Author Topic: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation  (Read 23853 times)

0 Members and 4 Guests are viewing this topic.

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8138
  • Country: ca
    • LinkedIn
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #25 on: September 15, 2018, 07:05:26 am »
without even looking at anything else : who the hell suspends a transformer in a hole like that ?
This super high quality, 10k$, 1kv isolated 4 channel 14 bit scope also has the transformers mounted in the middle of the PCB, and apparently, the high isolation development took over a year and was found to be the best solution:
https://www.eevblog.com/forum/blog/eevblog-1119-designing-a-1kv-isolated-oscilloscope/
 

Online Halcyon

  • Global Moderator
  • *****
  • Posts: 5981
  • Country: au
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #26 on: September 15, 2018, 11:54:07 pm »
I don't understand why there was so much hype for POE hat.

You can spend less money, and get a properly enclosed product, that can be reused for another project down the road if needed:
https://www.aliexpress.com/store/product/Cdycam-IEEE-802-3af-Micro-USB-Active-PoE-Splitter-Power-Over-Ethernet-48V-to-5V-2/1920360_32841790680.html
https://www.aliexpress.com/store/product/Xinfi-DC-12V-POE-Adapter-Injector-Splitter-Connector-IEEE802-3af-Active-10-100Mbps-For-IP-Cameras/1543317_32620368747.html

The only advantage I see is its slightly more compact.

Because POE is a widely used standard. You can plug a non-POE device into a POE port and not have it blow up. Those passive splitters/injectors are fine for crude installations, but when you forget and you plug it into the wrong device, the outcome isn't so good.

Also, those splitters use the unused pairs on a 10BASE-T / 100BASE-TX connection. If you used one of those splitters with a Gigabit-capable device, the best you'd get is 100Mbps as it's splitting off those "spare" pairs. A POE/POE+ device can run at 1Gbps while the switch also supplies the power along the same cable.
 

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 7236
  • Country: ca
  • Non-expert
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #27 on: September 17, 2018, 06:56:40 pm »
Because POE is a widely used standard. You can plug a non-POE device into a POE port and not have it blow up. Those passive splitters/injectors are fine for crude installations, but when you forget and you plug it into the wrong device, the outcome isn't so good.

Also, those splitters use the unused pairs on a 10BASE-T / 100BASE-TX connection. If you used one of those splitters with a Gigabit-capable device, the best you'd get is 100Mbps as it's splitting off those "spare" pairs. A POE/POE+ device can run at 1Gbps while the switch also supplies the power along the same cable.

Second link has gigabit for $11, as I noted above.
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 

Online Halcyon

  • Global Moderator
  • *****
  • Posts: 5981
  • Country: au
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #28 on: September 17, 2018, 11:11:47 pm »
Because POE is a widely used standard. You can plug a non-POE device into a POE port and not have it blow up. Those passive splitters/injectors are fine for crude installations, but when you forget and you plug it into the wrong device, the outcome isn't so good.

Also, those splitters use the unused pairs on a 10BASE-T / 100BASE-TX connection. If you used one of those splitters with a Gigabit-capable device, the best you'd get is 100Mbps as it's splitting off those "spare" pairs. A POE/POE+ device can run at 1Gbps while the switch also supplies the power along the same cable.

Second link has gigabit for $11, as I noted above.

Sure. However that isn't a "dumb" splitter (at least it doesn't purport to be). That device provides phantom power, just like other 802.3af devices do, hence why it's able to be used at Gigabit link speeds. It just has an additional DC barrel jack at the end. The point of the Raspberry Pi hat is so that additional equipment isn't required.



 
The following users thanked this post: thm_w

Offline viewgsm

  • Newbie
  • Posts: 5
  • Country: us
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #29 on: September 18, 2018, 01:51:36 am »
Well, guess what, we did the test  for NavoLabs PoE hat for you and it passes with flying colors :) the video is being edited and should be on youtube tonight, you can easy power up 2.5" HDD from the usb, also the ripple is very well under control, 20-40mV Peak-to-peak,  also using a DC load we are able to sink 1A easy from the USB port while powered by the NavoLabs PoE hat
https://navolabs.com/product/raspberry-pi-3-micro-poe-hat/    :-DD :box:
 
The following users thanked this post: ricktendo

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5031
  • Country: si
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #30 on: September 18, 2018, 05:25:30 am »
Nice one, very compact design there.

