Community discussions

MikroTik App
 
esorin0
newbie
Topic Author
Posts: 34
Joined: Tue Sep 27, 2011 8:40 am

netPower 16P power design flaw

Wed Aug 26, 2020 7:42 pm

Hello,
We have ordered 2x netPower 16P and they both put out current on the frame ground screw. It's the same voltage that's going in via the DC ports.
I am assuming that we did not receive 2 defective units and that this is the design. That being said, it's a bad design, since it pops our fuses all over the place and back feeding current into other devices via ground. There is the option to not use the ground, but that's not sane, if you know what you are doing.
 
User avatar
Lifz
Frequent Visitor
Frequent Visitor
Posts: 54
Joined: Tue Feb 26, 2013 1:05 pm

Re: netPower 16P power design flaw

Thu Aug 27, 2020 8:25 am

Did you read product specification? There is no information that this device can be powered by providing current through ground.
 
esorin0
newbie
Topic Author
Posts: 34
Joined: Tue Sep 27, 2011 8:40 am

Re: netPower 16P power design flaw

Thu Aug 27, 2020 8:32 am

Yeah, that's my bad here, device cannot be used in a -48VDC because it is not designed to be. It works fine without the ground, but that's not the right way to do it. Mixing positive and negative ground is a bad idea, as I discovered the hard way.
@Lifz, thank you for your quick and full of kindness reply. You sound as passive aggressive as Mikrotik support.
 
User avatar
BartoszP
Forum Guru
Forum Guru
Posts: 1896
Joined: Mon Jun 16, 2014 1:13 pm
Location: Poland

Re: netPower 16P power design flaw

Thu Aug 27, 2020 10:22 am

... both put out current on the frame ground screw. It's the same voltage that's going in via the DC ports....
Kind question: what is the voltage of the mentioned current"?
Real admins use real keyboards.
To quote or not to quote, there is the topic: viewtopic.php?f=2&t=168474
 
esorin0
newbie
Topic Author
Posts: 34
Joined: Tue Sep 27, 2011 8:40 am

Re: netPower 16P power design flaw

Thu Aug 27, 2020 12:26 pm

Same voltage that is applied on the DC input. In my case, 54V of 24V, depending on input.

Again, I understand the product is not usable in a -48VDC environment and that's my fault assuming it will work. I should note here that the similar product from Ubiquiti states very clear in their documentation that the device is not supporting -48VDC.
At the same time, I do not know (obviously, I am not a specialist) if it's normal to see this behavior. Maybe someone with power knowledge can say yes or no.
 
User avatar
doneware
Trainer
Trainer
Posts: 597
Joined: Mon Oct 08, 2012 8:39 pm
Location: Hungary

Re: netPower 16P power design flaw

Thu Aug 27, 2020 8:05 pm

#TR0359
 
User avatar
doneware
Trainer
Trainer
Posts: 597
Joined: Mon Oct 08, 2012 8:39 pm
Location: Hungary

Re: netPower 16P power design flaw

Thu Aug 27, 2020 8:13 pm

At the same time, I do not know (obviously, I am not a specialist) if it's normal to see this behavior. Maybe someone with power knowledge can say yes or no.
just telling my personal experiences.

we use -48VDC (36-72V) DC-DC converters with Mikrotik devices in telco environment. in this case, the input side is isolated, otherwise you end up seeing sparks if you try just try to use a 'basic' step-down solution. you can check it with a simple continuity-tester, most probably the ground on the DC-in jack is directly connected to the ground plane _and_ the grounding screws too.

stuff from mean well is pretty affordable and extremely reliable. our CRS/CCR series devices were converted to telecom power ~3-4 years ago, and they had 0 power issues. in your case however this would mean to install another IP67 enclosure with the DC-DC converter inside - not impossible but can be challenging.
#TR0359
 
User avatar
TomjNorthIdaho
Forum Guru
Forum Guru
Posts: 1119
Joined: Mon Oct 04, 2010 11:25 pm
Location: North Idaho
Contact:

Re: netPower 16P power design flaw

Tue Jan 05, 2021 4:03 am

Yea ... I just discovered the same problems with the netPower 16P

Telco 48 volt systems are positive ground.
The netPower 16P is negative ground

Soo, I can only guess that connecting a netPower 16P to a telco neg-48 volt system then connecting anything to ground potential is a really bad thing ( smoke and fuses ).

