Problem for undervoltage

[raspTest]

Hi everyone!
I'm developing a system that uses the rxtx libraries of java to communicate with an arduino, via serial with usb, and with a webcam (using opencv).
If I launch everything from the computer, the program does not have problems even after a whole day.
If I use the rasberry as a computer, I noticed that after a while 'in the system logs appears an undervoltage warning and below, the rasp disables and reactivates the USB and Ethernet, causing a malfunction of the program with the modification of the ports on the arduino and the webcam are attached.

The arduino is attached via usb to the raspberry, powered by it. It has several sensors and actuators on board.
The raspberry is powered with 5v by 2.5A

How can I solve this problem?

[Joel_Mckay]

Your power supply may be too low try a 5V 5A minimum supply if you have usb devices plugged in....
Otherwise you may need an external powered USB hub.

Often the USB driver power-save modes will periodically reset devices, and can be rather involved to disable.

Good Luck,
J

[drgeoff]

Your power supply may be too low try a 5V 5A minimum supply if you have usb devices plugged in....

The polyfuse on the micro-USB power input will not pass 5 Amps. A PSU that truly meets the specified 2.5 Amps at 5 Volts is adequate.

[raspTest]

On system's logs, I have a series of this messages:
Under-voltage detected! (0x00050005)
Voltage normalised (0x00000000)
rpi_firmware_get_throttled: 3 callbacks suppressed

[RaTTuS]

or your micro usb lead is made from damp string - get a much better [thicker] one

[Joel_Mckay]

Let us assume you already enabled max_usb_current=1 ...

This site is wrong for 3 reasons:.
https://www.raspberrypi.org/help/faqs/#powerReqs

1. The 5v header pins do not pass through the poly fuse.
2. While 1.2A for the USB port is fine, that optimistic minimum 400mA core draw should be more like 1.8A with a pi3B+ desktop running wifi. (that is a 3A draw)
3. Most cheap power supplies lie about the output, so de-rate by 40% unless you plan to fry eggs on the inductor or fets..

I have been building embedded devices since 1994, but please educate me if I am missing something.

[RaTTuS]

Let us assume you already enabled max_usb_current=1 ...
...

ignore that - that was only for older pi's nothing done in the last couple of years needs it

[markost]

Let us assume you already enabled max_usb_current=1 ...

This site is wrong for 3 reasons:.
https://www.raspberrypi.org/help/faqs/#powerReqs

1. The 5v header pins do not pass through the poly fuse.
2. While 1.2A for the USB port is fine, that optimistic minimum 400mA core draw should be more like 1.8A with a pi3B+ desktop running wifi. (that is a 3A draw)
3. Most cheap power supplies lie about the output, so de-rate by 40% unless you plan to fry eggs on the inductor or fets..

I have been building embedded devices since 1994, but please educate me if I am missing something.

pi3.PNG (39.53 KiB) Viewed 4041 times

[markost]

try to use the power supply arduino separately and not through the usb port

[Joel_Mckay]

@markost
That is the usb port input, not the 5v/VCC gpio header pin 2 and 4.
The 3B+ looks like it has kept this external shield power feature.

[LTolledo]

How can I solve this problem?

How about doing this:

1. Power the Raspberry Pi with its own 5v 2.5A PSU with thick microSUB cable.
2. Power the Arduino with its own PSU (5V 2A perhaps)
3. Connect the USB deviced thru a powered Hub..

Its still up to you to apply the solution....... or continue wondering why you still have undervoltage problem.....
anything else is irrelevant....

[markost]

@markost
That is the usb port input, not the 5v/VCC gpio header pin 2 and 4.
The 3B+ looks like it has kept this external shield power feature.

the power supply comes via the microUSB, it goes over the polyfuse. Here we have "first" 5v, which is the same as the GPIO expansion pin 2 and 4.

So, does the 5v header pins pass or not through the poly fuse?

[markost]

How can I solve this problem?

How about doing this:

1. Power the Raspberry Pi with its own 5v 2.5A PSU with thick microSUB cable.
2. Power the Arduino with its own PSU (5V 2A perhaps)
3. Connect the USB deviced thru a powered Hub..

Its still up to you to apply the solution....... or continue wondering why you still have undervoltage problem.....
anything else is irrelevant....

This is correct instead step 3.

3. Connect GND of both power supply together
and

4. remove 5V and GND from arduino USB, only TX and RX are needed

[Joel_Mckay]

So, does the 5v header pins pass or not through the poly fuse?

1. The GPIO header 5v pins do not pass through a fuse
2. micro-USB are rated for a theoretical 3A maximum, but the polyfuse will never allow this limit
3. As other users have said, a powered external USB expansion hub may be a simpler solution

Best of luck,
J

[markost]

OK,

can you show me on schematic where is the 5V input for the GPIO header (from where this voltage is coming)?
Obviously I'm looking at something wrong

Thank you!

[Joel_Mckay]

@markost
The two giant black arrows in the 3B+pow.png image show the pi's 5v rail and external VCC GPIO header pins 2 and 4 are connected with copper.

In schematic drawings people use symbols to label common things like ground and power rails to simplify the number of lines needed to describe a circuit.
The external powered USB expansion adapter option may be less complicated and safer if you are not comfortable with electronics.

Cheers,
J

[markost]

@Joel_Mckay

Thanks for your point of view.
I agree for the 5V power rail (all marked as 5V on schematic are on the same rail), but I don't agree that the 5V on the GPIO (pin 2 and pin 4) are not after polyfuse.

I don't have RPi3B+, only 3B.

If 5V on the GPIO doesn't go through the polyfuse than must go directly to the microUSB pin1.
If that is true, that mean that test point PP1&PP2 are connected directly with the GPIO2&4. But they are not connected (for RPi3B).

5V on the GPIO(pin2&4) are connected to the test point PP7.
Test point PP7 is behind the polyfuse.
So, for me, 5V on GPIO is after polyfuse.

Regards

[drgeoff]

The 5 volt pin of the micro-USB socket and the 5 volt pins of the GPIO header are on opposite ends of the polyfuse.

[raspTest]

thanks a lot! i will ltry !!!

[raspTest]

it seems to be solved! I connected the arduino's power supply of the externally, and then I connected arduino and raspberry through the usb for communication. for now the rasp has not launched any warnings on the voltage!

Thanks!!!