USB Power Supplies and cables
This test is performed due to the email send by Athom after the first shipment of the new Homey’s
- The first shipment has received a power adapter which unfortunately makes Homey Pro’s CPU underclock. Read the e-mail here »
But also for users that have to replace their adapter or cable in the future. And I have read a lot of nonsense on Slack and on social media around it, so let’s see if we can clarify.
TL:DR: Test it and if you can measure it. Bigger or newer isn’t better!
In the hope to come with some list of working 
or not-working 
Adapters I started testing:
Not on the picture, also tested with a IKEA Tradfri Zigbee Repeater and a Raspberry PI 4 USB-C Adapter.
First of all what (I think) I know:
- Homey, or the Raspberry PI module (CM4) needs approximately 5.0 volt.
- Homey self (incl. the CM4) uses most of the time less than 1000mA even with LED RIG ON (never seen it higher, maybe it uses it for flashing Firmware or under heavy compute load. I have not yet load tested it yet.)
- A Adapter with higher Amperage isn’t necessary better. If it provides 5.00 Volt it probably doesn’t work no matter if it is 1A , 2A or 20A.
- Real USB-C Power adapters (for Laptops, telephones etc.) communicate with the device to negotiate a higher voltage (9, 15 or even 20 Volt) to speed up Charging. Homey doesn’t use that.
- Some (cheaper) cables drop more power (volt) and that increases with longer cables, thinner coperwires and bad connectors adapters (USB-A to USB-C of vice versa).
So please stop the nonsense about bigger, higher better.
The result is not easy, testing Homey probably plugged and started more than 50 times, testing some adapter more than once, on different ports, etc…
Using some conversion adapters (USB-A Male - USB-C female, USB-C Male - USB-A female, and two extra short cables) and the USB Tester introduces also a small voltage drop.
That extra loss is expected (See USB hardware - Wikipedia wat the expected powerdrop could be in a cable and connectors ) but also makes good testing complex. On the other hand: If it works WITH the tester some time it probably has less chance to fail without the tester.
First I was measuring at the power adapter as that was the easiest to place the tester. But moving the tester to Homey, clearly showed it displayed a sleight lower voltage there behind the 1.x meter USB-C to USB-C cable.
So I also monitored the videoCoreUndervoltageOccured to be False to check for at least some hours. That is where we are doing it for, a stable Homey Pro Early 2023!
Some test I have repeated 2 or 3 times to verify, most adapters that fail, fail already very soon after plugging in and booting. Often also tested a Restart from software.
Some tests with adapters with the USB-Tester and short cables worked but failed with the longer cables or with the Homey Pro Ethernet Adapter and a long cable.
Conclusion
There was no easy option that absolutely worked with the Ethernet adapter, test it with what you have.
-
The Raspberry PI 4 USB-C Adapter is working for me, even WITH the USB tester in between. (Not tested with Ethernet adapter as the PI 4 USB-C Adapter has a fixed cable with USB-C male connector.)
-
Most USB-C adapters stick with ~ 5.0 Volt, you can probably skip them.
- Most USB-A adapters have a slight higher voltage, around 5.1 to 5.2 Volt.
- Most USB (-A) Adapters have a lower voltage without load (eg 4.88V) and raise if you plug in Homey to fe 5.10 or 5.20 Volt. These look better (Also older adapters)
- Some Adapters with USB-C & USB-A or Adapters with USB-A 2100mA and USB-A QC perform better with the dump USB-A as it often had a slight higher voltage.
Note: The place I tested I didn’t have an ethernet cable, Test With Ethernet adapter where performed without using the Ethernet for now.
It looks like a more than 5 years old adapter with specification: DC 5 Volt +/- 5% 1000+/-50mA
is even more stable than most newer adapters.
Without Load it starts with 5.17 Volt on the tester (behind the Ethernet and long USB-C to USB-C cable) and raised on plugging Homey Pro in the Tester to ~5.29 Volt +/- 0.03 Volt.
If I missed something, let me know but most is just what I experienced.
If you have a alternative good working adapter (preferably that is easily available) Post it here below!
Brand-And-Name-Of-The-Adapter
And a good Picture!
