I am getting more than a little perturbed over the trend I am seeing w/regards to RB951 models ‘bricking’ after what is assumed to be a power ‘spike’. I have had three now that have bricked themselves, two I was able to recover the first time. When it happened a second time the two I was able to recover are now stuck in netbooting/netinstall but will not take an npk package - so basically they are paperweights…
First off - the distributors will not warranty the item after ‘x’ period and a couple will not honor the warranty w/o a ‘say-so’ from Mikrotik. Well if the unit is bricked you can’t get a ‘say-so’ from Mikrotik because you can’t provide a supout.rif file for them to exam.
If in fact they are getting ‘bricked’ because a power spike - then how come ALL of my MT units are not getting bricked at the same location (none of the other units are 951s)? As an electrical engineer this tells me that there is an element of poor design whether it is the 951 or it’s the power supply supplied w/the 951. In my mind you should never be able to brick one of these due to a power spike unless it is so high that it actually damages the electronics (as in a lightening strike or something along that line). Both of these units were at locations that clients have their own computers connected to exactly the same electrical outlet as the these 951, yet their computers were not affected. There are in both cases outdoor MT units, also connected to the exact same outlet and they are fine (of course they are being supplied power over the PoE port by a non-Mikrotik power supply).
What I’d like to see is Mikrotik and their licensed distributors step up and replace these units. These units are obviously flawed, now whether that is a design, manufacturing, or component issue doesn’t really matter to me. What matters is a company standing behind their product. This kind of reminds me of the capacitor debacle a few years ago where the capacitors were bad on a number of 411 boards…
Don’t get me wrong folks I am a BIG MT fan. Been a user and contributor for more than 10 years… But this latest 951 issue is a real pain where I have to again shell out funds to cover the bad design/manufacture of someone’s else’s product because I stand behind the services I provide.
I have a lot of RB951Ui-2HnD and never had problem with them, but what are you describing is not unfamiliar. The electric shock does not come only from the power supply, mostly it will come from ethernet cables, specially if any of them runs outdoor.
Have you tried the reset jumper if it can cancel the booting loop? What do you get in Netinstall when you try to transfer packets? Have you tried netinstall with an older version, like 5.25?
This is a 951-2n indoor unit w/5 ethernet ports and no USB or serial port.
I have used both older and the newest netinstall versions along with ROS versions 5.24, 5.25, 5.26, 6.2, 6.14, and 6.17.
With Netinstall; I get it showing Ready for the RB in the status window, then I select the package(s) to install. Select install, Netinstall says formatting and partioning hard drive, I see very little packet flow, then there will be a burst of packet flow (maybe as long as 2 seconds), then the Netinstall window shows Ready again next to the RB.
That’s it…it can sit for hours and nothing more happens. Multiple attempts do no good, even powering it off overnight. Next day, try to turn it up it constantly reboots until you fireup Netinstall again, then you can see the unit again. But you can do nothing more with it than I have already described.
Bottom line is that there is a flaw somewhere in design or manufacturing that needs to be addressed. These units should never get bricked this way whether I can recover them with Netinstall or not. If they do brick then someone needs to be standing behind the product to repair/replace. These units were purchased only 7 weeks ago and they’ve already bricked twice.
These devices have working ESD protection on the power jack, so power spike to the jack can’t be the cause. Maybe only if you have power spike to the ethernet port as Caci99 said.
Not to be ‘snipity’ Normis but where would such a spike come from on the ethernet port? One Ethernet port of this unit was connected to a RB411 Ethernet port. The only other used port was connected to a PC. Both the PC and the RB411 are fine - no issues.
I use shielded cable with the drain wires connected to shielded RJ45 plugs - so everyone is grounded to the same potential.
There were no storms in the area. Power never went off overnight when this unit failed. I am ‘assuming’ it was a power spike based on other posts here of similiar circumstances/results. Really that doesn’t even matter - the fact of the matter is three units in different locations on different days all bricked. One I could never recover, the other two I was able to recover after the first incident. After they bricked the second time I have not been able to recover them with any means.
We would need to see the units to say more. All I know is that the overvoltage works well, and cuts off everything above 30V. Device doesn’t get damaged, just reboots.
Maybe the problem is with some other part. Please send us the serial numbers of all the damaged devices that you have returned previously, if they reached our RMA department, I could see what the issue was. Email support@mikrotik.com
The distributor would not issue an RMA to me to return the units - hence why I am perturbed over this. I can’t supply them with a supout.rif file and the units are over 30 days since purchase…
It just seems pretty damn odd that these units failed and the thousands of RBs I have out in the field, some older than 10 years, are still performing great. As I started out with…there is something inherently wrong with these units. If it had just been one I’d have figured it was just a fluke, but three…that’s more than a coincidence.
Thanks Normis and Uldis for chiming in on this - much appreciated.