build STM32 evaluation board for less than $20

Posted on May 17, 2011 at 12:21

I found convenient way how to build simple STM32 evaluation board.

I've used Schmartboard EZ prototyping board (mouser.com, $9.99) for 32 to 100 pin 0.5mm pitch TQFP. Soldering was easier than I thought - all you need is to do is heat presoldered grooves on board - it took me less than 15 minutes to solder 100 pin STM32 though I've never done SMT soldering before. I've added 8 power capacitors as recomended by ST (SMT or with leads), 3.3V regulator, 8 pin JTAG connector (20 pins standart is too big for this 5x5 cm board, I have 20 pin to 8 pin JTAG reduction), Boot 0,1 jumpers and USART1 connector. Crystal is optional, hovewer it fits nicely together with its caps. There was even room for few LEDs.

I was little worried about JTAG signal integrity, but it works without problems with parallel port Wiggler compatible JTAG at the maximum speed (400 kHz clock).

No problems with UART flash bootloader either - so JTAG is optional as well, if you are willing to debug with printf to serial line :)

I'm using eval version of Crossworks, works without problems(debug, flash programming). I've tried build in HW library examples, works well with one change - I had to increase timeout for HSE crystal oscillator start up time (it is crystal dependant).

I'm planing to use this board in small robot, so I've added four 2x13 pin headers to have more convenient acces to all 80 IO pins.

Posted on May 17, 2011 at 12:21

Hi ivan,

Interesting hand made board ... I like it ;) so much, you can have the same with lower cost using QFN36 or TQFP48 or 64 packages,

Here a small board developped for STM32 ( business card size :-). very low cost without quartz and using the internal RC and some capacitors with LCD + SWD connector.

Cheers , STOne-32

Posted on May 17, 2011 at 12:21

These are pictures of two boards:

board1 is with four headers. I have just unfinished board picture- unfortunately I've send 9V directly to chip so I've burned it and the board is away for chip removal.

Meanwhile I've build board 2 with addon PCB (7 segment LED, USB, UART, Boot jumpers,3.3V regulator + space for other stuff)

Posted on May 17, 2011 at 12:21

Hello Ivan,

can You post a photo of Your ''selfmade'' evaluation board? It might be interesting to see how it looks like.

Best regards

Squonk

Posted on May 17, 2011 at 12:21

''Ahh, i see, its a ST new product announcement!!

A ST32 with LCD driver on chip! ''

hmm could be, or maybe they are direct driving it, but too many segments for that

anyway that proto board stuff looks neat!

Posted on May 17, 2011 at 12:21

Dear all,

I see that board is creating some interest :) This is an early prototype

where STM32 is driving 128 Segment (Quadruplex drive) Glass through GPIOs and can also drive bigger LCDs with more segments and operating only from a CR2025 Coin cell Battery for a long period : more than 2 years :p

STOne-32

[ This message was edited by: STOne-32 on 17-01-2008 14:19 ]

Posted on May 17, 2011 at 12:21

''obtronix, there is no problem with segments with multiplexed mode LCD. Funny about this LCD board is that I'm developing the same thing here, but I see that at least one is ahead of me. ''

But don't you need a charge pump to bring the voltage level up on the coin cell output (to meet the LCD turn on, turn off levels near end of life of the coin cell) and voltage dividers for the bias (bunch of resistors -- board space), and a bunch of specialized software to drive all the waveforms. To get it all to work over temperature (you may need software contrast control) and within the narrow LCD voltages levels to meet life requirements, well it's not all too easy, you have to have a good understanding of the particular LCD fluid in your display, as well as the glass used etc.

Posted on May 17, 2011 at 12:21

thank I never saw that one, but I have look at simliar applicaton notes from other vendors.

my concern is not available throughput, but

''The problem is DC voltage on the LCD segment if something goes wrong in the software. Debuging and hence stopping SW control is not recommended. ''

that implies that you can never go into standby mode, so average current consumption will be high, whereas micro's with built in LCD controllers can run the LCD using 20 or so uA.

Posted on May 17, 2011 at 12:21

obtronix, have you seen

http://www.st.com/stonline/books/pdf/docs/6292.pdf

ST7 application note?

Processing power needed to control LCD in software is not problematic and with Cortex speed and very fast IRQs... you will not notice it. :) The problem is DC voltage on the LCD segment if something goes wrong in the software. Debuging and hence stopping SW control is not recommended.