MikroTik

I can't power my hAP AX2 with POE IN/OUT.

If I plug a POE device into the POE IN/Out it doesn't work - but if I plug that POE device into another POE source first and then plug it into the hAP AX2 POE out after - it works.

Seems very picky/touchy.

"PoE in Passive PoE"
https://mikrotik.com/product/hap_ax2#:~ ... in%20input

That's why.

can u elaborate a bit on this please?
as i have a hapax2 as well and couldnt get it to work as a ap i have to use an injector from my hapax3

i understand active poe all good but passive im not understanding i watch some docs and videos
so it can only support what it is so
24 to 24 and so on
but i plugged my ethernet cable in the poe port and it didnt power up any ap ?? states wan 1 is in/out
in config it is turned on for poe?

how is it spose to work in the hapax2??
is it only a one way out or only one wayt in?
and it is on the wan port where my internet comes in?? was and still a little confused
cheers

Coughy, which EXACT "POE device" are you trying to power? Make/model, please.

Mikrotik devices such as the hapAX2 provide a (actually a bit limited in current) out voltage that is at the same level as the voltage that powers the device.
So if you are powering the Ax2 with the standard 1.2A@24V mains adapter, you have on the PoE out port 0.6A@24V max, i.e. less than 15W.

And you have it on Mode B scheme, i.e. pins 4,5+ pins 7,8-.

A brief list of things that may go wrong, your PoE device may well:
1) need a higher voltage (like 48-57V)
2) need more power than the 14.4 W the Ax can output
3) use Mode A scheme only
4) actually force negotiation of active PoE

i was trying to get the hapax2 to poe power to a capax but couldnt get it to work i have used a injector and changed my set up
was just trying to understand what was happening
so the hapax2 is poe in and out correct?
the capax is poe in but wouldnt turn on not sure if i had it correct just trying to understand poe
cheers pete

PoE is a mess in itself, and Mikrotik doesn't help with the confusing documentation they provide.

The hap Ax2 is documented as having PoE out:
PoE-out ports Ether 1
PoE out Passive PoE
Max out per port output (input 18-30 V) 600 mA
Max total out (A) 0.6 A
Total output current 0.6
Total output power 16.8W

What is written elsewhere is that the powering scheme is Mode B.

The Cap AX is documented as having PoE in:
PoE in 802.3af/at
PoE in input Voltage 18-57 V
which in itself is "wrong", no way that anything lower than 44V is 802.3a compatible.
Poe 802.3af mandates voltage 44-57V, 802.3at mandates voltage 50-57V, AND compatibility with BOTH Mode A and Mode B powering scheme.

With the standard power adapter of the hap Ax2 at the most you have is 24V (which is NOT 802.3af/at compatible).

Whether the Cap Ax can actually be powered with passive PoE at 24V is not written anywhere, it may work or it may not.

But there are reports hinting that it could work with the correct connection and PoE set to "forced on".

For some strange reasons, both the hap ax2 and the hap ax3 have "everything" on ether1 (both PoE-in and PoE-out on the same port that is normally "internet" or WAN), see:
viewtopic.php?t=196611

The hap Ax2 can instead definitely be powered with PoE coming from a hap Ax3 as both use passive PoE (in/out) at 18-30V, the issue might be that the standard 24V/1.2A power supply of hap Ax3 simply has not enough "juice" to power both devices and you need a beefier power supply.

Whether the Cap Ax can actually be powered with passive PoE at 24V is not written anywhere, it may work or it may not.

You're right, the passive PoE-in is rarely mentioned for MT devices with 802.3 af/at PoE-in support. But it was explained multiple times (by MT staff) that all MT's PoE-in capable devices support passive PoE-in. Implicit data affirms this in multiple places:

• PoE-in voltage range, as you wrote, is wider than 802.3 PoE standards specification, which by itself hints at passive PoE capability
• text in product description
  e.g. description of cAP ax has this: "you can use cAP ax to power other devices ... or even another cAP ax!". The PoE-out explicitly specifies "Passive PoE up to 57V" ... combined with statement from product description it means that PoE-in should be passive-capable as well.
• officially endorsed related products, such as RB-GPOE power injector (actually included with cAP ax) which is passive PoE injector

And, probably to be added to the list (but it's not MT specific): only rare PoE-powered 802.3 af/at/... compliant devices refuse to be powered if PSE (power provider) doesn't perform full power handshake ... most of them simply accept power (if voltage is within specified range). And this behaviour is what IMO defines "passive PoE" ... powering without handshake.

Yes, but the dubitative form is anyway appropriate when the standard power supply is used, the 24V/1.2A are IMHO very "tight" to power both the "main" device and another one via (passive) PoE.

As said PoE is already a mess in itself, but the good Mikrotik guys, with their lacking (or misleading, or downright wrong) documentation only make things worse.

I don't think that it is too much nitpicking to state that this:

PoE in 802.3af/at
PoE in input Voltage 18-57 V

is incorrect. and that should have been wirtten as:

PoE in passive or 802.3af/at
PoE in input Voltage 18-57 V passive or 44-57V 802.3af/at

Yes, but the dubitative form is anyway appropriate when the standard power supply is used, the 24V/1.2A are IMHO very "tight" to power both the "main" device and another one via (passive) PoE.

Well ... one always has to do power budget calculations when doing any kind of PoE ... I'm pretty sure that even major vendors ship PoE switches where total power available for PoE is less than sum of individual port capabilities. And they count on the fact that not every PoE device consumes maximum allowed by standard and that devices even don't consume max of their rated power all the time and that maxima of powered devices hopefully happen at different moments ... that's called "over-subscription" and statistics somehow helps with it.
Indeed in case of single PoE out port statistics doesn't help much, OTOH not many (portion of totally shipped) hAP ax2 devices are used to provide PoE out hence shipping high-power power adapter would increase cost for majority of owners who don't need it (and will increase power consumption as switching power supplies run at lower efficiency when operating at partial loads). Luckily hAP ax2 uses external power adapter and if needed it's easy to replace it with a high-power (high voltage) one. After doing the power budget calculations


And while nitpicking:

PoE in input Voltage 18-57 V passive or 44-57V 802.3af/at

this could be simply written as
Nobody but nerds know the voltage required by 802.3af/at/... Not many even know the difference between "passive" and "802.3 af/at/..." and get stunned when they found out that PoE is not only one thing. And, BTW, standard requires voltage at PD (which in case of PoE-in hAP ax2 it is) of 37V-57V (af) or 42.5V-57V (at/bt Type 3) or 41.1V-57V (bt Type 4). But if daisy-chainging is to be a thing, then lowest PoE-in voltage should be around 51V (assuming not many small devices will be bt Type 4 PoE-out capable ... at and bt Type 3 require at least 50V on PoE out port of PSE, some little voltage is dropped by PoE electronics both on in and out).

As written on this forum already many times: it would be the best if MT entirely stopped the passive PoE stupidity (we're in 2024 when I last checked the calendar) and go with 54V and/or 802.3 at/bt on all new devices (no exceptions). Which includes their PoE injectors as well.