My needs are best suited by an F4 Mini with on board 100Mb Ethernet and no USB. So, I would say a Mini with Ethernet and an Arduino compatible carrier with USB.
Discussing new STM32F boards
(26 posts) (9 voices)-
Posted 3 years ago #
-
MediumKahuna - the smallest STM32F4 package with Ethernet is 100pins, a Maple mini is 48pins, so the board is going to be quite a lot bigger, with many unused pins, or double rows of pins. Could you prototype your needs with an STM32F4Discovery board?
Posted 3 years ago # -
Every time I'm trying to rethink "the development board", I'm reminded of this board called "number six" designed by Noah Vawter I think around the time that Arduino was created.
http://dl.dropbox.com/u/42394/numbersix.jpg
http://dl.dropbox.com/u/42394/numbersixbb.jpgI think it was a pretty unique, neat design. The 40-pin connector worked with the standard IDE 0.1" ribbon cables. The 40 pins map one-to-one with the 40 pins on the atmega chip. You could either plug a ribbon cable or a little breadboard adapter that's shown in the second image.
AnywayPosted 3 years ago # -
That 'number six' looks very nice.
There are other ways to make a small breadboard footprint. For example:
http://www.ebay.com/itm/Microchip-64-QFP-2-DIP-Breadboard-PCB-prototyping-board/250958560409
http://www.ebay.com/itm/Microchip-80-QFP-2-DIP-Breadboard-PCB-prototyping-board-/260443966844This might make it pretty straightforward to design a compact breadboard-friendly footprint with lots of other connectors and features. Of course, the volume is big, but it could be on a 0.3" pitch.
I should have a couple on Monday at our next Micromouse meeting, and one of the folks may have tried it already.
Posted 3 years ago # -
@gbulmer, very interesting design. Sad that pics don't show other side of the board, I'm very curious how other side is soldered.
Posted 3 years ago # -
siy -
Sad that pics don't show other side of the board, I'm very curious how other side is soldered.
Are you asking about attaching the pins that go into the breadboard?
AFAIK, there are two sets of right-angle header pins. These are asymmetric. So one side they go in, say, the lower row of holes, and the other, the higher row of holes.
I expect to have some Monday, so I should be able to post some pictures.
Posted 3 years ago # -
My initial impression is the same, but this does mean other ends of pins will not be in the same plane (i.e. one row will be longer than other). Of course it is always possible to cut pins a bit. It is possible that plastic holder for other pin header is moved from short end of pins to long one and then long end of pins is soldered into PCB. Probably this will allow avoid need to cut pins. Need to check this...
Posted 3 years ago # -
siy - I believe the two sets of pins are put in the opposite way round. I found pictures of an earlier version several years ago, and that is my recollection from those pictures. That isn't a guarantee, but it made sense then.
The pins are not cut, and the asymmetric length of the right-angle pins is used to get the pins (roughly) to the same plane on the breadboard by putting the short end of one set into the board, and the long end of the other set of pins into the other board holes.
I do like Noah Vawter's "number six" posted by Okie. That might be easier for inexperienced people to make.
Obviously it would be wider, and more bulky than that ebay breakout, but I think 0.6" pitch or less would be fine for most folks.
It could plug straight into a Micromouse 'chassis'.Posted 3 years ago # -
Agree, row pitch compatible with standard DIL sockets is the best choice (i.e. 0.3", 0.4", 0.5", 0.6", 1.0", depending on number of pins).
Posted 3 years ago # -
I may be a little late to this given the last post was a month ago and Olimexino was building prototypes before that!
However, one thing I'd like to throw into the mix is CAN support.
As mentioned above the STMF4* gives lots of nice things such as on-board floating point and a very high processing speed. This makes it a contender for advanced control and techniques, and useful for high-throughput applications.
One of the fields that requires high throughput and advanced control is automotive engineering, which also makes widespread use of CAN.
Can you guess which industry I work in yet?
In any case, I think an argument could be made for at least facilitating access to the CAN peripherals of the MCU (if not actually putting a CAN transceiver on the board).
Posted 3 years ago # -
I saw the PCBs for Pete Harrison's STM32F4-based micromouse robot yesterday at MINOS 2012:
http://www.micromouseonline.com/2012/03/22/decimus-4-beginsThey looked extremely good. He got them made at Seeedstudio:
http://www.seeedstudio.com/depot/fusion-pcb-service-p-835.html?cPath=185He paid extra for the blue solder mask and gold plating.
As you can see, that PCB service will do quite complex profiles, including large slots in the board (for the motors).
Base price is $9.90 for ten double sided 5cm x 5cm boards. $24.90 for ten 10cm x 10cm is boards. Plus postage, which is $8.10 to the UK.
Unusually, Seeed will let you 'panalize', ie. put several sub-boards on each board, but you have to cut it up yourself.
Posted 3 years ago #
Reply
You must log in to post.