Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[NEW HARDWARE REQUEST] Silicon Labs EFR32MG24 based Zigbee Coordinator USB adapter from ITead/Sonoff? #26

Open
Hedda opened this issue Dec 4, 2023 · 2 comments

Comments

@Hedda
Copy link
Contributor

Hedda commented Dec 4, 2023

@Daniel-zhan-itead @liangjia2019 @guozi7788 can you please forward a hardware request to ITead/Sonoff product managers?

I would like to suggest that you look into making and selling an upgraded variant of your ITead's Sonoff ZBDongle-E (Zigbee Coordinator USB dongle adapter) that would instead be based on a new Silabs EFR32MG24 SoC chip with more RAM + more Flash:

https://www.silabs.com/wireless/zigbee/efr32mg24-series-2-socs

Note that they also have various MGM240P (MGM240 radio modules based on EFR32MG24 with + 20 dBm power ampified transmitter):

https://www.silabs.com/wireless/zigbee/efr32mg24-series-2-modules

For development they have xG24-RB4188A (EFR32xG24 Pro Kit +20 dBm) and xG24-RB4187C (EFR32xG24 20 dBm Radio Board):

https://www.silabs.com/development-tools/wireless/xg24-rb4188a-efr32xg24-wireless-gecko-radio-board?tab=overview

https://www.silabs.com/development-tools/wireless/xg24-rb4187c-efr32xg24-wireless-gecko-radio-board?tab=overview

Also noticed that RF-Star already has a finished radio module ("RF-BM-MG24B2") that can be used for prototyping:

https://www.rfstariot.com/rf-bm-mg24b2-matter-zigbee-openthread-ble-multiprotocol-efr32mg24-module_p106.html

As well as the xG24-DK2601B (development kit for prototyping wireless end devices with only 10 dBm radio):

https://www.silabs.com/development-tools/wireless/efr32xg24-dev-kit?tab=overview

FYI, SparkFun Electronics released a similar MGM240P based "SparkFun Thing Plus Matter" (DEV-20270) development board:

https://www.sparkfun.com/products/20270

I believe some SoCs should be pin-compatible with their previous EFR32MG21 chips so I think in theory that you should be able to simply update your Sonoff ZBDongle-E relatively easily if you did not also want to add additional improvements to the board design.

Silicon Labs (sometimes also referred to as "Silabs") has released its new EFR32MG24 radio MCU SoC which is a direct upgrade replacement for its EFR32MG21 with SoC variants with larger flash storage, much more RAM and a faster CPU so those EFR32MG24 SoCs with more resources (especially RAM) than the EFR32MG21 you use in the current Sonoff ZBDongle-E model should be capable of handling many more Zigbee devices, and it in additional also on paper it offers improved reception and less noise sensitivity.

  • EFR32MG24 come with much higher RAM memory, with 128 or 256 KB variants, (compared to 64 or 96 KB variants of EFR32MG21).
  • EFR32MG24 flash storage has been increased to 1536 KB or 1024 KB, (compared 512, 768, or 1024 KB in the EFR32MG21).
  • Interestingly their similar ARM Cortex-M33 core runs at 78 MHz compared the 80 MHz core in the EFR32MG21
  • New feature is an integrated AI/ML accelerator NPU for edge interference which can potentially lower devices power consumption, though that AI/ML feature will likely not be useful when the adapter is used as a Zigbee Coordinator.

Note that there are variants of EFR32MG24 that have less RAM and Flash Storage or less dBm in power amplification, so be sure not use use one of those. You want to instead be sure to use one that has 256 kB RAM of data memory and 1536 kB flash program memory, (i.e. EFR32MG24B020F1536IM48 if do not want AI/ML acceperation). Also note that there are as well a few special variants that in addition offer onboard AI/ML acceleration (check out the EFR32MG24B220F1536IM48 variant in 6x6 QFN48 format) for not a lot of extra cost so might be worth getting that if available to offer a premium product.

Please keep the external antenna of the previous design as before, but highly recommend that you reconsider redesigning a new dongle for a such EFR32MG24 USB adapter so that it is a little smaller and with a longer USB plug so that there are no issues connecting it to all types of computers since your old Sonoff ZBDongle-E (and Sonoff ZBDongle-P) are infamously known for having a too short USB plug so will not work without a USB extension cable in many cases if the computer has its USB ports sunken too deeply. It is also too wide plugging it in directly normally means that the users can not use other USB ports that are too close as the wide enclosure of the ZBDongle design blocks other ports.

Keep in mind the board still needs to have good electromagnetic shielding -> https://en.wikipedia.org/wiki/Electromagnetic_shielding

Suggest that you call the updated variant "Sonoff ZBDongle-E Pro" or something similar to indicate that it is a more high-end version, however do continue to sell your existing "Sonoff ZBDongle-E" and "Sonoff ZBDongle-P" dongles as sure people will still want those too.

PS: Also recommend you consider always shipping at least a 1-meter USB 2.0 extension cable with good shielding with this USB dongle in the same package as it really makes a night and day difference in noise from EMF interference, (which is why the Home Assistant SkyConnect USB dongles ships with a 0.5 meter long USB extension cable, though it has been proven to work better with an even longer extension cable).

@wolfgang-muc
Copy link

I'm not able to flash the ZBDongle-E.

Silabs flasher (Chrome on Windows 10 or Ubuntu 22.04) shows imediately after selecting the file:
Auf darkxst.github.io wird folgendes angezeigt:
Failed to parse firmware: PythonError: Traceback (most recent call last):
File „/lib/python3.11/site-packages/universal_silabs_flasher/gbl.py“, line 118, in from_bytes
for tag_byres, value in parse_silabs_gbl(data):
File „/lib/python3.11/site-packages/zigpy/ota/validators.py“, line88, in parse_silabs_gbl raise ValidationError(„Image is truncated: tag value is cut off“)
Zigpy.ota.validators.ValidationError: Image is truncated: tag value is cut off

Setting the dongle in boot mode and start the transfer nothing happens.

@Hedda
Copy link
Contributor Author

Hedda commented Apr 12, 2024

@wolfgang-muc that is off-topic as nothing to do with this specific request, please create a separate issue so not to hi-jack this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants