LiveCode Forums

Hi,

anyone knows or tested a standalone on Rpi with latest Raspbian ?
In the IDE LC704 it works correct. I can let the steppermotor run.

Only when creating the standalone for Linux ARMv6-HF it does not execute, although it is allowed to execute as a program.

Should it work on Raspbian or do i need an other distribution?

Thanks!

Heya Sphere,

I don't have a Rpi to test with, but am curious if your actually running the IDE inside of Raspian, and if so, which version of Raspian are you using?

Hi bogs,

well maybe i have an old version, i quickly saw version 2.1 2017...

Hm. Later I'll set up a vm (I'll be gone most of today unfortunately), hopefully I can find that one around

Hi Sphere,

did you compile the standalone on the Raspi?
If so, try to compile on a Win/Mac/linux desktop.
I had success with all my trials to create standalones when compiling on Mac.

But I usually run everything with 6.5.1 from the IDE on the Raspis because the 7.0.4 version (which is the only one that is able to create standalones) is too slow. Especially you won't get an (approximately) exact timing with 7.0.4.

Hi Hermann,

yes i did compile on the Raspi using LC704.
In LC651 i did not see the option for linux-ARMv6. (maybe i installed the wrong one, so i donwloaded again. WIll install tomorrow in my lunchtime).
But had to save to legacy5.5 in LC704, else it won't open on lc651 i noticed.

I also noticed that moving the mouse does influence the motor.

Ah now i read your sentence, so it is not possible to create a standalone from lc651 on raspi?

Ok i can try to compile from win or mac and see if it then will execute.

btw. it seems my raspbian is old, so i will update it also to the latest 2018 version. https://www.raspberrypi.org/downloads/raspbian/

Thanks!

Don't update. Or the LC IDEs don't run any more 'clean'.
Or use
http://forums.livecode.com/viewtopic.ph ... 53#p140253

I have 7.0.4 working to compile standalones on the latest NOOBS on Pi Zero W, nothing too complicated but hasn't fussed at me so far.
(I also have Xubuntu on Pi 3 that builds with LC 7.0.4)

@Sphere.

But had to save to legacy 5.5 in LC704, else it won't open on lc651 i noticed.

Yes, but you don't need Unicode for your motor! That's what made the first versions of LC 7 slow.

I also noticed that moving the mouse does influence the motor.

This shouldn't occur.

Don't you use "send in <time>" for your motor?
My experiences are that the software clock of the Raspi allows (close to) "exact" intervals of not less than 1 tick for "send in <time>".

@SparkOut.

I have 7.0.4 working to compile standalones on the latest NOOBS on Pi Zero W, nothing too complicated but hasn't fussed at me so far.(I also have Xubuntu on Pi 3 that builds with LC 7.0.4)

Xubuntu/Lubuntu is also fine here on Raspi 2/3.

With Raspbian Jessie I had, IIRC, on Raspi A and 2 scenarios where the color wasn't OK (BGR instead of RGB) and transparency had failure.

Glad we have a stable enlarged Raspi party.

Yes maybe Ubuntu Mate is a good choice, referring your link.

I did not yet use the send in time.
First i had a LED version which referred to the logic inputs of the H-Bridge motor driver.
This all blinked nicely, and now i have the wires connected from GPIO to the motor driver.

A full positive pulse is 1 uSecond, so now i use the 0.001 millisecond. If increased in time the motor will slow down.
The number of steps is 200 for 360 degrees at 1.8 degrees per full step. The driver can go up to 16th steps which then is 3200 steps for 360 degrees. For selection of these Full, half, quarter, eight and 16th steps there are 3 MS lines with which you can select the mode. Works all ok.

So for now i have it working with 2 handlers and a few repeat structures. This works.
After this i want to try to rewrite it to Send in time, perhaps that will have less influence in time.
I only can work on this when i have some spare time at the job.

The idea behind is to be able to test a manual (turn) switch with say up to 4 positions which have to be tested up to 6000 times or more. Switching from position 1 to 4, and with a delay of about 9 seconds between every switch position and back and forth. So the basic with the motor works.
The force of the motor was about 2Nm and the switch can differ, sometimes 0.7Nm of force is enough to have it switch a position. (but have to test that yet with some kind of setup).

If first used a driver board from RS-online which was around 15 euro and blew that one up and the first Raspi too, don know why. But now i have 5 china-boards with the same driver ic 4988 for 5 dollar . They go up to max 2 Amps.

Hi sphere,

I really doubt that the software clock of a Raspi can time 1 millisecond.

There is a python example that also goes in ramps:
https://www.rototron.info/raspberry-pi- ... -tutorial/
They use 208 milliseconds cycles.
This will certainly depend on the driver. But with 1 milliseconds intervals you have disturbances alone by mouse moves on the fastest desktop ...

Interesting is your work at any rate. Thanks for reporting for us.

Hi Hermann,

yes i used info from the same link you pointed at for connection to the raspi.

I can check when i have time with an oscilloscope to see what the actual pulse rate is from the steps.

I followed your advise and created the standalone from windows, this now executes ok on the raspberry.
I just forgot to include the gpio library, but when i started the IDE and loaded the stack, The standalone started working OK.

THis is also an interesting board (as the Gert-board seems not to work on the raspi-3) the swiss-pi whcih you can plug onto the raspi: https://www.axiris.eu/en/index.php/i-o-cards/swiss-pi

Interesting is your work at any rate. Thanks for reporting for us.

No problem at all