Mail Archives: geda-user/2012/07/24/19:53:27
Edward Comer:
> My md5sum for stm32f100lm.pins failed, so I re-downloaded it and then md5sum passed. When I ran it, this was my output:
>
> ecomer AT pennyroyal:~/pcb_projects/diy-symbol/pintosym$ ./pintosym.pl stm32f100lm.cmd
> printing to stm32f100lm.adc.LQFP100.sym (adc_pins cutout r:DAC l:ADC:s w:8)
> Use of uninitialized value in concatenation (.) or string at ./pintosym.pl line 886.
> Use of uninitialized value in concatenation (.) or string at ./pintosym.pl line 886.
> printing to stm32f100lm.adc.LQFP48.sym (adc_pins cutout r:DAC l:ADC:s w:8)
> Use of uninitialized value in concatenation (.) or string at ./pintosym.pl line 886.
> Use of uninitialized value in concatenation (.) or string at ./pintosym.pl line 886.
> ecomer AT pennyroyal:~/pcb_projects/diy-symbol/pintosym$Â
Hmm, strange I don't get thoose complaints about uninitialized value.
I have to check that out later.
> In that directory I now see some very attractive sym files with
> simulated tear marks at the top and bottom, named
> stm32f100lm.adc.LQFP100.sym and stm32f100lm.adc.LQFP48.sym,
Great, thank you, that is good to hear.
> which I assume are two parts of the same chip.
No, it is basically the same chip, but in different packages. One in a
lqfp100 and the other in lqfp48. So it cannot be parts of the same
chip.
It just shows that you can start with the lqfp48-sym, then when need
araises, exchange that sym for a sym for a bigger package/chip without
rearranging the nets in the schematic, the pins will be in the same
position.
If you compare the two, you'll see that one has less pins, and that
the pin numbers differs.
> I have only been peripherally following the conversation about
> breaking up a complex chip in this way so as to make the schematic
> more readable.
> Will gschem and gsch2pcb correctly handle these two symbols as a
> single chip in PCB??
My understanding is that for the geda suite to treat it as the same
chip in pcb, it must have the same footprint and the same refdes.
///
If you replace the "!" lines in the .cmd file with e.g.:
!f pwr_pins tt pins $type eq "pwr"
!m jtag_pins tt pins (JT|NRST)
!m misc_pins tt pins (OSC|BOOT)
!j sys_pins pwr_pins jtag_pins misc_pins
!> .sys.%l.sym 3 sys_pins rect jtag, boot and crystal pins
! arg l:OSC,BOOT r:TRST,TDI,TMS,TCK,TDO,RST b:.
!m adc_pins tt pins ADC
!> .adc.%l.sym 3 adc_pins cutout analogue input and outputs
! arg r:DAC l:ADC:s w:8
It will generate two sym's for the same pkg (3 = LQFP48) from
different pin sets (sys_pins vs. adc_pins).
If you use booth theese sym's (with the same footprint and refdes)
in the same .sch file, it "should" turn up as a single device in pcb.
I did some tests, but gsch2pcb complained about a refdes U? which is
nowhere in the schematic. I tested this with refdes=IC1. Maybe it is
related to:
http://www.delorie.com/archives/browse.cgi?p=geda-user/2012/07/19/16:06:08
///
Now I see the pin file is strange:
$ grep OSC_ stm32f100lm.pins
12 5 C1 5 pas OSC_IN
13 6 D1 6 pas OSC_OUT
81 5 C1 5 pas PD0 OSC_IN
82 6 D1 6 pas PD1 OSC_OUT
it has duplicate lines for OSC_IN/OUT for the smaller packages.
Page 25 and 27 of
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00251732.pdf
have the same data, which means I copied it correctly.
Maybe I should include some check for duplicate pins.
Regards,
/Karl Hammar
-----------------------------------------------------------------------
Aspö Data
Lilla Aspö 148
S-742 94 Östhammar
Sweden
+46 173 140 57
- Raw text -