Oke, forgot I connected the RPI 7"touch again. This morning disconnected the powerpin on 5V and system booted and stayed online:

#raspberrypi:~/PiModules $ sudo python pico_status1.4_hv4.0.py

  **********************************************
               UPS PIco HV4.0 Status
        Version 1.4   Relase Date 13.03.2024
       Compatible with Firmware 0x0168 and up
  *******************************************

  System Clock read from Date................: 2024-11-22 07:49:48
  System Clock read from HWCLOCK.............: 2024-11-22 07:49:47.974122+01:00
  UPS PIco RTC Registers direct read.........: 2024-11-22 06:49:49

  UPS PIco Firmware..........................: 0179
  UPS PIco Bootloader........................: 41
  UPS PIco PCB Version.......................: 54
  UPS PIco PCB Defaut Battery................: Battery LiPO 'L' 0x4C
  UPS PIco Set Battery.......................: Battery Li-Ion 'I'       0x49

  Raspberry Pi Core Temprature...............: 46 C
  Raspberry Pi Core Temprature read by PICo..: 46 C
  Embedded Fan Temp. Threshold...............: 45 C
  Embedded Fan Speed.........................: OFF % of Total RPM

  Powering Mode..............................: Cable POWERED
  Power Backup Set Mode......................: Battery Backup

  Battery Level..............................: 4.1 V
  Super Capacitor Level......................: 3.7 V
  EPR/PPoE Powering Level....................: 0.0 V
  RPi(USB) Powering Level....................: 4.9 V

  Running Time on Battery....................: 1 min
  Battery Charger Status.....................: ON
  Max Allowed Battery charging curent........: 0.44  A
  Max Allowed On Board SCAP charging curent..: 0.0  A

  Power Source(s) Cable checking timeout.....: 0.3 Seconds
  Restart on Hold time.......................: 00 Seconds
  FSSD Durtion...............................: 30 Seconds
  Firmware loop..............................: Healthy - Running Properly
  Still Alive Timer..........................: Timer Disabled
  Raspberry Pi OFF waiting time..............: 5 Minutes
  Magic Switch Funtionality..................: de-Activated

  UPS PIco HV4.0 RS232(s) Setup..............: RPI_RS232_OFF     PICO_RS232_OFF
  -12/+12 V RS232 Driver Connection..........: none
  RS232 PIco routing.........................: not implemented yet
  RS232(s) PIco sleep/wakeup data setup......: not implemented yet
  RS232(s) PIco sleep/wakeup data received...: not implemented yet

  Internal Relay State.......................: OFF
  User LED Green Activity....................: OFF
  User LED Blue Activity.....................: OFF
  Green LED mapping setup....................: none
  Blue LED mapping setup.....................: none

  A/D Converters Correction Factor...........: 00
  A/D Converter 0 ...........................: 0.0
  A/D Converter 1 ...........................: 0.0
  Opto Coupler/Digital Input Level...........: 0.0

  *******************************************
   Powered by PiModules with help from ChatGPT 4.0
  **********************************************

#raspberrypi:~/PiModules $ uptime
 07:50:30 up 38 min,  1 user,  load average: 0.11, 0.09, 0.09

Why does it say:

Running Time on Battery....................: 1 min

Next what I did is reboot the RPI, it remains working.

Next step is shutdown and this also works.

Last test is to build back the power supply to the PI screen, which means we have 2 devices on I2C bus and also current/power drawn from GPIO1 5V.

Conclusion with 7"PI screen connected -> device keeps rebooting and is not stable. 7"removed from power pin.

After removal of the screen the device is stable again.

Only thing I can now test is to power the screen with a separate power supply.

Result with external PS for the screen is that the unit stays stable so far.

Now reading on the site op RPI:

RPI Screen Documentation

They say to use GPIO4 for powering and not GPIO2 (which I used, so I will try).

Ok, I think we are getting a step closer:

[    8.798312] hwmon hwmon1: Undervoltage detected!
[    9.120425] Console: switching to colour frame buffer device 100x30
[    9.138061] vc4-drm gpu: [drm] fb0: vc4drmfb frame buffer device
[    9.404845] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[    9.404868] Bluetooth: BNEP filters: protocol multicast
[    9.404889] Bluetooth: BNEP socket layer initialized
[    9.413495] Bluetooth: MGMT ver 1.22
[    9.431906] NET: Registered PF_ALG protocol family
[    9.959925] bcmgenet fd580000.ethernet: configuring instance for external RGMII (RX delay)
[    9.961073] bcmgenet fd580000.ethernet eth0: Link is Down
[    9.993522] brcmfmac: brcmf_cfg80211_set_power_mgmt: power save enabled
[   12.830188] hwmon hwmon1: Voltage normalised
[   14.046845] bcmgenet fd580000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off
[   14.846477] hwmon hwmon1: Undervoltage detected!
[   22.914210] hwmon hwmon1: Voltage normalised
[   37.570561] systemd[1032]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set
[   46.970243] hwmon hwmon1: Undervoltage detected!
[   52.980742] hwmon hwmon1: Voltage normalised
[   57.000224] hwmon hwmon1: Undervoltage detected!
[   63.038446] hwmon hwmon1: Voltage normalised

As you can see, whenever the screen is connected it detects a Under-voltage. But ONLY when the UPS hat is on, as I used this device before without hat and never had problems.

