I'm really starting to get worried. R52H failed yesterday on a tower install. It had been running great for 5 months. Most of the failures have happened in a few days of install though. It looks like about 10% failure rate.

Almost all of them are in Rootena (R2T58W-19) installs in RB133's.

None are going to Omni's, as talked about in previous posts.

Any idea's?

We just received 10 defective R52Hs, right out of the box ... the best one had a signal strength of 29mW.

I've NEVER had issues before with MT hardware. Have any idea about warranty / RMA's?

We just received 10 defective R52Hs, right out of the box ... the best one had a signal strength of 29mW.

How do you test cards when you receive them new? Equipment, software, setup, etc? I have no clue and i'd hate to get a radio setup and up a roof or tower or wherever to find out it does not work well even if it does in the office right after being set up...

could anybody enlighthen me please?

Nemecio.

Up until this point, I've not really tested prior to install... When we first began working with MT, we tested setup's on the test bench with some small patch antenna's.. Then we developed the standard setup for AP's, Backhauls with WDS bridges, Client CPE's with NAT, and Client CPE's with WDS bridge. I've got a backup file for each that we ftp to the new board, then go in and restore, then setup the wireless settings which don't restore.

I do have extra coax going from test bench to roof antenna that connects to a nearby AP, and use it for testing coax jumpers etc... and now we use it to test suspected bad radios.

Are these radios being run with default TX power?

Has it ever changed from anything besides default even for a few seconds?

I used to change them to fixed at 26, but more recently, I've left them at default.

I used to change them to fixed at 26

There you have it. Overdriving cards will blow them every time.

First off, it was unclear whether the default power setting actually worked in earlier versions of MT. However, we did start leaving it at default in later versions.. The cards STILL FAIL with default settings.

Second, the power output rating on the R53H is 350mw... There YOU have it.

I have not found any single factor across all the failures.. One was in RB133 with long CAT5 and 18v POE, so I thought it might have been insufficient power.. So we've only used 24volt since then, and all of the subsequent failures have been short CAT5 runs.

350mw is less than 26dBm. Do the math.

Yes 398mw. But then again, that also depends on whether the card was actually transmitting at 100%. That's why when the radio is idle there is low power draw by the radio. In fact, at idle, there is close to 0 mw coming from the output.

Having ###holes on the forum gets us NOWHERE.

But then again, this has gotten off thread... I've made the point that the 26dbm setting (mainly used in early SR5 days) hasn't been used in a long time, changed after the first few failures.. The failures CONTINUE.

The diversity chip referred to elsewhere seems to be the question mark. Does anyone that has any USEFUL knowledge of the R52H care to comment?

Call me what you want. It's not my fault nor Mikrotik's that you are too stupid to read the manual and understand if you overdrive a cards power - even for a short amount of time you will blow the card.

If you have doa cards out of the box that is life. We're dealing with ESD sensitive devices - it happens.

Five months and it quit working? Did you overdrive the power? Are you using a polyphaser on your antenna? Are you using shielded CAT5? If so, is it grounded to the same ground as other equipment and most importantly the tower? Are you using a CAT5 surge protector?

There are a hundred reasons why failure occur. Copping an attitude and blaming MT for your stupidity just make the MT guys and myself sit back and laugh - it keeps the forum lively.

I originally posted the message about getting 11 out of 15 R52H cards DOA about 4 weeks ago. We stopped using them right then.

All of our settings were left default (and always have been for the last 6 years using MT). Mikrotik has a QA problem with these cards right now. We even had 1 fail on a tower that had been working for 2-3 weeks, with all power settings at default.

How are you going to blame MT for something that has been working for 2-3 weeks?

I have 5 x Gigabyte minipci with 19dBm default power and they are working fine at 23dBm. My opinion is that R52H are low quality cards.

There is definately something up with them.. Since around the 1st of the year, the default tx power setting has been used, and the failure rate has increased. A problem that we never had with R52's or SR5's and XR5's.