Tho by the looks of it that's not a isolated supply so careful plugging anything grounded into that raspberry pi.
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3731
  • Country: lv
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #31 on: September 18, 2018, 07:46:58 am »
Tho by the looks of it that's not a isolated supply so careful plugging anything grounded into that raspberry pi.

Also looks that there are no rectifiers nor reverse polarity protection diode  :palm:
Claiming "IEEE 802.3af POE" is so wrong.
 

Offline LapTop006

  • Supporter
  • ****
  • Posts: 467
  • Country: au
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #32 on: September 18, 2018, 11:01:00 am »
Tho by the looks of it that's not a isolated supply so careful plugging anything grounded into that raspberry pi.

Also looks that there are no rectifiers nor reverse polarity protection diode  :palm:
Claiming "IEEE 802.3af POE" is so wrong.

Some PoE chipsets such as the Si3402 have the diodes integrated, although for a use such as this I'd use external diodes, not sure what chipset they're using on that one.
 
The following users thanked this post: ogden

Offline ricktendo

  • Regular Contributor
  • *
  • Posts: 115
  • Country: hn
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #33 on: September 18, 2018, 02:42:28 pm »
Some PoE chipsets such as the Si3402 have the diodes integrated, although for a use such as this I'd use external diodes, not sure what chipset they're using on that one.
The chipset the board is using is the: Si3402 B-GM

I purchased one while I waited for the official Pi PoE Hat to be released, I added a fan connector to help with cooling.

 

Offline viewgsm

  • Newbie
  • Posts: 5
  • Country: us
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #34 on: September 18, 2018, 04:36:24 pm »
It does, take a closer look, plus it has integrated diodes too !
 

Offline viewgsm

  • Newbie
  • Posts: 5
  • Country: us
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #35 on: September 18, 2018, 04:40:34 pm »
Also looks that there are no rectifiers nor reverse polarity protection diode  :palm:
Claiming "IEEE 802.3af POE" is so wrong.

Dude, do some research before you speak, or at least look good at the photos, it has integrated diodes and it has external too, and its IEEE 802.3af compliant.
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 8070
  • Country: gb
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #36 on: September 18, 2018, 07:02:41 pm »
Does appear to not be isolated, though. That can be a massive pain.
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #37 on: September 18, 2018, 07:07:01 pm »
without even looking at anything else : who the hell suspends a transformer in a hole like that ?
Most of the thin TV PSUs.


yup. through hole ones...
or if the transformer is on TOP of the board.
in this thing it is hanging BELOW the board causing stress on the copper lamination ! BAD design !
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline ricktendo

  • Regular Contributor
  • *
  • Posts: 115
  • Country: hn
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3731
  • Country: lv
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #39 on: September 20, 2018, 05:15:24 am »
Also looks that there are no rectifiers nor reverse polarity protection diode  :palm:
Claiming "IEEE 802.3af POE" is so wrong.

Dude, do some research before you speak, or at least look good at the photos, it has integrated diodes and it has external too

Dude, please be so kind and show (with arrows) where are rectifiers. Seems, you don't even comprehend what rectifiers I am talking about - obviously not about diode at buck converter output [rofl]

Quote
and its IEEE 802.3af compliant.

Dude, you are The One who needs to do some research - about IEEE 802.3af compliance.  :-DD

Hint: clause "33.4.1 isolation"
« Last Edit: September 20, 2018, 05:21:22 am by ogden »
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5031
  • Country: si
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #40 on: September 20, 2018, 05:49:00 am »
Found another two PoE hat designs... Opinions?

eBay auction: #263811907135

https://www.tindie.com/products/canerdurmusoglu/power-over-ethernet-hat-for-raspberry-pi3-model-b/

The first one does look better made and it does have real opto feedback so its output should be nice and stable. It has decent sized capacitors on it to smooth things out including some ferrite filters to really clean it up. The parts look reasonably well cooled and it has troughhole connectors. Looks quite good. Tho the isolation doesn't look that great(But it is there so that's a big plus) and the capacitors that bridge the primary and secundary side look kinda small to meet the kilovolt voltage ratings and the way that they route the output back to that capacitor makes a significant loop area in there and probably makes a good antenna to radiate out switching noise.

The other one i couldn't reverse engineer very far from the photos but its quite neat in how compact it is, hopefully they did it properly and met the spec.

