v1.3.0 & v1.3.1 & v1.3.2 FINAL bug reports and comments

[sirhc]

We are looking into this right now and will try and push out a fix.

IF you log into the switch and go to the Device/Status Tab what does it report for the 48V in the upper right corner?

[sirhc]

There was a routine put in v1.3.X which was mainly for the WS-12-DC unit that would prevent people from powering 48V devices if they supplied the switch with less than 43V.

Remember the WS-12-DC unit is not our WS-12-250-DC which will take any input voltage from 9-60V and allow 24V and 48V POE.

The WS-12-DC unit expects conditioned 48V input.

This routine will be removed from all models except the WS-12-DC unit in v1.3.2 and we hope to push this out within the hour.

We apologize for any inconvenience this may have caused.

[astounding]

I'm looking at the switch that petelarsen & jknight are talking about. This is what shows in the upper-right area of the Device >> Status tab:

Board 48V 12.9Vdc
Board 24V 6.6Vdc
Board 3V 0.3Vdc

That seems highly unlikely, considering it's powering 24V devices and functioning just fine except for the 48V failure.

[sirhc]

Would it be possible to ask you to down grade that unit to v1.2.4

I would like to know if it is firmware related or possibly the voltage sensors are damaged like the current sensors got damaged in this post from storm damage.

viewtopic.php?f=17&t=920&p=6965&hilit=damage#p6965

If the voltages are reported wrong with v1.2.4 we may need to RMA that unit and see if the sensors failed or got damaged.

Either way we are releasing v1.3.2 that will disable the input voltage check on all but the WS-12-DC

I am running v1.3.2 through tests right now to release shortly

[sirhc]

I'm looking at the switch that petelarsen & jknight are talking about. This is what shows in the upper-right area of the Device >> Status tab:

Board 48V 12.9Vdc
Board 24V 6.6Vdc
Board 3V 0.3Vdc

That seems highly unlikely, considering it's powering 24V devices and functioning just fine except for the 48V failure.

Another reason the board can become confused and read improper DC voltages is from BAD or incompatible SFP modules.

Does this unit have SFP modules installed and if so could you try removing them to see if the DC voltages return to normal?

It may require a reboot after the modules are removed.

Communications with the SFP modules and the voltage sensors use the I2C bus.

Also what board Rev is this (see the Device/Status Tab bottom of top left section)

If you have an early Rev B board there were changes made to Rev C and Rev D boards to help with this issue with incompatibility with some SFP modules.

But in the mean time v1.3.2 has been released which removes this input voltage check on all but the WS-12-DC model.

As well we can not yet rule out board damage from surges that may have damaged the sensors. I own a WISP as you all know and I installed v1.3.0 and v1.3.1 on all my towers and I have not seen this at all. So with that being said the possible causes which we are trying to determine are as follows:

1) Incompatible or damaged SFP Module (Board Rev may play a role on incompatible SFP modules)
2) Board sensor failure or damage

If you want us to help you determine this you will need to do some testing as described above in other posts.

[Jknight]

With regard to the switch with the grayed out 48v options, I have downgraded it to 1.2.4 and the 48V ports are working again.

Attached is a screenshot:

Also, regarding SFP modules, yes it has an SFP module, Cisco, unsure which model specifically.

[sirhc]

With regard to the switch with the grayed out 48v options, I have downgraded it to 1.2.4 and the 48V ports are working again.

Also, regarding SFP modules, yes it has an SFP module, Cisco, unsure which model specifically.

You can safely upgrade to v1.3.2 as we took that check out for all but the WS-12-DC unit(input voltage).

But the question still remains why are your voltages wrong? You will need to do some testing.

Is the voltages being reported wrong because the SFP modules are causing an issue with the I2C communications or is the board possibly damaged?

Please pull the SFP module(s) and check the input voltages (you may need to reboot after pulling the SFP).

If the voltages are incorrect without the SFP modules then either the board is damaged or something failed.

If the voltages still report incorrectly after removing the SFP modules I may ask you to visually inspect the board.

[WisTech]

Not sure if this is normal operation, but when you disable a port (to test radios, etc. without creating a loop that disables port(s) on a switch, it kills power to that port. You have to turn POE off, then save, then turn it back on, and save (all while the port enable box is disabled) to apply voltage back to the radio while keeping the LAN disabled.

[billbuchan]

We have over twenty switches, in fact all of the AC versions and are keen to try some DC with wind and solar in our deep rural relays here in the UK.

Extreme Discovery Protocol apparently there are Linux packages that will add many discovery variants into the switches which just get better and better

We're just about to install our 17th Netonix switch here in Rural Scotland.. I'm going to buy another five just to overtake you ;)

---* Bill

[Eric Stern]

Not sure if this is normal operation, but when you disable a port (to test radios, etc. without creating a loop that disables port(s) on a switch, it kills power to that port. You have to turn POE off, then save, then turn it back on, and save (all while the port enable box is disabled) to apply voltage back to the radio while keeping the LAN disabled.

Actually, I just tested this and on my switches disabling the port is not killing power to the port. What are you doing to make this happen?