Can there be multiple products with multiple imps?

[joesgarage]

Im new, forgive my ignorance - Im confused.

I have 2 imp-001 and (2) April demo boards. Both imps have been configured for wifi.
I created a project via the online guide, (Connected Product 1, with the development device group labeled as" test 1"). The 1st imp 0c:2a:69:00:43:23, was assigned to this, and coded a blinking led.
It works.
I installed the 2nd imp 0c:2a:69:00:39:c3 into this April board, and the same code was downloaded, and works. Great.

I now turned on both imps in their respective April boards, and created a 2nd product (Connected Product 2, with the development device group labeled as “test 2”).

I would like to assign 1 imp for Connected Product 1, and the other imp to Connected Product 2.
But when I click on “assign” for test 2 devices, and enter the mac address, I get the error:

1 device failed to move to test 2 development_devicegroup
Details
ID CODE STATUS TITLE DETAIL
0c:2a:69:00:39:c3 NF001 404 Resource Not Found No device found with that ID.
source: data[0]

Can this be done? Perhaps Im not understanding how the product development accounts work ?

Also, how does the product know what code to send to what imp ?

[hugo]

As you have worked out, the code is associated with the April (which has a unique ID chip on it - it’s a little SOT23 3-pin device right by the socket, connected to pin 6) - whatever imp001 you put in there - assuming it’s on your account, anyway - will run the code that is assigned to that April.

Hence, you need to use the device ID of the April when you’re assigning to the device group, not the MAC address of the imp.

The device ID of the April starts with “23” and is 16 hex digits long. You can see this in the “devices” sidebar of the code editor, or in the list if you click the “chip” icon (bottom icon in left bar)

[joesgarage]

of course. I now see the part on the schematic and this now makes sense. I had to re-blinkup this April and now it shows. 2 products now working with different blink code.
Thank you again !

[mlseim]

@joesgarage,
And if you want to move up to a newer Imp with more pins, the versions go up to Imp006. I really like the Imp003 because of the price VS the capabilities. You can get an Imp003 for $30:
https://www.digikey.com/en/products/detail/murata-electronics/IMP003-BREAKOUT/6205491
Size … it’s only a little bit bigger footprint than the April board.
Best of all, no SD connector pins to worry about (poor connections, etc).

Mouser has the Imp004 … for the same price as the Imp003 actually …
https://www.mouser.com/ProductDetail/Murata-Electronics/Imp004m-Breakout/?qs=u4fy%2FsgLU9PzSAZWTTwbIQ%3D%3D

Although not really recommended, I’ve been powering an Imp003 directly with a LiFePO4 battery w/solar charger … connected directly to the 3V3 power input pin.
See this project writeup: https://www.catpin.com/mailalert/

Note, I first tried a ESP8266 but it had poor WiFi reception. The Imp003 has amazing WiFi distance. Then I used a normal LiPo battery but it failed when the temp got cold. So I decided to use a LiFePO4 battery and since it’s nominally 2.3V, I just connected directly to 3V3 pin. It’s been running over a year now without any problems. The write-up tells the story. Schematics for all 3 attempts until I got it right.

You probably see a lot of articles about ESP32 / ESP8266, which is a nice WiFi controller for $10. But the Imps with the “agent” have a high level of security, ease of online programming, and have served me much better than the ESP family of controllers. Also, the ESP controllers are fairly good at POSTing to remote website API’s on the internet, but are very hard to work with the other direction. To access them online from a remote website is really hard because you don’t have an “agent”. Opening ports on your router to access them is not so wise. Dealing with the Arduino IDE can sometimes be frustrating when trying to use it for other microcontrollers like the ESP32. The IMP Central IDE is so nice to use.

[hugo]

Though 2.3v is definitely below the recommended voltage even for the imp003 (2.7v+ is the spec), the CYW43362 is specced to work at a lower voltage than the CYW43438 in the imp004m, so that probably accounts for why you’re seeing out-of-spec success.

One thing to note is that battery holders tend to add significantly to the operational impedance seen from the supply, so putting a big bulk capacitor (330uF+) across the power rail can help with this - just like we have on the battery powered sensor node reference design.