[Rood_Dragon]

what about using a Micro SD? Startech makes a lot of  adaptor that plugs in to a usb port (several styles)
http://www.startech.com/HDD/Card-Readers/USB/micro-sd-usb-micro-usb-adapter~MSDREADU2OTG

[TheOtherCliff]

We are interested in getting something like currently for flight data logging.
Blackbox/Openlog is also very interesting.
The problem is that to add anything at all to CC3D,  we need to remove / compress / move off some other functionality.  CC3D is also lacking in CPU power, making it less desirable for advanced features, even if there was room.
Any dev that wants to jump in to CC3D coding is certainly welcome to propose fixes that he is willing to implement. 

[peteng]

Is it possible to replace the main CPU on the CC3D with another (same footprint)  with more memory and higher clock speed?

[hwh]

As far as I know, no.   The 20 kb of ram on the STM32F103CB on the board is the maximum ram in the 103 family in that package.  Even if the package/pinout of a chip in another family with more ram matched the 103, changing family would probably mean that the firmware wouldn't run or wouldn't use the additional ram.

[Lob0426]

Some people are replacing the STM32F103CBT6 with the STM32F303CBT6. It has double the RAM at 40K and Double the program storage at 256kb. It uses a Cortex M4 rather than the M3. They are using Betaflight 2.5.4, so there is a firmware difference to make it work. They lose a UART as far as I read.

I do not know if they are accessing the extra Ram or Program storage.

I ran across this last night. I know this is an old post but thought you might like to know.
http://www.rcgroups.com/forums/showthread.php?t=2614433

[hwh]

That can be done but it's of limited use to most people.  Other than the difficulty of unsoldering/resoldering the cpu chip what you end up with is a new type of flight controller.  It's not a cc3d anymore, can't run the stock cc3d firmware, and requires programming as a new board.  And after all that effort you still don't have a board that's anywhere near as good as the current STM32F4 based boards.

Since cc3d boards with STM32F3 chips would never be common it wouldn't be worth the effort for the project to write and maintain firmware for them.   It's only of use to someone skilled enough to do the chip change and wanting to run non-standard firmware.

[Lob0426]

@peteng asked if a "CPU" change was possible, yes it is, But as you said it requires a new firmware. At the time you (hwh)posted it was not possible. It is now supposedly.

I could handle the chip replacement. I have a rework station. The Firmware would be a big challenge for me. Installing a bootloader, then writing a firmware. And supporting this is not worth the LP effort. The firmware would have to coded by those that want to try it with LP.

I thought it was an interesting post.

I think it would be better to redesign a new DIY board for the upgraded unit. Add an SD card for the logger and a Baro! LOL. I am all for a newer CC3D board with some added features!

[f5soh]

You can do all you want, all is possible.

But like Hank says, we cannot seriously include a new board that only used by a few users or just for you.

Add a new board means:
- New firmware: in this case start including STM lib not currently used.
- New Usb ID / different from the one used for CC3D
and all that comes in background:
- Gcs / Hardware page
- Artwork
- Setupwizard
and many other stuff i'm sure i miss...

For a result something hybrid and rickety.
No mag, no baro...

Destroy current CPU and solder a new one (STM32F303CBT6) don't take long in my opinion.

If you want to contribute, you're welcome. Even for current working hardware and already supported.

[nitromethane]

If I may add a couple of cents to this discussion, I think that for LibrePilot STM32F4 MCUs is the way to go forward (would still like to see some CC and CC3D legacy support) and that people should look at something as FalconF4 controllers (or RevoNano if someone ever decides to start producing them again...  ) as an alternative to the old CC3D controllers.

Then there's also the option that we try and make a Nano compatible (or very close to it) LP "official" board. Firmware is already done.

But yes, if someone wants to and can code some STM32F3 support in to the LP firmware, that would be sweet.