I have a board that comprises a Photon with headers connection (ie it has onboard regulation) and a 5V fan circuit that provides DC-DC regulation via PWM. The question is whether I need / should use a regulator if I intend to use 5V USB power supplies (ie 5V@1A Apple wall wart)?
Logically the Photon board happily will regulate that down to its 3.3. My circuit cleans the 3v3 out to make it similar to the old Core 3v3* ref. The 5V fan will happily take whats given to it, ie whatever the USB wall wart gives it.
The issue with a regulator is they rarely output what they receive, ie I’d need more than 5V input to get a full 5V for the fan to operate at 100%. I could live with 4.8v out, but then I’m likely caught trying to source a low dropout regulator.
So, should I skip using a regulator?
Should I apply any filtering or caps?
If that’s all the current you need and your always going to use a wall wart USB adapter then what would be the benefit of adding another regulator on top of the one that is built into the wall wart and the Photon already?
You shouldn’t see any voltage drop on a 1A wall adapter.
If the fan speed regulator is indeed PWM, it likely does not require much headroom.
If it is designed to run off 5V, then that’s what it’ll do.
I would suggest a mix of decoupling caps on the 5V rail, because the fan controller is likely to generate a good amount of hash on the 5V line if you let it. I’d suggest a range of values of cap, all installed with short leads as close to the fan controller as poss. Start with 3, say: 10nF, 100nF and 10uF.
Avoid ceramics for the 10uF part, because sometimes with long (inductive) leads to wall-warts you can introduce other problems, I’d just use a standard aluminium electrolytic for that.
The photon should be quite happy regulating down to 3.3V, as you say.
I’m not sure if I’m understanding the problem correctly but one thing that it might pay to check is the data lines on the USB wall adapter. I’m not sure what hardware is inside apple USB chargers but some of them perform a USB ‘handshake’ over the dataline to negotiate current draw. There is a few different standards for this BC1.2, USB2.0, USB3.0 and proprietary Apple 1A/2A.
I designed a USB charger at my old company for one of their products and it had an integrated control chip, I believe if there was no resistance on the data lines it would default to 500mA, maybe less. Furthermore, I imagine the photon could easily draw over 1A during data bursts so as AndyW says, it could pay to have some local capacitance.