The power supply is a 5V 2A USB at this moment, but the idea is to use passive POE (which will be similar in capacity). Therefore it is quite strange that it complains about under-voltage.

Is this something you can elaborate on?

Furthermore regarding LEDS:

1. SYS led shows heartbeat (on-of intermittent)
2. TMP LED orange, sometimes on/off, but mainly ON

What is your question here:

1. do you have wit your set our small high current LiPO battery ?

The battery I use is LiIon:

KeepPower 18650 - 3500mAh - LiIon

A photo of LEDS still needed?

Thanks for all help.

Hi,

The Running time on battery........................ 1min means that when system is out of cable power will run with battery back up for  60 seconds and then shutdown. You can  change this to 2 min, 10 min or less, or infinite (shutdown the system when battery is low)

The display is drawing so much current (max is 3.5 A@5V) but 3A@5V is better with margin. Alternative is that your batteries are weak and can not supply so much current to boost converter.

I check and answer to you at the evening

BR Pi Master

The screen only does:

Power requirements: 200mA at 5V typical, at maximum brightness.

Which is quite low (and reason I picked the RPI one).

RPI normally does way less than 2A, when connected to 24V POE you will only have 10W, which is like 2A @5V as well.

Batteries can do what is needed and are brand new (bought for this purpose).

Have been tested in my previous post in a SkyRC MC3000, so I am sure these are good as well.

I can test later as well with a different charger. Might see if that goes better.

If I have new I will also post here

Well, as I had a little time at my hands. I switched the power supply.

Initially I used a Samsung 5v @2A, which did not work (suspected by you)

This Samsung is a quite new adapter, but also capable in adaptive fast charging.

So switched to a somfy (dedicated 5v@2A) and no more reboots (with screen enabled)...

I keep it running for now to see if it sustains.

Later I will buy a Active Poe to Passive Poe converter so I can do the last and final test.

After all it seems like it was a "faulty/not properly suitable" charger for a RPI...

Keep you posted, will see how to adjust the rc.local script so it is embedded in the systemd environment.

Hi,

Better to use 3A@5V PSU, or even 3.5A@5V. Generally if more current is rated PSU then better.

You do not need to use converter of the PoE to PPoE, you loose money, the core idea of the PPoE implemented in the UPS PIco HV4.0 is to save money, you can use just an injector with cost of 1-2 EURO, like this one, and PSU of 12 or 24VDC

PPoE 1

PPoE_2

PPoE_3

UPDATE: I checked links to be sure are 1GB, and updated them

To rest of your posts will answer at the evening.

BR Pi Master

I understand what you are saying, however I have "kiosk" holes in my wall prepared which only have CAT6A connections. These CAT6A are powered by HP/Aruba PPOE switches. At the moment I prepared these screens (years back) I never checked that RPI can only do passive POE.

Of course it is a bummer, however not having the ability to add additional cabling. So 2 options remain:

1. Add a Passive POE injector in my "server room", patching from the Aruba to this injector and then to the kiosk.
2. Add a converter in the kiosk to convert Active POE to Passive POE. Currently ordered a sample from Ali: Converter

Very curious how it will work as this will be the last piece in this puzzle to start some real-life testing (before scaling up to a dozen units in the house).

Did you check an alternative for the rc.local yet or nothing specific.

Additionally, could this info help in adjusting the RTC procedure:

RTC Link

Hi,

I understand what you are saying, however I have "kiosk" holes in my wall prepared which only have CAT6A connections. These CAT6A are powered by HP/Aruba PPOE switches. At the moment I prepared these screens (years back) I never checked that RPI can only do passive POE.

However the core idea behind of this was to simplify and use low cost PPoE (also a PCB space was too small for Active PoE)

Of course it is a bummer, however not having the ability to add additional cabling. So 2 options remain:

1. Add a Passive POE injector in my "server room", patching from the Aruba to this injector and then to the kiosk.
2. Add a converter in the kiosk to convert Active POE to Passive POE. Currently ordered a sample from Ali: Converter

Must say, that I learned form you about such converter. Thank you good to know. You must decide your self what is better, and lower cost.

Very curious how it will work as this will be the last piece in this puzzle to start some real-life testing (before scaling up to a dozen units in the house).

Me too, as I said, I personally even not know that such device existing. Please keep us informed !

Did you check an alternative for the rc.local yet or nothing specific.

As I said to another post, I will be able to work on UPS PIco HV4.0 end of this week and weekend, as we are now overload with firmware for  the M.2-UPS, and productions matter -  we are small company. However I propose you a solution.

Best Regards

Pi Master

@sebas2024 

Additionally, could this info help in adjusting the RTC procedure:

What do you mean ? please advise what do you need in more details

BR Pi Master