3V3 pin powered by USB port on an IMac. According to the Microchip data sheet for the Rn42 it draws a maximum of 30ma while transmitting.
Thanks for the detailed explanation!
Good to hear, was this with Wi-Fi + Cloud enabled?
3V3, tried both via laptop and 5V 2A (Sony phone adapter) but had same results
Also just tried 3V3* just now but no luck as well.
Wifi enabled with TCP, no cloud connecting an Acoustic Guitar to the Internet
Added in another RN42 from a different Arduino project (had to unsolder but oh well shit happends) which is 5V, did some rewiring and it doesn’t lose Wi-Fi with that one so it looks like its power related.
Gonna do some more testing to make sure
/edit: stress tested and remains working, gonna use the 3V3 converters like @ScruffR suggested and power the original RN42 module separately.
Still find it a bit weird that it can’t even supply ~35mA constantly on its own, hopefully the Spark Photon will be better
Maybe due to the constant cloud connection it just uses too much total?
/edit 2: now it reverts to breathing blue again, remained working for a few minutes so hopefully just a simple glitch
30mA doesn’t tell the whole story, it is likely the transients that are killing you, as the thing goes to transmit, it would not surprise me that you see 300mA spikes caused by the TX firing up - of course these are very short duration, then it settles down to the 30mA they talk about in the datasheet.
Decoupling caps are your friend here.
The CC3000 is doing the same evil shit on the core’s 3V3 line, so if they happen to do it around the same time, it gets pretty ugly (another problem of coexistence.)
Experience says you’ve got a combination of RF and power problems, and you’ll need to solve both to have a reliable solution.
The good news is that the Proton will support a coexistence interface to let BT/BLE and the WiFi play nicely together.