Sure, the one that failed on a tower after five months may very well have been attrition, although we usually only have those problems with severe lightning etc....

My question is whether it's something with the Rooteena. Our environment is very windy- maybe a static problem that no other cards have been affected by.

I'm not blaming anyone.. There may be a logical reason for the failures. We just need to know as there are probably 100 installed now.

How are you going to blame MT for something that has been working for 2-3 weeks?

Because it should last longer than that. If you bought a new car and the transmission went out after 1000 miles, would you blame the dealer?

No offense, but you seem to be the one who have the attitude lately. Don't get me wrong, you have many many very good posts on here which have been very helpful for users, but lately I have noticed an extremely high number of posts from you which simply say "read the manual". Not really helpful.... Yes, people could do a better job of asking questions, but you can simply say, give us more info so we can help.

At any rate, randallr has made it clear that they have not changed the power from default for a long time, so I would say it's time to drop the "you screwed it up" theory.

Again, don't take it as me yelling at you... just saying you can ease off a little.... Be cool!

Joe

Ok, I will re-phrase.

What is your exact setup end to end including what type of power supplies, routerboards, grounding, surge protectors, poly phasers, etc?

jwcn,
Saying someone doesn't know what they are doing instead of looking at a possible problem shows YOUR ignorance.... You come across as a real know-it-all who has everything perfect. May your karma bite you in the a... Now for our experience in the R52h debacle... We have a stack of 10 r52h's that have failed on the bench and in the field. There are two failure modes. Most often it looks like the antenna switch blows due to something we suspect is static discharge. This has been with 12V battery power, 12V poe, 24V poe, whatever... They were very well grounded to common grounds. Some had polyphaser discharge protection. We have done everything we could to protect them. In the very same box and on some of the very same boards, SR5's have shown no problems. We have also seen R52H's with blown amps. Blown as in a small hole in the amp where they exploded. The three that this happened to all experienced antennas that were swung off line ( 90+ mph bent the brackets twice and ripped it off the wall once ). It's like the amp tried to compensate for the slowly changing signal and went farther than it should have. We now are off R52h's Back to Ubiquiti for highest power and trying the engenius for a little less.

I don't claim to "know it all" I do have ample experience making the same stupid mistakes.

So far, the only installs that the R52H's have failed in are Rootenna installs with the plastic standoffs. Don't think any of the ones that are better grounded enclosures have failed, but may just be coincedence. None of the regular R52's have failed, or SR5, XR5's.
Most seem to just loose their signal level. I'll look at the logs and it's a quick drop off to either next to nothing or nothing.. A few of them occured right after changing channels on the associated AP- like the CPE with the R52H blew itself out trying to re-associate.

I have tried the WiliGear 400 card and seems to work good. Although it will not run in RB411- that board shuts off right after boot. The RB133, RB333, and RB600 all worked fine. Haven't done any benchmarking yet with this card, but it may be a decent replacement for the R52H needs at a similar price.

I originally posted the message about getting 11 out of 15 R52H cards DOA about 4 weeks ago. We stopped using them right then.

All of our settings were left default (and always have been for the last 6 years using MT). Mikrotik has a QA problem with these cards right now. We even had 1 fail on a tower that had been working for 2-3 weeks, with all power settings at default.

make sure the equipment is properly grounded, that they have never been used with the pigtail disconnected and that the power settings are default (you already said they are)

make sure the equipment is properly grounded, that they have never been used with the pigtail disconnected and that the power settings are default (you already said they are)

Our equipment is all in metal enclosures and well grounded. We always use either default power settings or "card rates" with power betwen 4-10dBm. We still continue to have problems with R52H cards dying. Sometimes they lose receive sensitivity. Sometimes they lose 20dB of transmit power. Sometimes a single card will prevent the routerboard from booting. Sometimes everything looks good except the latency and jitter on a link is terrible. Replace with an R52 and everything is good. (too bad I don't have an unlimited supply of R52).

Sometimes a single card will prevent the routerboard from booting.

