External (non-USB) power supply for Teensy 2.0: General thoughts, bypass Caps?

Status
Not open for further replies.
Hello, beginner question here:

I want to supply several boards, among them two with a Teensy 2.0 in a socket, with some common, reasonably dimensioned 5V power supply,
probably some 5A/5V Meanwell device. I don't know the full power draw yet, but I suppose it is above the 500mA that work painlessly over USB.
Probably somewhere around 2A at 5V.

I will cut the 5V pad on the teensy, to split up USB power and the rest (USB will go to a Raspberry Pi probably powered from the same 5V power
supply, so don't know if this is strictly necessary, I just have to access the serial port from there).

Will I need some bypassing capacitors as soon as I cut the pad, so that the Teensy power stays stable, like most people recommend for logic chips?

What would be the best way to power 1 Raspberry Pi, 2 Teensy 2.0 (one draws quite a bit of power for LEDs, the other less so) and some
components linked to the Teensies from a common 5V PS? I don't think I can feed enough power over the USB, even less so from a Raspberry Pi as the USB Host.
If I cut the pad, connecting a good regulated PS to the GND and 5V pins of the Teensy probably most sensible, or am I wrong here? Also: Are
the usual Meanwell Power supplies (in perforated enclosure or plastic brick) good enough to power the Teensies in the long run without
frying them, or will I have to add another regulator?

TIA
Ralph
 
The deeper question here is what else is being powered with the Teensies and Pi. If all you have is the assorted T2+Pi running along in a steady state most power supplies will produce the target output without problem, since they have onboard caps designed to cope with the generally pretty erratic USB 5V.

Where this gets tricky is if you have say a massive motor, LED wall or other highly variable demand. These have the potential to cause the supply voltage to rise and fall as their demand changes. In which case decoupling caps to stabilize supply to the microcontroller parts will improve things, especially if you are using analog read since supply changes can show up as apparent changes in measured value. Better quality supplies do better at regulation, and if things are really noisy sometimes you need multiple supplies. The basic test would be to run any high demand item from the power supply with expensive stuff disconnected, aim for worst case changes in load and measure the 5V rail for changes either with a scope or with a mulitimeter looking at both AC and DC. It is also possible to build simulations of your power supply and loads, but possibly overkill for this sort of project. In anycase not one that can be answered without access to the actual hardware to be used.
 
Status
Not open for further replies.
Back
Top