Can't execute RoBoRC.exe problem

[ries]

Dear all,

I have installed win xp sp2 on the roboard and it works!
After copying the demo roboard the program can't run.

I have installed the redistribution 2008 package, the winio folder and the roboard dll's and libs.
The message that i get is:
This application has failed to start because the application configuration is incorrect.....

I have checked all the missing dll's with dependency walker but the program still won't run.

On my lap where MSVS2005 TS is installed the program runs.

What could be the problem?

Thanks,
Ries

[roboard]

This application has failed to start because the application configuration is incorrect.....

1. Please ensure that winio.dll and winio.sys are put in the same directory of RoBoRC.exe (or in the System32 directory of WinXP).

2. install VC2005 SP1 Redistributable Package (http://www.microsoft.com/downloads/details.aspx?FamilyID=200b2fd9-ae1a-4a14-984d-389c36f85647&displaylang=en) and try it again

[ries]

Hi Roboard, thanks for the reply.
I did all that, but shall do it again.
I have been reading about this error message and they mention copying the manifest and microsoft.vc80.crt folder into the same folder as where you execute the program.

A bit overdone, but at the moment it doesn't want to work.

<Edit>
By copying the C:\Program Files\Microsoft Visual Studio 8\VC\redist\x86\Microsoft.VC80.CRT folder to the same folder as where the roborc.exe is located and the manifest in the release from the example, the program would run.
Now it doesn't read the servo's :S
</edit>

Regards,
Ries

[roboard]

Hi Rise,

there is VC 2k5 and VC2k8 compatiable issue, could rebuild under your vc2008.
if you want to use vc2008.

[ries]

Hi Roboard,

Im only using vc2005, I dont have vc2008 .

But could it be that the manifest and the crt are not 'seen' by the software?
Maybe the path is incorrect.
Currently im away from work so hopefully tomorrow I can check why I need to put to folders in my roborc.exe folder.

Regards,
Ries

[ries]

Dear Roboard,

I didn't had the time yet to figure out the manifest problem.
But when trying the roboard as is (with the manifest crt folder copied into the same folder as where roborc.exe is located) I can't read the
ERROR: failed to read the position of channel 0.

I have KONDO KRS-788 installed on channel 0.

The command given is: roborc.exe CAPTURE 0000.txt -CH 000000ff
channels 0-7 give the above error and the rest of the channels give a zero.

What am I doing wrong here?

<edit>
More info about my problem: according to the manual roboard_training_sw_v15.pdf page 41&42 i have put the pull down/pull up switch to pull down and set the bios items accordingly.
Measuring with a voltmeter the voltage on the pwm pin 0 is zero volt.

I have written a test program to put a 'high' on pwm pin 0,
But the pin didn't go high.
My test program:

#include "stdafx.h"
#include "roboard_dll.h"

int main(void)
{
if (rcservo_Initialize(RCSERVO_USECHANNEL1)) {
while(1) {
rcservo_Outp(0,1);
}
rcservo_Close();
}

return 0;
}

I use VC2005 on my laptop and copy the .exe from the release folder to the roboard.

Page 52 of the above mentioned manual says that RoBoIO recognizes Roboard's CPU, and doesn't run on other PC.
But i can start the .exe on my laptop????
Do I build the .exe wrongly?
What am I doing wrong......

Pls advice, thanks!
</edit>

Greetings,
Ries

[roboard]

Hi,

1.
RoBoRC.exe can only run on your laptop without arguments (to see help messages). Running RoBoRC.exe with arguments on a laptop, you will get an error message in roboard.log:

ERROR: RCSERVO lib fails to initialize (unrecognized CPU)!

2.
For the manifest problem, if you install VC2005 SP1 Redistributable Package in RoBoard, you should be able to launch RoBoRC.exe directly (without copying any manifest crt folder).

3.
Please ensure that in RoBoard BIOS, Bit 6 of Port 3 is set to 0 and that you download the newest RoBoIO binary. Then try the following code to see whether the GPIO function works.

Code: Select all

#include "stdafx.h"
#include <stdio.h>
#include <conio.h>

#include "roboard_dll.h"

int main(void)
{
    if (rcservo_Initialize(RCSERVO_USENOCHANNEL)) {
        printf("set GPIO pin 0 (PWM channel 0) to high\n");
        rcservo_Outp(0,1);
        getch();

        printf("Close the above test\n");
        rcservo_Close();
    }

    return 0;
}

An additional remark: if you use the RoBoIO DLL version, you need to copy RoBoIO.dll to the folder of your test program.