And to top it off , there is no built-in 24-Volt regulator to power 24-Volt POE devices.

as is , it's just crazy ( both thumbs down !!! )
Last edited by TomjNorthIdaho on Tue Jan 05, 2021 5:57 pm, edited 1 time in total.
 
marekm
Member Candidate
Member Candidate
Posts: 248
Joined: Tue Feb 01, 2011 11:27 pm

Re: netPower 16P power design flaw

Tue Jan 05, 2021 2:11 pm

It would be nice to have a single isolated DC/DC converter with 36-72 V DC input (telecom power) and both 24 and 48 V DC outputs (powering the netPower 16P). The voltages don't need to be very precise, so just one may be regulated and the other one taken from a different winding on the same transformer. I have a power supply with such output voltages, but only with 230 V AC input (powered from UPS, which means more power losses, another conversion from DC to AC and back to DC).
 
pe1chl
Forum Guru
Forum Guru
Posts: 7185
Joined: Mon Jun 08, 2015 12:09 pm

Re: netPower 16P power design flaw

Tue Jan 05, 2021 4:51 pm

These devices are designed to be powered by wallwarts, and to provide 24 or 48 V power via PoE to devices like access points.
They are not designed to be powered from a grounded system like telecom power (-48V referenced to ground).
But I would not power them from a +48V system that is referenced to ground, either!
Use an isolated power supply, either a wallwart or an isolated DC-DC converter.
At most, use a 48 to 48 volt isolated DC-DC converter and a non-isolated 48 to 24 V buck regulator with common minus to provide the 24V.

Indeed it would be nice if someone finds a ready-made device that does provide isolated 48 and 24V from 48V input, at approximately the right power level, and affordable cost.
 
User avatar
TomjNorthIdaho
Forum Guru
Forum Guru
Posts: 1119
Joined: Mon Oct 04, 2010 11:25 pm
Location: North Idaho
Contact:

Re: netPower 16P power design flaw

Tue Jan 05, 2021 6:10 pm

These devices are designed to be powered by wallwarts, and to provide 24 or 48 V power via PoE to devices like access points.
They are not designed to be powered from a grounded system like telecom power (-48V referenced to ground).
But I would not power them from a +48V system that is referenced to ground, either!
Use an isolated power supply, either a wallwart or an isolated DC-DC converter.
At most, use a 48 to 48 volt isolated DC-DC converter and a non-isolated 48 to 24 V buck regulator with common minus to provide the 24V.

Indeed it would be nice if someone finds a ready-made device that does provide isolated 48 and 24V from 48V input, at approximately the right power level, and affordable cost.
Re 48 Volts , on some of the Mikrotik product documents ( printed and on-line ), Mikrotik references "-48V DC telecom".
 
pe1chl
Forum Guru
Forum Guru
Posts: 7185
Joined: Mon Jun 08, 2015 12:09 pm

Re: netPower 16P power design flaw

Tue Jan 05, 2021 7:54 pm

Re 48 Volts , on some of the Mikrotik product documents ( printed and on-line ), Mikrotik references "-48V DC telecom".
Yes but that is a different class of device, that should have isolation on that input (DC-DC converter).
 
User avatar
TomjNorthIdaho
Forum Guru
Forum Guru
Posts: 1119
Joined: Mon Oct 04, 2010 11:25 pm
Location: North Idaho
Contact:

Re: netPower 16P power design flaw

Tue Jan 05, 2021 8:48 pm

It would be nice if Mikrotik had a complete DC solution for battery power that a WISP/ISP/Telco could use.

It is inefficient to use AC utility power UPS systems for alternate backup power. And most 1500 watt UPS systems do not have the battery capacity to last more than one hour.

I wish Mikrotik had a complete gel-cell battery backup solution that could be used on all newer Mikrotik products. Something which could support 1 hour to days on battery power when utility power fails ( and something compatible with existing telco -48 volt systems ).

Generators use fuel and getting fuel to a mountain top tower in winter can be impossible at times. Telco -48 DC battery power systems have been around for almost 100 years. The problem is Mikrotik needs to design products which can optionally utilize these power systems.
 
pe1chl
Forum Guru
Forum Guru
Posts: 7185
Joined: Mon Jun 08, 2015 12:09 pm

Re: netPower 16P power design flaw

Tue Jan 05, 2021 10:35 pm

There used to be a product like that for 12V but it was discontinued. Apparently this kind of thing does not sell.
 
User avatar
TomjNorthIdaho
Forum Guru
Forum Guru
Posts: 1119
Joined: Mon Oct 04, 2010 11:25 pm
Location: North Idaho
Contact:

Re: netPower 16P power design flaw

Tue Jan 05, 2021 10:51 pm

There used to be a product like that for 12V but it was discontinued. Apparently this kind of thing does not sell.
re ... There used to be a product like that for 12V but it was discontinued ...

Yea the mUPS. I've got a couple dozen of them.
They could of been a great product. My issue with them is if you have 4 mUPS at a tower, then you also had 4 non-shared-common batteries.

If I had a Mikrotik product wish-list re the subject of Mikrotik DC powered POE switches/routers , it would likely start with something like the Netonix WS-26-400-IDC ( a negative or positive 48-Volt POE switch - with built-in 24-Volt regulator to also power 24-Volt POE devices and/or 48-Volt POE devices ). - except it would run ROS - and SFP+ ports - and an optional mid-span tower mounted nema enclosure.

Who is online

Users browsing this forum: Dzejkob and 38 guests