Helpdesk

Hello,

I'm having issues with getting the Pienergy min work on my raspberry pi 3b+.

I measure 5v on the output from the energy, and when I press the on button the red power led on the Pi lights up, after that there is no boot. 
If I hold the power on botton for 1 sec the green led on the pi starts flashing and the Pi start to boot until the PiEnergy shutdown (this is expected behavior). Please help.

Hello,

Usually pressing the power button for a second will start the voltage converter on the board, and 5V will be provided to the Raspberry Pi. 

The moment the Raspberry Pi gets this 5 V, it should start up. 

When pressing the button again for a second, the voltage converter shuts down after about 10 seconds. There is also an autostart firmware, where the voltage converter will be on default.

When 5V is provided to your Raspberry Pi and it does not boot, you can try connecting GPIO3 to GND for a short time; this should start your Raspberry Pi.

Other possible sources of trouble could be the power supply is not providing enough current, or there is a hardware defect on your Raspberry Pi or Pi Energy.
In this case you could try booting your Raspberry Pi with a different 5 V power supply over the 5V and GND GPIO pins to test it.

Best regards 

Tim (Joy-IT)

Hello,

Okay thanks for the help, I will try once I get home from work.
what does connecting the GPIO3 to gnd do?

As I said I see the solid red pwr led on the Pi board as I turn on the PiEnergy mini but no act green led flashing, which I assume means its not gotten to read from sd card stage in the booting stage yet.
Is there any gate before that occurs? meaning any current limits etc?
My PSU that I connected to the pienergy should be a +12v 3amp one, but I will also try with thicker gage wire and see if I can find another PSU.

Connecting GPIO3 to GND usually triggers a wake up and will let the Raspberry Pi Boot if a power supply is connected.

There should not be another gate.
A 12 V 3A PSU is definitely sufficient for the Raspberry Pi.

Best regards 

Tim (Joy-IT)

Thanks Tim,

And normally the PiEnergy connects GPIO 3 to gnd?

No, the PiEnergy-Mini does not connect GPIO 3 to GND.

The PiEnergy-Min simply supplies 5 V to the Raspberry Pi, which causes it to start automatically as soon as power is applied.

GPIO 3 is only connected manually to GND if you want to restart the Raspberry Pi after it has been shut down manually. Briefly shorting GPIO 3 to GND triggers the wake-up function, so you don’t need to disconnect and reconnect the power supply to start the system again.

Best regards 
Tim (Joy-IT)

Sadly I tried shorting the GPIO3 to gnd when the red pwr led was on from the PiEnergy but no boot.
I didnt find any new PSU but hopefully I will try that tomorrow

Hello again,

I tried today again with a new PSU 24v @1.75Amp which should be plenty of power. Exactly the same behaviour, red pwr led but no boot, I also measured the +5v fan output on the PiEnergy and it was within spec.
What should I try next?

//Martin

Question: is the PiEnergy mini supposed to generate the 3.3v as well? It seems that when I have the PiEnergy plugged in and powered I read nothing in the 3.3v pin but I do if I plug the pi into the usb pwr.

and I can also confirm that during the power-down phase of the PiEnergy (once you've hold the button for >1 sec) the 3.3v rails comes alive and hence the Pi boots.

okay I've done even more investigation into this:

When pressing the start button on the PiEnergy mini:

+5 pin (2 & 4) raises to ~5-15v

+3.3 goes to ~0.74v (GPIO 2 & 3 as well)

GPIO4 goes to 2.14v

compared to when I run from the usb:

+5v applied

GPIO4 goes high to ~3v

3.3v is applied on pin1

Sorry it was a typo, 5.15v

Right okay the 3.3v is generated by the Pi PMIC, but why is the GPIO4 left floating from the PiEnergy? When I bench test the PiEnergy I can clearly see the +5v and GPIO4 goes to high 3.3v.
Something seems to be leaving the GPIO4 in "limbo mode" sort of like the Pi is trying to make it low and the PiEnergy's pull-up draws it high?
Not sure why that would hault the boot process, but as you saw as well from the measurements with usb cable, the GPIO4 goes high directly at boot.