EDIT: And yes the ebay module is the best i seen so far in this whole thread.
« Last Edit: September 20, 2018, 05:52:37 am by Berni »
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3731
  • Country: lv
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #41 on: September 20, 2018, 05:49:49 am »
Found another two PoE hat designs... Opinions?

eBay auction: #263811907135

https://www.tindie.com/products/canerdurmusoglu/power-over-ethernet-hat-for-raspberry-pi3-model-b/

Both isolated. Good. Isolation is "must have" for anything that is more than just hobby toy for living room installation.

Ebay module clearly have two rectifiers meaning - will work with any polarity. It also have optoisolated feedback, thus supposedly good regulation at any load. From what I see, eBay module is best I have seen (visually).

That IVMTECH module seems to be "OK-ish" but unfortunately no isolated feedback, nor input rectifiers are visible on top of the daughterboard PCB. There are some hints that most likely some components are on the bottom. Hard to tell. Have to see both sides.
 

Online Halcyon

  • Global Moderator
  • *****
  • Posts: 5981
  • Country: au
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #42 on: September 20, 2018, 06:03:56 am »
Found another two PoE hat designs... Opinions?

eBay auction: #263811907135

https://www.tindie.com/products/canerdurmusoglu/power-over-ethernet-hat-for-raspberry-pi3-model-b/

Both isolated. Good. Isolation is "must have" for anything that is more than just hobby toy for living room installation.

Ebay module clearly have two rectifiers meaning - will work with any polarity. It also have optoisolated feedback, thus supposedly good regulation at any load. From what I see, eBay module is best I have seen (visually).

That IVMTECH module seems to be "OK-ish" but unfortunately no isolated feedback, nor input rectifiers are visible on top of the daughterboard PCB. There are some hints that most likely some components are on the bottom. Hard to tell. Have to see both sides.

The feedback on that site doesn't look too good.
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 8070
  • Country: gb
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #43 on: September 20, 2018, 02:57:36 pm »
Also looks that there are no rectifiers nor reverse polarity protection diode  :palm:
Claiming "IEEE 802.3af POE" is so wrong.

Dude, do some research before you speak, or at least look good at the photos, it has integrated diodes and it has external too

Dude, please be so kind and show (with arrows) where are rectifiers. Seems, you don't even comprehend what rectifiers I am talking about - obviously not about diode at buck converter output [rofl]

The rectifier is internal to the IC. This is not hard to understand.
 

Offline ogden

  • Super Contributor
  • ***
  • Posts: 3731
  • Country: lv
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #44 on: September 20, 2018, 07:16:16 pm »
The rectifier is internal to the IC. This is not hard to understand.

I just missed to carefully read DS of Si3402. Two on-chip input diode bridges confirmed, case closed.
 

Offline viewgsm

  • Newbie
  • Posts: 5
  • Country: us
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #45 on: September 21, 2018, 06:05:44 am »
intro to the test video for the NavoLabs PoE hat

 

Offline Red Squirrel

  • Super Contributor
  • ***
  • Posts: 2751
  • Country: ca
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #46 on: September 21, 2018, 12:22:31 pm »
What kind of boggles me is that they would release this as is.  That's a pretty major flaw.   Didn't they test the design first?   "It passes electrical rules check, ship it!"   :-DD
 

Online oPossum

  • Super Contributor
  • ***
  • Posts: 1453
  • Country: us
  • Very dangerous - may attack at any time
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #47 on: October 03, 2018, 07:12:28 pm »
A fix using a LC filter. 4.7 uH + 100 uF x 2


 
The following users thanked this post: ricktendo

Offline chefkoch84

  • Contributor
  • Posts: 41
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #48 on: October 22, 2018, 08:10:24 pm »
FYI:

Dave is not the lonley ee nerd anymore - he goes full mainstream ROCKSTAR.

This is the german yellow press of IT:
https://www.heise.de/make/meldung/Aerger-mit-dem-Raspberry-Pi-Power-over-Ethernet-HAT-4196194.html

Gangnam Style - Style

cheers.
 

Offline darrellg

  • Supporter
  • ****
  • Posts: 106
  • Country: us
Re: EEVblog #1122 - Raspberry Pi 3 PoE Hat FAIL Investigation
« Reply #49 on: November 09, 2018, 04:54:58 pm »
The Raspberry Pi folks have posted a mea culpa discussing what went wrong and what they've done to fix it.

https://www.raspberrypi.org/blog/poe-hat-revision/
« Last Edit: November 09, 2018, 04:58:57 pm by darrellg »
 
The following users thanked this post: ogden


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf