder.
all settings was set to RPM and ticked.
attached are the debugs from today and logs from saturday and today.

Sime

[Matt]

NIStune_2011-06-18_1939_51.csv
- Opened this one up and has no AFR (only consult registers)

Then opened up these two:
NIStune_2011-06-20_1528_06.csv
NIStune_2011-06-20_1530_52.csv

Both of these when opened show the LC-1 AFR and lambda on the screen

If the log recording is started with consult only and then wideband data is connected after the log should be stopped and user prompted to save. Not sure if it did that for you or not, but if the recording is started it will not add new data (like from the wideband) mid way through. A new log has to be started. I'll recheck that functionality again this week

each time nistune asked if I wanted to keep the file and i chose 'yes'. the log file didn't stop, it just appended, creating a log file almost 10 minutes long instead of smaller single 10 second versions.

I've made a few fixes for the connectivity of units and the order at which log player is record mode or recording for next version. Doing more testing tonight for the 0.10.6 release

[bachig24u]

Hi Matt,

had a fairly sucessful day with 0.10.5 TYPE 1, CR31 addy & binary.

WB disconnections were limited to a once off WB data window text disappearing but then recovering on it's own within 30s or so.

log recorder has all relevant data and stopped and started creating new files each time.

wideband input tracer performing normally.
No usb reseat required at all.

Only one software crash to report:
The car was about to go into motion for full load 1/8 mile standing start when software crash occured during a sync to ECU operation.
Wideband was connected, log recorder on (grabbing maf,tp,rpm,lambda), afr input tracer running (updating TP cells instead of AFR, TPS decel cut 'ticked'), timing load scale displayed and also a fuel map open.

I only wish I had debug running.

the second strange happenings were;
consistent WB display issue noted: WB screen display LAMBDA but gauge view screen shows AFR figures even when LAMBDA selected in configuration settings.

Noticed intermittently 'speed' value was not showing in gauges view or consult window although mechanical speedo working fine.
Is this related to WB data going missing intermittently or do I just have too many selections in nistune going?

[Matt]

I've noticed an issue with not all bytes in the data stream on Type 1. This was picked up when the timing value for CA18 was being glitchy. I've identified the problem and made a fix in the 0.10.6 release I was testing last weekend. It may be linked to the speedo issue

consistent WB display issue noted: WB screen display LAMBDA but gauge view screen shows AFR figures even when LAMBDA selected in configuration settings.

Okay will look into this one

With regards to log player/recorder during driving I tested various combinations of disconnect/reconnect and noticed some issues with logging modes so will look into that tonight before putting the next release up (also fast playback still to be fixed)

[Matt]

consistent WB display issue noted: WB screen display LAMBDA but gauge view screen shows AFR figures even when LAMBDA selected in configuration settings.

There are two gauges for wideband. You should see:

L02 - Lambda wideband value
LAFR - 14.7:1 based AFR wideband value

These display regardless of 'lambda' setting inside File-Configuration