4

Some people claim that they have fixed the Pi's USB hotplugging so that power spikes don't hard-reboot the Pi. They have extended the (orange) capacitor next to the USB hub by another one for that.

If this works with a 10V 220μF one (w/o breaking anything), could I use 1000 (or even >= 10000) μF instead? (Such capacitors are cheap and I have that money.) I'd like not to break my Pi at all.

Al Klimov
  • 141
  • 1
  • 1
    I didn't think there were any issues with the A+/B+/Pi2 with regard to hotplug. What model Pi do you have? – joan Oct 24 '15 at 19:31
  • @joan B (not B+) – Al Klimov Oct 24 '15 at 21:53
  • If you do it really really really slowly, close your eyes, cross your fingers, and think happy thoughts, it might be okay. – goldilocks Oct 25 '15 at 01:26
  • @joan Was it because of linear regulators in model A and B ? Just curious to know the reason behind that flaw. – dhruvvyas90 Oct 25 '15 at 05:39
  • 1
    @dastaan You'd have to check but from memory the USB peripherals caused a voltage dip during insertion which caused the reset. I don't think it was the fault of the linear regulators, but the power chain as a whole. I know the A+/B+/Pi2 are meant to be much improved (and more efficient given that linear regulators are no longer used). You'd really need an answer from one of the electronics type people. Adding capacitors help prevent the voltage dip. – joan Oct 25 '15 at 09:02
  • 3
    easiest solution is to put a powered USB hub at the point where the user hot plugs acting as a buffer to the Pi. This is what I do with my old Bs that otherwise reboot when WiFi is plugged in. – rob Oct 27 '15 at 14:45

1 Answers1

1

Yes, you can add a capacitor with a larger value - however, this will not help:

  • Changing the capacitor value will not enhance the circuit's ability to dampen the power spike; look here http://planetcalc.com/1980/
  • There is a polyfuse that will cut off the circuit at 140 mA, meaning you cannot support devices that draw >140 mA anyway

Other possible solutions:

  1. Get a powered USB hub
  2. Directly connect a more stable power supply to the 5V pin on the USB ports (not recommended)

I DO NOT TAKE ANY RESPONSIBILITY FOR ANY DAMAGE CAUSED

However, if the capacitor draws too much current and causes one of the polyfuses to "blow", it should return to working condition after about a day.

sunny-lan
  • 111
  • 3
  • No, you can't use a simple resistor there. Either it does nothing, because there is another path from the supply input to the supply output you are trying to stabilize. Or else the resistor is between the cap and the supply you are trying to stabilize, in which case the cap does nothing. Or else the resistor is in the only line from the input to the supply you are trying to stabilize, in which case it drastically limits the net current which can be supplied. – Chris Stratton Nov 16 '15 at 01:44
  • Yes, sorry. I have made some edits – sunny-lan Nov 16 '15 at 01:59
  • You are incorrect to say that changing the capacitor's value will not help with the damping - effectively when you plug in the peripheral, you have your onboard capacitor trying to charge its onboard capacitor, so their proportion matters. However, the potential problem is when you first plug in the pi with a bigger capacitor - that's the problem you were trying to solve with a resistor, but it is not one that can be solved with a simple resistor. – Chris Stratton Nov 16 '15 at 02:21