Custom Shield - MicroSD/FRAM

Hardware
141

Time for some lunch and make a new board :smiley:

But i have to work on re-routing the board. Good practice for me :smiley:

@bdub, can u try 2 x FRAM connected to SPI and see if there’s interference?

I see in the datasheet (another fram) that with 2, they hooked up the HOLD pin to the mcu as well.

UPDATE

Learnt how to desolder an SOIC-8 Chip and quickly fabricated a breakout board which only took 10 minutes.

I’m jealous that @BDub, board is 2 pitch narrower than mine :stuck_out_tongue: Kidding :smiley:

Have to clear some work before i test this out tonight :smiley:

1 Like
142

@kennethlimcp,

Your one step closer to Profection & Production ! We all learn from this as well ! Thanks @BDub & @kennethlimcp for sharing your success.

143

I'd be 99% certain it would be fine. I think what you are really testing is the code that breaks up the two chips based on address... and that looks good. That said, I can add another Spider to my breadboard and try it tonight. It would be interesting to see how long it takes to store and retrieve data, and it would also be interesting to see how we could allocate memory for buffers at runtime using these FRAM chips. That's probably more work than I can offer at this point though. You should see my list of things to do... I made some progress on the NFC shield library and testing so I need to get back to that.

144

@BDub,

I will try my best to test them out since i have devoted the Month of April to complete this Shield and i have 7 more days to push myself! :smiley:

Please work on the NFC so i can wrap this up and move on to it :stuck_out_tongue:

Cranking some numbers and formulas to wrap up another project before i can even power up my core X_x

*SCREAMS!*

2 Likes
145

I updated my Test Code above to fix an issue with error checking.

Was this:
if (addr > FRAM_ADDR_MAX) return -1;

Changed to this:
if (addr > (FRAM_ADDR_MAX * 2) + 1) return -1;

1 Like
146

@BDub,

With this working, i would like to try making a library from scratch. Have never attempted this but is definitely on my TO-TRY list and the FRAM has limited functions which makes it easy for me to learn!

Will update you guys when i have completed the library and would need to trouble you guys to help scrutinize it.

I’m thinking of some #define so that users can easily set the 2 x FRAM memory size

*launching Eaglecad and starts re-routing the board :D*

147

I just heard that squeal all the way over hear! :wink:

148

:scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream: :scream:

I hope i didn’t scare your children :smiley:

@wgbartley, you’re making me so busy due to the other Freeboard thread you started! Haven’t even had time to test it out but it’s on my chrome TAB :wink:

149

TL;DR

Status of this Shield - 80% completed.

What to be done:

  1. Make sure that the FRAMs are working fine through SPI
  2. Check through once more on the Pinouts on R4 since some changes are made (D2 & D3 for CS)
  3. Add in Solder masks and label
  4. See if more pads can be added for pins for option to change pins if needed (So far, pads have been added for all CS pins)
  5. Check pin compatibility with SRAM, FLASH etc.

UPDATES

This is R4 after correcting the FRAM pin 5 and 6 connection. Time for bed :clock1: :smile:

150

My favorite thing about this thread is the time difference between us. There's always lots of progress posted here when I check in the morning and early afternoon.

1 Like
151

Take this with a grain of salt… your non-45 degree tracks drive my OCD nuts :slight_smile: I’m sure it will work even if it was squiggly though. Just a standard way of doing things, for me.

I think your two solder jumpers look different. Might be good to put them on the same side with a nice silkscreen label J1 & J2 (jumper) or S1 & S2 (strap). I like Sparkfun’s solder jumpers:

152

Hahaha! Sadly I have OCD as well and I was just too tired to clean it up last night :stuck_out_tongue:

I spent more time making it neat than testing most of the time. Lol.

Do you know what’s the pad they use? Or I can download their board schematic and take a look :slight_smile:

Thanks for the input on this! Gonna look pretty with the solder mask and all :smiley:

I would love the jumpers on the same side but this might end up with more vias as some traces are on the bottom. I’ll take a closer look later

UPDATE

I’m toying with the idea of just leaving the HOLD and WP tied to high and not changeable by the user. Well i can do so but there’s so many vias that’s gonna be added for that. Thoughts?

Let me justify my decision.

For WP pin:

control writing to a status register

I haven’t thought of a scenario where you want to use a pin to disable the status register due to software not being able to handle it.

For HOLD pin:

This pin is used to interrupt serial input/output without making chips deselect. SO becomes High-Z, SCK and SI become don’t care.

Basically, using the CS pin, you will also be able to achieve the same effect on the FRAM. CS says: SO becomes High-Z. Other inputs from pins are ignored for this time.

Also, for HOLD to work, CS must be held LOW for it to work.

That’s another prerequisite and we will be better off controlling whether the FRAM is selected using their respective CS pin!

So…not user changeable :stuck_out_tongue:

153

Does the SD/microSD get cranky on Microcontrollers in general?

It reads the SD type but is unable to find FAT16/32 partition. I would like to blame the PCB i made than to know that this is commonly occurring.

But theres’ only like 6 wires to get the MicroSD card running and i don’t see how it can screw up the reading of partition…

What do you say? :smile:

154

kennethlimcp, have you tried slowing down your SPI clock? If it works at a lower bus speed then it may be your PCB. :smile:

1 Like
155

@peekay123,

any idea what’s the default speed?

156

Yup, default uses SPI_CLOCK_DIV4. The latest pull request I made to BDub for the microSD library allows for full speed (DIV4) and half speed (DIV8). In your code, after the call to SD.init() or card.init(), add this line to change the SPI clock:

SPI.setClockDivider(SPI_CLOCK_DIV16);

You can try DIV8, DIV16 or DIV32. If none of those work, then there is a different problem.

157

I am curious about how everyone uses the headers on the Spark. Does everyone just use the male headers supplied with the board or does anyone use female headers with the long pins like an Arduino so that you can stack other boards or jumper wires on top?

158

Since the Spark has an antenna, and ships with male headers, it’s mostly been the “top dog” in the shield stack up so far. I think if you did it the other way around it wouldn’t be as useful in the breadboard, and you might cover the antenna too much. I usually just plug it in a breadboard and then I have two connection points per pin to work with… even a mini breadboard works well.

159

@peekay123,

In the Spark-CardInfo.cpp it uses if (!card.init(SPI_HALF_SPEED, chipSelect))

So i’m on DIV8.

I checked the Arduino forums and seems like there’s always some topics popping up about the FAT16/32 partition not being found…

But since you guys did get it to work and i did get it working few days back, I’ll hook up a new board and test it out.

It’s weird since i didn’t do anything to the board except to remove the FRAM…

160

kennethlimcp, if I got a nickel for every time that happened to me… :wink: