Using bstc/bstl with Fedora

aderfy

Hello,

I'm trying to download some code to the propeller chip using either bstc or bstl under Fedora and a normal serial port (not USB to Serial adapter). The serial comunication seems to work however both programs hangs forever and doesn't seems to have any reaction on resets or power off/on cycles.

So far, what I see is:

When the PC is booted for the first time, both bstc and btsl exits immediately (seems some sort of timeout).
Then I started minicom with a serial speed of 115200, I can type characters and see the echo back. I also disconnected one of the serial port pins to be sure that it is the propeller chip that echoes back the character and it is, so I guess that the serial connection is correct.
After that both bstc and bstl hangs forever trying to do I don't know what.
I tried bstl -t (test undocumented), I see that it echoes back whatever I type prefixed with a 'b', so again I guess that something is working correctly.

Is there something I can do to understand what's wrong ? Is there some debug setting that can help to understand where it hangs ?

Regards.

Peter Jakacki

Run BST from a Linux terminal and you will see you are missing some lib files for sure. I can't remember which but use the terminal and if you don't have problems with finding and installing lib files then all will be well.

aderfy

No, I'm using the command-line tools bstl and bstc not the GUI IDE. There are no errors about libraries. I'm sure that it is something with the serial port connections but without any debug output it is impossible to understand what.

jazzed

Can you try writing a small program to EEPROM instead of RAM ? It's possible that the program is actually being loaded but propeller can reset before the program starts. If you put the program in EEPROM it should boot regardless of a loader induced reset problem.

bstc -p2 -d <port> <program>

aderfy

But the program (bstc or bstl) should not exit after downloading the program ?
Also bstl without a binary to download should just identify the chip, it never does that.
I'm just trying to comunicate with the chip.

And where are the sources of these problems, are they open sourced ? It seems not.

aderfy

Solved!!!

It turned out it was the crappy RS232 schematic I was using to program the chip. Thankfully the chip wasn't fried after all the attempts.

Here is the schematic. It is identical to the datasheet except for the DTR/RES pin that is using a third transistor just like the TX pin. Just in case anyone is interested: