Page 4 of 4

Re: New STM32F3 devices (Cortex-M4)

Posted: Tue Dec 04, 2012 2:11 pm
by Giovanni
Small update.

The demo in the repository works (HAL, PAL only), now it is matter of validating the other device drivers except ADC and I2C that will have to be created from scratch (and the ADC driver will have to support linked ADCs on this one).

Giovanni

Re: New STM32F3 devices (Cortex-M4)

Posted: Tue Dec 11, 2012 9:44 am
by ojo
Hi,

I compiled the F3 discovery demo and found one minor bug which prevents compilation on Linux (but not on Windows):
Makefile, line 75:
LDSCRIPT= $(PORTLD)/STM32F303xc.ld
should be:
LDSCRIPT= $(PORTLD)/STM32F303xC.ld

to be able to link the executable.

Regards, ojo (who is looking forward to test this on the f3 discovery board soon)

Re: New STM32F3 devices (Cortex-M4)

Posted: Tue Dec 11, 2012 10:14 am
by Giovanni
Hi,

Thanks for finding, I will fix it as soon the SourceForge SVN repository is up again.

Giovanni

Re: New STM32F3 devices (Cortex-M4)

Posted: Thu Dec 20, 2012 12:19 am
by davidji
I'm interested in the oscilloscope idea. I destroyed a DSO nano v2 because it's not isolated. Until then, it was pretty handy. In any case, the STM32F3 can do two channels at 10Msps using interleaving - an order of magnitude better than the DSO nano. The required external components are two resistors (divide by 10) a capacitor (frequency compensation). You can overcome the isolation problem by hooking up the F3 discovery up to a Raspberry Pi, and then access the data over the network - the PI can be powered from an isolated supply and the ethernet is isolated up to 1.5KV.

I'd rather the connection to the PI was USB, but I gather the STM32F3 chip needs a new USB driver. Am I right about that? I'll give writing one a go, but haven't coded anything in C for a decade or so, and never done anything embedded, so don't hold your breath.

The rest could be pretty simple - the scope software would run under Linux (or Android via USB OTG). It requests a sample from the F3. The F3 collects as big a sample as it can fit in memory, and then returns it at leisure. At best, only half can end up on the display, but you get to work out where the trigger is retrospectively.

After that, I can try to work out how to run two ADCs in interleave mode to double the sample rate.

Re: New STM32F3 devices (Cortex-M4)

Posted: Thu Dec 20, 2012 9:27 am
by Giovanni
Hi,

The F3 uses the old USB peripheral already found on the F1 and the L1, the driver works already on the F3, there is a test application too under ./testhal/STM32F3xx/USB-CDC.

I also planned a specific ADC driver supporting interleaving. An open source HW/SW oscilloscope could become a really popular project IMO.

Giovanni

Re: New STM32F3 devices (Cortex-M4)

Posted: Sun Dec 23, 2012 1:07 pm
by davidji
I've built that project, and flashed it onto the board using openocd. When I connect minicom, I see the shell prompt, but when I press enter, the flashing led stops flashing and nothing else happens :(.

I'm using the latest from svn. Should I expect this to work at the moment?

Re: New STM32F3 devices (Cortex-M4)

Posted: Sun Dec 23, 2012 1:27 pm
by Giovanni
It works here so it must be a difference in the build environment, what compiler are you using?

Giovanni

Re: New STM32F3 devices (Cortex-M4)

Posted: Mon Dec 24, 2012 12:16 am
by davidji
Yes - updated summon arm toolchain and re-installed, now it works fine. Thanks.