Release 0.1.2

[jbelanger]

There is a new release available for the IOx. The firmware is available here and the ini file is available here. An msq with the default settings is available here. There is also a zip file with these 3 files available here.

This new release ports the features added to the TinyIOx release 0.1.0 as described here: viewtopic.php?f=9&t=1102#p6383. There is also a small change to the offset of the GPS variables but that should not have any impact on anyone unless those variables were read from a CAN device (I am not aware of any setup using this at this time). This change makes it easier to poll the basic position/speed data over CAN and will make it possible for MS3 to log the data in an upcoming version.

Jean

EDIT: It is now preferable to configure the IOx as a CAN device in an MS2/Extra or MS3 project instead of having a standalone IOx project. If you do not do it this way, you may have errors and warnings that will require some changes to the default data. Also, you won't be able to use some of the features if not done this way.

[masterx81]

This evening i'll try it
Having the id of the iox changed, i still need to flash it via serial interface?
Thanks!

[jbelanger]

Yes, the CAN downloader is still assuming a CAN ID of 5 so if this has been changed then you'd still need to use the serial one.

However, I will have a look at making another version that allows any CAN ID.

Jean

[masterx81]

I get a signature mismatch....
Current configuration 'IOExtender Rv0.1.2'
Controller required 'IOExtender Rv0.0.12'
Firmware version IO Extender Alpha V0.0.13

I've edited the ini to match and seem ok, but also the default msq has the same problem.
Plus i can't reload the old msq...

[jbelanger]

That means that you have not loaded the correct firmware to the IOx: it still has the old version. And I just checked and the s19 file linked above is correct and has the correct signature.

Never change the signature in the ini to match because then you corrupt the data and possibly the firmware.

Jean

[masterx81]

Experimented myself
Yes, i've done a mess with fw/ini files
Seem all to work, tomorrow i'll recalibrate the sensora and gauges

[jbelanger]

Good to see you've corrected the issue.

By the way, I have updated the ioedl_ms3.exe file. This now can take another argument after the baud rate which is the CAN ID. So the command is now:

Code: Select all

ioedl_ms3 COMx 115200 6 io_extender.exe

where 6 is the CAN ID. You can omit the CAN ID and it will default to 5 so the old command line is still valid.

Jean

[masterx81]

The default id for iox isn't 6? I've changed mine to 5...

[jbelanger]

The IOx default CAN ID is 5. The TinyIOx default CAN ID is 4.

Jean

[masterx81]

Tried it today, it work like a charm! So much more precision with the 10 bit pwm and 12 bit adc!

[masterx81]

After the update of the iox every time that i open tunerstudio i get this error

Error: Unable to resolve Expression: getChannelScaleByOffset(gen_pwm_xaxis_canidC,gen_pwm_xaxis_offsetC)
EcuOutputChannel::Error executing formula: getChannelScaleByOffset(gen_pwm_xaxis_canidC,gen_pwm_xaxis_offsetC)
Error:
No OutputChannel found for offset: 453
Error: Unable to resolve Expression: getChannelTranslateByOffset(gen_pwm_xaxis_canidC,gen_pwm_xaxis_offsetC)
EcuOutputChannel::Error executing formula: getChannelTranslateByOffset(gen_pwm_xaxis_canidC,gen_pwm_xaxis_offsetC)

repeated a lot of times. What can be?

[jbelanger]

The problem is that you probably are not using this as a CAN device in an MS project. The default value for the x axis for the table is the RPM on the MS. But that means that there must be a device with a CAN ID of 0 in the project of TS won't be able to resolve this.

Since you're using a curve (I assume), you haven't edited the default x axis for the table so it is still the default. You could use an MS project with the IOx as a CAN device (which is now the preferred method since there are many things that require it to work). Or you could switch to a table, change the device for the x axis and change back to a curve. Just make sure the load values are still correct after this (all the values will look strange in the table but ignore it).

Jean

EDIT: I have edited the first post to reflect this.

[masterx81]

I'm using it as a can device in the ts project...
The device is listed in the projct properties.
I've also tried to delete the device ad re-add it, same problem.
Note that on generic pwm out A i'm using a curve based on map on the ecu/can id 0...
The problem seem on channel C that is from an IOX input (also changing it as table both axis are on iox)

[jbelanger]

Can you post your project? Use the TS archive function and post the .tsproj file. There might be something I missed or something that TS doesn't like in what I do in the ini.

Jean

[jbelanger]

I should add that sometimes you need to not only change the device but change the axis variable. Even if you see something there that makes sense, TS is simply displaying something but doesn't actually change the value in the msq and the error message remains.

Jean

[masterx81]

Here is the tsproj... Now i'll try to change all in the PWM C channel
But why only on C and D channels there is the error?

EDIT: I've tryied to change all, nothing,the problem remain...

[jbelanger]

I got the error messages on C and D (a lot of them). If I change the x axis to AD1 for both C and D (and go back to a curve for C and turn off PWM D) then the error messages go away.

I had a look at your original msq and there is data corruption for both C and D with an offset and size that doesn't make sense. I don't know if this msq was created when you tried to load the new code and got the old code and changed the signature to match. If that was done then that could have corrupted the data. If that's not the case then something else corrupted the data.

For now, I would say that you should simply do as I did and that should solve this msq problem. If you see that again then there is something else that corrupts the data and that needs to be solved. You can also use the msq I attached which I corrected from your project.

Jean

[masterx81]

Ok, this time i've fixed it reselecting the x axis and all seem to work.
I don't know what have broken those settings, maybe the first fw update that was gone wrong. It's strange because after that i've started with a new iox msq.
I don't know, but now seem all ok...
Thanks!!!

[jbelanger]

It is possible that there is a bug in the code that corrupts the data and that my testing did not discover it. So if this does happen again then I'll have to do some more tests to try and find it. But it might have been some other external causes in which case you will likely not see it again.

Jean

[masterx81]

I hope so
By now, the fw works well, the 12 bit adc and 10 bit pwm has really changed the accuracy of my gauges, now they are REALLY precise