RPI module CM3

What do you think?

+1 for WinIoT compatibility! Now if we can get it on BeagleBone… :whistle:

What do you mean no SD card support? Did you see the CM3L version?

Sorry, the lesdixia is kicking in… However, you did write… “only sdcard support no flash store”. So, that seems wrong, too. The CM3 version has 4GB eMMC but no SD card support while the CM3L has no eMMC and SD card support.

They kept good on their promise. This is wonderful. A BBB module should be easy enough to do. It’s just a specialized board that would carry the existing module. :clap: :dance:

My only gripe is concerning the Dev kit. I think they should have placed everything the Raspberry Pi 3 has on the kit.

Hmmm… So, GHI could create an Octavo module that’s pin compatible with the CM3? I’m having a hard time thinking of a scenario where that would be beneficial (except to GHI & Octavo…). I haven’t taken it to the point of comparing pins but I suspect you wouldn’t have access to the best parts of the BBB (the PRUs…) if you did that. An Octavo module would probably cost double what the CM3 costs… So why?

No need to do this with the Octavo as by design it is relatively easy to use the device itself on your own custom board. The large BGA pitch and the fact that all the high speed design is kept internal means you can easily build your own custom solutions with little effort on a 4 layer board.

The GHI OSD3358 SOM (if it every gets released :whistle: ) will be a great solution for those who don’t want to do BGA routing.

My own gripe with the Raspberry Pi was the lack of real UART ports. The original Pi modules, including the Pi3 only have 1 UART

I see that this new module has 2 which is better but still not enough. Embedded designs by nature tend to still use legacy UART ports for a number of external devices. Card readers, RFID readers etc are generally UART based.

Of course, you can add USB to UART but that makes hanging bits off your design which is not ideal and also bumps up the cost.

@ Dave McLaughlin - Pair it with a G30?

Would be easier to just slap on a UART HAT.

1 Like

Actually a SPI to UART bridge IC would be cheaper.

I suspect that’s essentially what the UART HAT is. You can’t just lay an IC on the Pi, though. Of course it’s more expensive to buy the HAT. Depends on your need.

1 Like

Then you need to write and debug code for that part. No thanks. :stuck_out_tongue:

An SPI UART would make more sense but then you need a driver for this whereas native ports are already supported. Of course, you could use a USB to UART bridge on the board and add a USB HUB but that ramps up cost and complexity to the board design.

@ Dave McLaughlin - Exactly how many UARTS do you need anyways?

How about a USB to Quad UART bridge?
[url]http://www.ftdichip.com/Products/ICs/FT4232H.htm[/url]

4 minimum as follows:

  1. MODEM - direct TTL serial link. No USB as it is embedded
  2. Modbus RTU - Slave for client data access
  3. Modbus RTU - Master for communication with VSD
  4. TTL serial for Zigbee interface

@ Dave - The specs on this new board made me think of you.

2 Likes