Powering my Rpi Retro Arcade

Hi, all!

I started a project this summer, and actually got tweeted by Hackaday about it, which was really cool, but I simply haven’t been able to finish it.

Here is the link to the project, if you’re interested: https://hackaday.io/project/14623-raspb … etroarcade

At any rate, the entire thing is technically operational, but I keep having issues with supplying power to the device.

The raspberry pi keeps sending me the ‘low power’ colored square in the upper right corner. At first, I thought my wall wart simply couldn’t supply enough current, since it is attempting to run a 10.1" TFT LCD and RPi all at once.

I upgraded to a 5V, 3A power supply, and it still gave me these warnings, so I dug a little deeper:

When everything is hooked up and the device is at full throttle, it only draws about 1.4A. So current draw isn’t the issue.

The interesting thing is that when no load is attached to my power circuit, the input and output voltages are about 5.2V, but when I attach the RPi and LCD, the voltage drops to 4.82-4.86 V, which isn’t quite enough to support the Pi.

Why is this happening?

Attached is a schematic and an image of the power circuit.

Anything helps. I’m at my wits’ end here. Thanks guys!!

Perhaps use an adjustable regulator to supply a couple tens of a volt more?

I thought that I couldn’t boost a voltage using a regulator/op-amp any higher than my input voltage, which is 4.86 V…

To clarify, you mean something like an LM317?

Yeah, but you will have to be extremely careful about going too high. Your current input of less than 5 volts might be the issue. Surely your components can tolerate a tenth of a volt over voltage. A ten turn pot will let you dial in a more precise output than a simple trimmer. You could power up with you DVM in place and adjust until you get to a steady 5.1, or at a lower voltage that will make the undervolt warning go away. That relay might be adding enough resistance to account for the voltage drop. Another, probably safer but more complex solution, would be to power the Pi and LCD with a separate power supply. Just throwing ideas out there.

Ok, cool! I’ll try it out and let you know if it works.

I’m trying to keep it as one power supply input for now. Thanks so much!

I could be wrong but I’m pretty sure that no regulator is going to boost your voltage.

In fact, from the TI LM317 datasheet the input voltage needs to be a minimum of 3 volts higher than the output voltage.

I had assumed, quite incorrectly, that his power supply was being stepped down via an LM 7805 or equivalent. After re reading his first post the adjustable regulator indeed won’t provide a higher voltage from a fixed 5 volt supply. You are, of course, quite correct. Most of my power supplies are 12 volt so my assumptions are driven off that. Still, my bad. It’s just so much easier to step a voltage down than to boost it up. In light of this I suggest a 12 volt supply in the 5 amp range and a UZ1085 adjustable regulator. They can output up to 3 amps and are very stable.

So I can do that, and probably will, but just so I know, why is the voltage dropping down when a load is connected?

I understand that it might do that if the current demand is high, but the current demand is not high in terms of the max sustainable current for the power supply.

Thanks!

Good question. It could be your power supply isn’t what it is cracked up to be. It could be a weak rectifier inside the power supply. It could be several things, but I suspect it all arises from that 5 volt power supply. Perhaps you could test it with a different load to make sure it isn’t your circuit which might have a part drawing the voltage down. A string of LEDs, a big rheostat or anything that would present a load should work.

Probably poor load regulation using the wall wart.