As reported elsewhere on this forum I've just returned my first and last RB333 to my distributor because of booting problems. Interestingly it gave no trouble when I took it from the envelope, sat it on the testbench and connected to it via ethernet - I upgraded it, configured it as far as I could and installed it in a Rootenna. Then I populated the card slots with two R52H's and an Engenius EMP-3601 and the problems began - the board would power up but only boot occasionally. And this continued even after the cards were removed.

I assumed it was the 333, but now have to wonder.

I'm getting very uneasy.

Bill - that's an interesting image. Could you;

1. Tell me where you got it? It looks like something out of Winbox but I can't find it's equivalent in any Winbox of mine. Is it related to a particular Board or RouterOS version?

2. Interpret it for the amateur/thicko class (eg. me). It seems counter-intuitive to me, in that I thought -75 was a weaker signal than -69 so you'd be looking for a lower throughput rather than a higher one.

3. What is the TX CCQ? My best guess is that RX CCQ is the % of uncorrupted packets received c/f the total number transmitted, but unless the device at the other end is relaying information about its own RX CCQ in the background I can't envisage what TX CCQ is.

4. If the R52H was dead when it was installed, how is it you are receiving anything?

Bill - that's an interesting image. Could you;
1. Tell me where you got it? It looks like something out of Winbox but I can't find it's equivalent in any Winbox of mine. Is it related to a particular Board or RouterOS version?

My winbox (latest version) shows the same, but actually I noticed these changes after start using ros3.x family. So could be both or either one of them.

2. Interpret it for the amateur/thicko class (eg. me). It seems counter-intuitive to me, in that I thought -75 was a weaker signal than -69 so you'd be looking for a lower throughput rather than a higher one.

SweetSunday, It can be I am an even more thicko, but don't understand your question. But actually don't understand what info Bill was trying to tell us here to start with....

3. What is the TX CCQ? My best guess is that RX CCQ is the % of uncorrupted packets received c/f the total number transmitted, but unless the device at the other end is relaying information about its own RX CCQ in the background I can't envisage what TX CCQ is.

Good question. I never understood how one card can tell how its signal is received on the other end?[/quote]

4. If the R52H was dead when it was installed, how is it you are receiving anything?

Good question! The graphs shows the other card is sending well, by showing the receiving result of this end. I don't see what is wrong with the remote card apart from the fact that the signal strengths between this end send and the other end receive is a bit different. But that can have many other reasons.

Indeed, I have to agree with SweetSunday that I don't understand what it is Bill is telling us here.....

Bill - that's an interesting image. Could you;
1. Tell me where you got it? It looks like something out of Winbox but I can't find it's equivalent in any Winbox of mine. Is it related to a particular Board or RouterOS version?

My winbox (latest version) shows the same, but actually I noticed these changes after start using ros3.x family. So could be both or either one of them.

2. Interpret it for the amateur/thicko class (eg. me). It seems counter-intuitive to me, in that I thought -75 was a weaker signal than -69 so you'd be looking for a lower throughput rather than a higher one.

SweetSunday, It can be I am an even more thicko, but don't understand your question. But actually don't understand what info Bill was trying to tell us here to start with....

3. What is the TX CCQ? My best guess is that RX CCQ is the % of uncorrupted packets received c/f the total number transmitted, but unless the device at the other end is relaying information about its own RX CCQ in the background I can't envisage what TX CCQ is.

Good question. I never understood how one card can tell how its signal is received on the other end?

4. If the R52H was dead when it was installed, how is it you are receiving anything?

Good question! The graphs shows the other card is sending well, by showing the receiving result of this end. I don't see what is wrong with the remote card apart from the fact that the signal strengths between this end send and the other end receive is a bit different. But that can have many other reasons.

