Particle Photon rapidly alternating between white LED and D7 LED when powered with VIN pin

So I have a particle Photon that I’ve been using to interface with a PLC. It has been working fine so far, code is being flashed successfully and it is sending data up to the Particle cloud. However, the Photon only works when it is being powered through the microUSB port so if I want to run it at all I need to have it connected to a computer or some other type of 5v USB source.

This won’t work as I want to run it off of the 24V power supply inside of the PLC cabinet. But when I connect it to the power supply (through a 5v L7805CV regulator) to the VIN pin, it works normally for a few seconds - going through setup and eventually breathing cyan to indicate that it’s connected to the cloud - and then suddenly starts flashing the white LED and the D7 LED rapidly.

I’ve tried switching out the 5v regulator with a 3.3v (LD33V) regulator, I’ve tried using the 3.3v regulator to power it though the 3.3v pin, and I’ve tried using the 24V source from the cabinet and a benchtop power supply as the input to the regulators and each time the result has been the same.

However, when I connect the photon directly to my benchtop power supply set at 4v, it works fine.

Why is this happening? I’d rather use a voltage regulator to supply the photon instead of a DC-DC converter or something as it’s less hassle and takes up less space. Am I doing something wrong?

Have you checked the quality (noise, stability, max. current, …) of your 5V supply?
A video of the issue and a schematic of your setup would be good to.
Is your 24V supply DC?

Here’s a video of what’s happening
The schematic is very simple, you can see it in the video. It’s just power going into the regulator and then having the output connected to Vin on the Photon.

I have not checked the quality of my supply, would that be in the datasheet or is that something I can measure? I believe the max current is over an amp, so it should be enough to power the photon.

Does the regulator and/or any component on the Photon get rather warm?
What is the noise that starts at the same time your Photon starts playing up?

You should add the recommended caps to the regulator.

I’d check with an oscilloscope.

@slop_jones, using a linear regulator to go from 24v to 5v is a large leap. That makes for 19 volts times the current consumbed by the circuit of lost power being dissipated by the regulator. At a typical 40-80ma, that makes for .75 to 1.5 watts!!! Chances are the regulator is thermally shutting down. The input and output capacitors are SUPER important as they will prevent the regulator from going into oscillations of death. I assume you followed design guidelines for type and value for these caps.

I highly recommend getting a switching (buck) regulator to produce your target voltage (5v or 3.3v).


I’ll echo what @peekay123 said.

Yes, the 5v regulator is getting very hot. The noise is probably the giant air compressor that I’m next to.

It does seem like the regulator is oscillating and making the particle continually restart. I don’t have any input or output capacitors on the regulator so that may be something that I can try. I’ll see if I can get my hands on a buck regulator and try that as well, because I don’t like the idea of all that wasted energy.

1 Like

So I added capacitors to the regulator as well as a big ole’ heat sink. I have two 0.1uf capacitors in parallel on the VIN leg to ground, and one 0.1uf capacitor on the output leg to ground. Based on the picture below.

The voltage regulator still gets very hot. ( my power supply was reading 0.09A 24V ) It worked for longer than it did without the capacitors and the heat sink but it still eventually oscillates in the way it was earlier.

Am I doing something wrong here or is using a voltage regulator like this inherently not going to work?

It really isn’t tenable. As @peekay123 pointed out - you are converting approx 19V @ 90mA into heat, that’s about 1.7W. You might be able to do it with a much more efficient heatsink, and/or forced airflow, but really, you’d be far better using a switched mode unit instead.

I’m sure a bunch will chime in here with their favourite unit from ebay/amazon/adafruit/wherever.

These work in a fundamentally different way, and do not simply burn the unwanted voltage off as heat.