Pi 2 B, Jessie, 7" screen, restarts on screen sleep

voltage/current drop along the power line. I find connecting a large cap across the gpio 5v pins solves a lot of power problems. See:

--
nev 
getting the wrong stick end since 1953

Interesting - how does that help?

Daniele

It's a poor solution, but it might work. In addition to the DC voltage drop if the on board capacitance is not enough to supply the current surges, it will create AC voltage drops which cause the board to reset. So adding bulk capacitance at the connector can help, but this is not a great solution.

--

Rick

I don't understand how it could solve a problem that only occurs at the moment the display is turned *off*.

Daniele

Hmmm... how much do you understand about inductance? Honestly, I don't know what your problem is. You might want to hang a scope on your power rails and see what it looks like when you turn off power. Who knows, maybe it's a software problem.

--

Rick

The impression I get is that your power supply (or more likely the connecting lead if its a cheap USB one) was already borderline. The upgraded OS may be making more demands, or demands in a more 'uneven' manner.

When the screen is put to sleep, it likely goes through some configuration code, and at the same time the Pi will be sending information round its services "Hey guys the screens going away" - all of which will briefly draw more power, and sometimes cause a brownout.

If you have a meter stick it across the GPIO power pins and watch the voltage while going from idling to making the Pi do some real work.

--
W J G

As Rickman said, use a (fast enough) oscilloscope. A meter or even a logic probe[*] is unlikely to react fast enough to show anything.

• yes, I'm old enough to have used them. They were fine when clock speeds were in the 1-2 MHz range but totally out of their league with GHz clocks.

--
martin@   | Martin Gregorie 
gregorie. | Essex, UK 
org       |

As some one else mentioned, it is a quick and dirty fix to overcome line voltage drops (in the order of microseconds) caused by the inductance of the power line from the PSU to the Pi generating negative spikes during short term heavy demand by the PI and/or its peripherals.

The one in the image is the Pi2 I use for CAT control of my HF radio transmitter. Without the extra capacitance the under volt "spikes" produced by the sudden extra load at various times cause the low voltage square to appear on the screen and sometimes cause the Pi to reset. The cap stopped all of that.

It also prevented problems with a Pi B+ used as a media centre that would get upset when the TV dongle and TVheadend were being powered up as they were brought into use. It also makes the Pi[1] I use as a web server much more stable.

A good oscilloscope shows the voltage drop spikes quite clearly when connected to test points within the PI. A much better PSU and better cables to connect it to the Pi would be a more elegant solution but that

nothing as they were bits in the spare parts box.

For me it works. So I'm happy and job done. Others may not be a lucky.

[1] that is my original Pi that has been running my, and a few other, websites 24/7 since about June 2012.

--
nev 
getting the wrong stick end since 1953

I had one. I'd build another but I don't know anyone who smokes expensive cigars.

--

Graham. 

%Profound_observation%