R5SHpn - MAJOR PROBLEMS - Cards burn out in just a few seconds


We have a rather large shipment of R5SHpn and we are finding these wards last only a few seconds prior to going up in smoke when the WLAN interface in turned on.

I am about ready to send hundreds of RB and wlans cards back and demand full refunds. Plus the two dozen plus we went through that burned up within minutes of powering up the wlan.

Who else is seeing this problem?

Tom Jones

Did you conect an antenna to them?

If the driver does not known the card correctly this may happen

I don't think he installed them in laptops

re: Did you connect an antenna to them?
YES ( I know the impedance load on the RF output is needed and also understand SWR)

re: If the driver does not known the card correctly this may happen
This is 100 percent new Mikrotik 411s and R5SHPn

re: I don't think he installed them in laptops
Installed in brand spanking new 411s

FYI - we have installed thousands of microwave links. This problem appears to be new. We have used hundreds of 411s in the past and many R5SHPn wlan cards in the past. We have several hundred new 411s and new R5SHPn which simply go up in flames (quit working) the moment you try to connect.

Tom Jones

We have a 250 new cards, we were going to install. Out of the box, the first 10 or so quit working after initial wlan enable.

Here are some serial numbers. I have more also - but these are the ones I have in front of me right now:

Router Board 411
SN: 364904 A4B492/443
SN: 364904 7378AD/443
SN: 364904 753517/443
SN: 364904C7A8AD/443

R5SHPn
SN: 381501FFE 4B7/443
SN: 381501FFE 303/443
SN: 381501FFE 323/443
SN: 381501FFE 329/443
SN: 381501FFE 336/442

And yes - I have verified power supplies and antenna connections.

On new installation, it will scan (site survey), however when you start using it (connect) then it quits right away on 1-st wireless transmit packet.

The 411s appear to die and no longer function when the R5SHPn is removed (no PCI card).

What is the "/433" on the end of the serial numbers ?

Unless you say no, we are going to pop a new R5SHPn into a new 433 and see what happens.

HELP -

Tom Jones
FYI - we do know our WiFi. We have installed thousands...

I have a question not quite related:
Why do you use 411+card? It's somehow old style wireless. I give up a long time ago and switched to 911+antenna or SXT lite. And is cheaper too.
Anyway it may be an bad lot of cards and Rb's, but is quite hard to belive.

RouterBOOT version???
RouterOS version???

Tech spec:

RB411 max power consumption with extension card: 12W
RB411 power consumption without extension card: 3W
R5SHPn power consumption: 10W

(3 + 10 = 13) > 12

Cannot you set the power regulation of the card before you switch it on? Or are you switching it on without any settings so it works on its maximum (or maybe beyond its maximum) power level?

In the Mikrotik forum - a post from MikroTik Support:
http://forum.mikrotik.com/viewtopic.php?t=62978

janisk posted "RB411 should be capable to work with R5SHPn card"

- - - - - - -
Also - the original PDFs specifications from Mikrotik for the R5-SHPn did NOT indicate the watts the card actually would draw from the mini PCI buss. They simply stated "800mW transmit power"

When was the Mikrotik documentation changed to also include "Power consumption Up to 10W at maximum power" ?
I have two Mikrotik PDFs on the same product and one with and one without the power consumption.

- - - - - -

The problem is I have qty- 250 RB-411 boards and qty-250 R5SHPn cards we recently ordered and now I see they can't work together!!!

some findings re RB411 with R5SHPn :

1st - re "(3 + 10 = 13) > 12"
I have two digital meters, one measuring DC volts and the second measuring amps. To calculate total watts I use volts * amps.

Here are some of my findings:
-Slowing down the CPU speed of the 411 from 300 to a slower CPU clock setting does save up to almost 2 watts.
-The highest peak wattage I have seen is almost 6 peak watts with an average of less than 1/2 of that. Not 10 or 12 or 13 watts. These values were measured putting the 411 through it's paces including (speed test to self, wlan speed test up & down, heavy internet access and cold power on).
- It appears that the amps drawn is a constant and is not proportional to the input voltage when tested using 11 to 24 volts via power jack or POE over Ethernet. ((The amps drawn remain the same - even if you increase the voltage !!!) Thus I suspect the 411 internal voltage regulator is acting more like a vari-resistor (different resistance at different voltage - or like a cheap linear power supply). Additionally - the formula Watts = Voltage * Amps is not a constant across various voltages. With the 411 the only thing a constant is Amps. As Voltage increases the total Watts increases. Any good switching power supply input load will decrease Amps as Voltage increases to maintain the same regulated output voltage under load - but not the 411 on-board voltage regulator. ( I am stating a 411 draws more Watts with a higher voltage POE than it draws with an 18 volt POE or a 12 Volt POE).

I next performed some power and POE testing. Here are my findings:
- The RB-411 with a R5SHPn installed operating with a 12 volt power supply connected to the power jack appears to work good. No burn-outs or problems.
- The RB-411 with a R5SHPn installed operating with a 24 volt POE will fry the 411 with the same configuration - as tested good prior with a 12 volt direct power supply connection.
- The RB-411 with a R5SHPn installed operating with a 18 volt POE appear to work ok - I am still testing them on the bench.

At this time, I am beginning to think the 411 with a higher amperage/wattage PCI card actually has a problem with the 411 on-board voltage regulator when the input voltage is greater than 18 volts. In other words, the problem does not appear to be total watts, it appears to be input volts when under load.