GPIO4 is connected to the microcontroller on the PiEnergy board.

During normal operation, the microcontroller keeps this pin pulled to 3.3 V. When the power button on the board is pressed, the microcontroller pulls the pin low to signal the Raspberry Pi to shut down. About 10 seconds later, the main 5 V power converter is switched off.

The “limbo mode” you described could occur if the Raspberry Pi is also trying to pull this pin low.

On the PiEnergy side, this could indicate a problem either with the microcontroller itself or with the 3.3 V voltage converter that powers the microcontroller.

To check whether the converter is working properly, please measure the voltage on the back of the board at the small round test pads labeled “+” and “G”.
The voltage there should be approximately 3.3 V.

You could also measure the voltage at GPIO4 on the PiEnergy while it is not plugged into the Raspberry Pi. This can help determine whether the Raspberry Pi is pulling the pin low.

Hello,

When I measure GPIO4 on the PiEnergy without it connected to the raspberry pi I measure 3.3v

Is there any default or setting on the PI that can set the status of GPIO during boot?

I don't think there is a default setting where the Pi sets the gpio.
There are some settings like 1Wire for example that could use this pin, but this should also not affect the booting process.

Based on the described behaviour, I would assume that the problem lies with the Raspberry Pi rather than the PiEnergy.
I assume you don't have another Raspberry Pi to check this, right?

Sadly no, I do not have another Pi

according to the datasheet from the PI GPIO4 should have a pull-up as a default state.

Yes, the pin is pulled high at start-up, but this has no effect on the PiEnergy. And conversely, the status of the pin should not affect the boot of the Raspberry Pi.

There seems to be a way to overwrite the default behavior of a GPIO pin:

import RPi.GPIO as GPIO



GPIO.setmode(GPIO.BCM)



gpio_number = 8



GPIO.setup(gpio_number, GPIO.IN, pull_up_down=GPIO.PUD_UP)



print("GPIO no " + str(gpio_number) + ": " + str(GPIO.input(gpio_number)))

I'll see if that helps

So there is no mechanism in the Pi that can halt it from booting due to floating GPIO pins?

I don't know of any mechanism that would prevent the Pi from booting due to floating GPIO pins, including GPIO4.

The only somewhat related case would be a manually implemented shutdown routine that monitors GPIO states and powers the Pi down depending on a GPIO condition. However, even in that case the Pi would first have to boot before such software could run.

I agree that would be most likely

I have now measured the pienergy on bench and between + and G there is 3.3v.

The gpio4 is 3.3v high during booting from USB and 2.45v with the pienergy

I think I have solved it now and it seems to be a too slow ramp up of the 5v for the PMIC to start. I measure the raise time with the pienergy and from usb cable and the pienergy was slower. Also I tried to start the pienergy on bench and press it onto the pi header for an instant 5v and it booted

Hi Martin,

the ramp up time usually takes about 5 ms. I have never observed that this is causing a problem with the boot process of the Raspberry Pi.

Do you have the opportunity to measure the ramp-up time with an oscilloscope to see if your device is taking an unusually long time?

Sadly not, I will see about getting my hands on a oscilloscope.
But this is the only reason I can think of that the PI is not booting?

It does seem possible that this is the cause on your device, especially since the Raspberry Pi boots normally when the PiEnergy is already powered and then connected to the Pi.
However, without additional testing it is difficult to determine with certainty what exactly is causing the issue.

On Monday I can most likely get ahold of an oscilloscope.

Otherwise maybe it is possible to add a decoupling capacitor to increase the ramp up time?

Unfortunately, an additional capacitor would increase the ramp-up time rather than decreasing it.

would the ramp up time different with another firmware? I know there are 2 versions one where the button is in use and one where the voltage regulator turn on as soon as it sees power

The other firmware would not make a difference. It would only skip the step where you press the button to start the device, but the actual time it takes for the output voltage to ramp up would remain the same.

then the next step is only to measure the ramp up time?