[ries]

Hi Roboard, thanks for the reply.

I reinstalled windows xp sp2 again.

1. I did see another error on my laptop when initializing the rcservo lib:
RoBoRC.exe capture 0000.txt -ch ffffffff
Open 0000.txt to write... file already exists. overwirte? (Y/N) ...success
Initializing RCSERVO lib (for KONDO KRS-786 & KRS-788)......fail!
But this doesn't really matter.

2. Tried a lot, even a new xp install (sp2) but I still need the manifest together with the 3 dll's (msvcm80.dll, msvcp80.dll, msvcr80.dll).
But at the moment this solution works.

3. Done all that. Switch is in pull down position. GPIO36 = 0.
GPIO00-07 = out=0;GPIO10-17 = out=0; GPIO20-27 = out=0;

Compiled your code on my laptop and run it on my RB100.
When measuring the GPIO00 it reads 3.3V. So i assume its CMOS.
I have changed your program so i can set the GPIO00 to low and this works as well.

Full of hope and running the roborc.exe... nothing happens.
I still get the error fail to read the position of channel 0.

Is the 3.3V output voltage on the GPIO too low for the KRS-788HV servo's?

Could you throw in some more ideas?

Regards
Ries

[roboard]

Hi, ries,

RoBoRC.exe records error messages in "roboard.log". When RoBoRC.exe fails, you can see the cause in that file.

3.3V is ok for KRS-788. We have previously tested RoBoRC.exe with KRS-788 and see that it works. I will make the same test again for you, and will report the result tomorrow.

Also, to help locating your RoBoard problem, please do the following steps on your RoBoard:

1. remove KRS-788 from RoBoard.
2. set GPIO Port 0 Bit 0 as "IN" (input) in BIOS.
3. boot into WinXP.
4. measure the voltage of GPIO pin 0 (PWM channel 0).
5. report the voltage value to us.

[ries]

Dear Roboard,

I must let you know I appreciate the help you're offering very much!
A very good service, even on a sunday you're given your help.

Again thanks very much and i'm gonna test what you wrote and let you know the outcome.

At the moment in the bios i have set the ports to GPIO. This is ok?
I read in the manual about GPIO and than about PWM.
Or should I set the ports to PWM?

Regards and many thanks for your help.
Ries

[ries]

Dear Roboard,

the voltage measured on the GPIO00 IN is 6.7mV (so it's 0V).

I have measured the output pin of the RCB3 board of the KHR-1HV (the robot that's gonna get the roboard) and it's 5V
So i hope the khr-788HV is TTL compliant.

Hope this will help!

Regards,
Ries

[roboard]

Hi, ries,

I have tested the RoBoRC.exe with KRS-788HV again, and it works.

It seems that your RoBoard has some trouble on either its GPIO36 pin or its PWM pull-up inverter.

Please measure the voltage of the "Reserve" pin at the I2C/SPI connector (refer to the RB-100 HW introduction slides) under WinXP. If it is high, then your GPIO36 pin may be anomalistic; if it is low, then your PWM pull-up inverter may be bad. You can mail to tech@roboard.com for fixing your board.

[ries]

Dear Roboard,

Thanks for this update. I shall measure tonight.

Are the KRS-40xx series also supported with capturing?

In the KHR-1HV there are 2 KRS-4024 servo's and when capturing the readings are zero.
But when altering the captured text file i can move those servo's.

Regards,
Ries

[roboard]

Are the KRS-40xx series also supported with capturing?

Hi,
in our test, RoBoRC.exe can capture the feedback of KRS-4024.

[ries]

Dear Roboard,

Sry for the late reply, i have been a couple of days away for work.

Please measure the voltage of the "Reserve" pin at the I2C/SPI connector (refer to the RB-100 HW introduction slides) under WinXP. If it is high, then your GPIO36 pin may be anomalistic; if it is low, then your PWM pull-up inverter may be bad

I have measured the reserve pin of the I2C/SPI connector.
It is measuring a 100mV.

Regards,
Ries