Can somebody with some inventory please confirm my findings?

If I find the 18 volt POEs with a 411 and R5SHPn mini PCI card has problems during my continued testing, then I will re-post my additional findings / warnings.

North Idaho - Tom Jones

EDIT: The 18 Volt POE I am testing with is a Laird 18V PoE Supply +Surge

EDIT: Thinking about it, I would suppose the excess wattage dumped by the on-board voltage regulator as voltage goes up is lost as heat. Thus in very cold icy tower locations I would guess a hotter voltage might keep the RB boards warmer. And for very hot locations and energy sensitive locations (solar cell & battery), the lowest voltage you can reliably run at could greatly reduce heat and save up to 50 percent energy from a limited supply. Anybody - any thoughts ?

I have a question not quite related:
Why do you use 411+card? It's somehow old style wireless. I give up a long time ago and switched to 911+antenna or SXT lite. And is cheaper too.
Anyway it may be an bad lot of cards and Rb's, but is quite hard to believe.

My answer to your question:

RB911-5Hn Max TX power 23dBm
R5SHPn Max TX power 29dBm

(29 - 23 = 6)

The R5SHPn has an additional 6 dBm transmit power. The extra 6 dBm transmit power makes it 400 percent more powerful. Every additional 3 dB is double the power - thus 2x2 (3 dB + 3 dB).

And if I remember by math correctly, that effectively almost equals a 200 to 400 percent greater working distance diameter operational range greater than the RB911-5Hn. I need the distance because many of my multi-point customers are over 10 miles and I want a high CCQ and a high connection rates and remain operational when the worst of the snow storms come through our area here in North Idaho.

Note - we have verified 150 meg rate 100 percent CCQ connections at 18 miles connected to a multi-point 90 degree sector antenna and can easily pass 100+ meg through this link on a speedtest.net bandwidth test. The only reason it was not faster is because my upstream internet connection was only 135 meg at the time. These days we have a 10-gig fiber connection to our equipment.

UPDATE:

My techs have tested almost 3 dozen with no problems.
R5SHPn
RB411
Laird 18V PoE Supply +Surge

Note - the combination below results in burning out 100 percent all tested configurations:
R5SHPn
RB411
24 Volt POE

At this time - I would suggest never using a 24 volt POE on a RB411 and I suggest only using up to an 18 volt POE on a RB411. The RB411 on-board voltage regulator just can not properly handle the higher voltage without burning up when under a heavy load. The additional voltage is simply just dumped as heat on the RB411.

It is possible the RB411 boards that you got has some problem and that why it stops working with the R5SHPn card. Try to put the card into different RouterBoard and check again.
For the faulty RB411 boards please contact RMA.

It is possible the RB411 boards that you got has some problem and that why it stops working with the R5SHPn card. Try to put the card into different RouterBoard and check again.
For the faulty RB411 boards please contact RMA.

- Every RB411 with a R5SHPn connected to a 24 Volt POE we have tested has resulted in the RB411 burning out. We have tested more than a dozen (new - out of the box). If you perform some google searches, you will find this appears to be is a common problem.

https://www.google.com/#q=RB411+R5SHPn+problem


- Every RB411 with a R5SHPn connected to a 18 Volt POE we have tested has worked. We have tested more than a dozen (new - out of the box).

-Every R5SHPn mini PCI card works in a RB433 with 12 Volt, 18 Volt POE and 24 Volt POE. Even if the R5SHPn was in a RB411 that burnt out during the time the mini PCI card inserted.

I will be contacting my distributor for an RMA/exchange on the new RB411 boards that have burnt out (about two dozen). I will also be exchanging about 150 new 24 Volt POE injectors for 18 Volt POE injectors.

Note: Mikrotik changed some specification capacity capabilities on the Mikrotik documentation for the RB411 and R5SHPn. The original Mikrotik documentation did not state the RB411 mini PCI power availability and the original Mikrotik documentation for the R5SHPn did not state the power draw requirements. If you look at the newer Mikrotik documents, the RB411 clearly states it can not supply the wattage needed for a R5SHPn mini PCI card. The total wattage is 1 watt higher than the RB411 capability (as noted in newer Mikrotik documentation).

The newer Mikrotik documentation for the R5SHPn has an additional line stating:
Power consumption - Up to 10W at maximum power

10 Watts is a friggen huge power draw to generate 29dBm .8 Watts (point eight watts). That would mean the card draws over 9 Watts to generate .8 Watts in RF energy. Additionally, my measurements of the R5SHPn mini PCI card were less than 1/2 of what Mikrotik states in the documentation.

I suspect two things:
-1, the R5SHPn has a peak draw of less than 5 Watts.
-2, the on-board power supply of the RB411 utilizes an insufficient capacity linier voltage regulator. As input Voltage increases from 12 Volts to 18 to 24 Volts, the RB411 on-board linier voltage regulator becomes unstable under load. The total Watts the RB-411 on-board voltage regulator can handle at 24 volts is about 1/2 (half) of what it can handle at 12 volts. The difference between 12 Volts and 24 volts is just dumped as heat through the RB411 on-board voltage regulator.

In summary - in my opinion - - - NEVER use a 24 Volt POE on a loaded RB411 (use 12 or 18 Volt POE - only).

Can anybody including Mikrotik please check and verify my findings - thank you

North Idaho Tom Jones