rusefi.com

Current status: SD + TTL v 0.3 Ordered on March 5th 2014.

While stm32f4discovery has a micro usb connector which could be used for serial-over-usb, sometimes a second serial protocol is desired.

Based on FT232RL


usb_mmc_spi.zip
Schematic.pdf
BOM

Current status: New order placed, should get the board any day now.

This board has two connectors because depending on your situation you can snap it on at two different locations - either in the bottom-left corner of the discovery, or in the middle of the right side.

TODO: add exact pin ranges

russian wrote:One of the reasons for all this is to gather pieces for the future board which has to be our own board for the reasons of price & size.

Then I say definitely bluetooth. After it gets it's final resting spot it'll be nice to get data wireless.

I'd also suggest using USART1 or USART3 then you can program wireless also.

The boot loader is located in system memory. It is used to reprogram the Flash memory by
using USART1 (PA9/PA10), USART3 (PC10/PC11 or PB10/PB11), CAN2 (PB5/PB13)...

Opendous has a good SD card KICAD reference picture found here http://code.google.com/p/opendous/wiki/LPC313x_DIP_Board uses this connector http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=3M5607CT-ND

Does anyone know the height on a standard 2.54 female header like
?
And the height of an ANGLED male header like
?

I wonder if we can snap our communication border under the discovery using vertical female and have an angle male on the communication module - would this angle module fit under discovery?



In terms of the location that would place it under the audio jack of the discovery. So, question is the female header is taller than the angled header? On my amazing draft vertical female is black, the communication module is red and the male angles header is blue.

Update: the box is 9mm tall


and the vertical header is only 8.5mm tall, so this would not work

Perhaps of interest, see assembly technique found on this page, where the connector is put on the edge of a PCB.

Couple new ideas for this module:
we should have two 10x2 headers on it so that we can snap it on the discovery two ways: we can snap it to the left of the discovery using the bottom-left corner as on the original picture - this case our header would sit from PE13 to GND. We should also have another 10x2 on the other side - this one would let us span it to the right from the board center - that would be from PB6 to PA13.

We should also provide a 5x2 connector for SPI bus used for other external modules.


The pinout on the SPI connector should be
1 3.3V
2 GND
3 5V
4 CS1
5 CS2
6 CS3
7 MISO
8 MOSI
9 SCK
10 CS4

Left 10x2 header:
CS1 PE13
CS2 PE14
CS3 PE15
CS4 PD10
CS_SD_MODULE PD11

Right 10x2 header:
CS1 PD7
CS2 PB7
CS3 PD5
CS4 PD3
CS_SD_MODULE PD4

SPI1_SCK PB3
SPI1_MISO PB4
SPI1_MOSI PB5

USART3_TX PC10
USART3_RX PC11

Ooops, I see D2 now.

To add via, simply add trace to pad, then press V for via. Add the via at the edge of the pad, such that part of the via touches the solder pad. Hmmm, I see what you are saying. G1, G2, G3 and G4 should be added to the symbol. Then they will get a net name, then you should be able to add the trace, then you can add a via where ever you desire.

Schematic notes FT232RL, was there some reason to believe it would be the CP2102 or CP2103? It's my understanding that the goal is to use Seeedstudio OPL which specifies the FT232RL.

I'd like to stress silk screen. It's really handy for assembly purposes. Also labels on the headers is really handy.

Update: so the issues was NOT about the 10x2 module - the module is fine.

The issue was about KiCad generating two .drl file in this project - and apparently Oshpark only supports one .drl file per PCB, they have this warning in their FAQ but I have missed that.

Also when you get the holes thing ironed out, double check the hole size on pin 1, the square. Many default KICAD libraries have that hole slightly smaller than the rest, and it makes it a bugger to insert the connector. So it's worth double checking the pin 1 via.

russian wrote:where to get an SD module compatible with 'ST-TF-003A'.

So I was really going to give up on this module and I have started looking for some easy-to-solder replacement, only to realize that this module has other names - http://www.dipmicro.com/store/CON-MICROSD1 sells it as tfwf1-ebxx-1x0x and eBay sells it as TransFlash TF Micro SD Socket

Still a nice global part number would be nice.

We made some progress:
1) all footprints were right
2) SD card works (but it is a little bit unrealizable, not sure if the issue is with the card or the fact that I do not have a 4.7nF capacitor)
3) an issue: we've got the USB pinout upside down
4) another issue: even with the hacked USB wiring, the chip is recognized as 'Unknown device'. I have no idea if that's a result of the lame wiring or if I need to initialize the FT232 chip with some default firmware to get it recognized by Windows or if that's an issue with my Windows instance

PS: the green wire on the picture is soldered to the signal side of C353 on the schematic.

Guys, can you please research this? Do I need to initialize the FT232 chip? If not, why does it not ask for a driver? Is it because of the lame wiring on the previous picture?

FT232 requires a driver install.
http://www.ftdichip.com/Support/Documents/InstallGuides.htm

abecedarian wrote:FT232 requires a driver install.
http://www.ftdichip.com/Support/Documents/InstallGuides.htm

Sure thing it needs a driver. My problem is that Windows does not recognize it as 'TTL232R' - it recognizes it as 'Unknown device' and it does not want any drivers! Now my question is why would windows have any USB layer issues with this chip - could it be the wiring, could it be what?

I think "reset" pin 19 is buggered, See reference found here http://puma.freeems.org/docs/Puma_spin1_schematic.pdf on page 15. I know that circuit worked, and there aren't many differeences. The cable and the 47pF are differences, but I don't expect them to be that significant.

kb1gtt wrote:I think "reset" pin 19 is buggered

Active low reset pin. This can be used by an external device to reset the FT232R. If not required can be left unconnected, or pulled up to VCC

from here

So I guess you are suggesting connecting it to GND? But the datasheet has promised that we can leave it unconnected?

I've got the module working!

Reset pin was not the problem - wiring was the problem. It was not simply flipped, it was totally messed up - so, I've swapped the data wires one more time and this got the module up&running! Oh, and the RX/TX pinout on the connector was wrong, too. So, while rev 0.2 is workable, I would not call our biggest success so far

Hi rusEFI,
Could you please explain to me, How to jump or wiring for FT232RL .
I got some problem like this.

Thanks,
Mr.X

Current version of the module is know to work right away. Frankenstein communication module is same as this small PCB and again, it is expected to work as is.

didn't get if c358 is ceramic or electrolytic (dejavu?)
http://www.digikey.com/product-detail/en/CC0805ZRY5V6BB475/311-1371-1-ND/2103155
but on schematics it has polarity...

I use ceramic c358 - do they even have electrolytic in this package?

I agree that that's confusing. Same story with c354