Indeed, I have to agree with SweetSunday that I don't understand what it is Bill is telling us here.....[/quote]

My understanding is that TX and RX signal strengths should be very close to each other if you're using the same card on both ends, with the same tx-power settings and you're using the same antenna to tx and rx (ie. you're not using the tx-a/rx-b feature). Bill has 10db of difference, which means that one card is not transmitting with as much power as the other, or one card has less rx sensitivity than the other.... or both

When both wireless devices are MikroTik ROS, the Wireless/Registration window you're seeing does show you both RX and TX signal strengths because the receiving MikroTik box sends the RX information back to the sender. In fact if you try to connect a non-MT box (eg. a Windows laptop), you will only see the RX information in that window.


To contribute my 2c to the list, we used to use CM9 cards and never had any strange problems. Only recently we switched to R52H's, and I have already replaced 3 brand new R52H cards this week.

i also have a bad problem with R52H for 3.3km

site A : RB532 - R52 - Grid 24dbi - signal -60 - apbridge 2.4Ghz-B only
Site B : RB600 - R52H - Grid 24dbi - signal -80 - station 2.4Ghz-B only

then i change R52H with R52 and signal more good
site A -50 and site B -55

My understanding is that TX and RX signal strengths should be very close to each other if you're using the same card on both ends, with the same tx-power settings and you're using the same antenna to tx and rx (ie. you're not using the tx-a/rx-b feature). Bill has 10db of difference, which means that one card is not transmitting with as much power as the other, or one card has less rx sensitivity than the other.... or both

I agree, partially. I have just had a case that an installer of mine fit a RIC/522C that was to point at an rb333 unit with sector antenna. Since the sector was higher then the CPE he should tilt the RIC/522C upwards which he forgot to do. The result was that the RIC/522C had a strong signal strength from the AP but that the AP had one that was almost 20db lower! It was even so bad the connection regularly dropped. By pointing the CPE upwards the problem was solved and both readings became almost equeal (1-3dbm difference).
Conclusion: it can also be miss alignment of antenna's.

When both wireless devices are MikroTik ROS, the Wireless/Registration window you're seeing does show you both RX and TX signal strengths because the receiving MikroTik box sends the RX information back to the sender. In fact if you try to connect a non-MT box (eg. a Windows laptop), you will only see the RX information in that window.

I never knew this. Thanks for that info!


To contribute my 2c to the list, we used to use CM9 cards and never had any strange problems. Only recently we switched to R52H's, and I have already replaced 3 brand new R52H cards this week.[/quote]

I had a lot of R52H cards this winter that saw some failing. But the failures dissapeared when I put them in another unit (same rb though). Then they suddenly worked fine again. This way out of the lot of 10 I had to replace 6 in other boards, 3 always worked fine and 1 seems to be really dead.
Later lots I never had anymore problems with....
Must have been a line of ´monday morning product´.

Can someone please tell me how do I check if some R52H card is having hardware problems using the winbox utility (or other tools) ?

Can someone please tell me how do I check if some R52H card is having hardware problems using the winbox utility (or other tools) ?

gi,

in winbox under wireless go to "registration", double click on the link you want to see details on under "signal" and status

If an R52H is ill it will transit a lot lower than it should.

Thank you for your reply, but I dont have a clue what are good and bad values. What TX power should i expect? -17dbm is stronger that -25dbm? A TX/RX CCQ of 30% is good or bad? Where should i look, and what range of values should be normal?

TL

I also have about 20 R52H cards here now which are all DOA. They are about 20dB down on the TX for 5GHz only, 2.4GHz is ok. These were all installed by experienced technicians and never worked. I still have 2 boxes of 50pcs and am quite worried as I can not get any help getting them replaced as they won't RMA them. I have never had a Ubiquiti or EnGenius mini PCI card failure in 3 years, so it is not the installation procedure.

I am a fan of r52s.......
But, since a while, i am also a fan of china products..... I have about 20 wireless mini-pci from china and taiwan, don't mention names, for not to be erased.... ;;), and about 30 r52 and 4 r52h, taken all in about half a year span. No china has died, but one r52, and all the r52h are not listening to me anymore......
Maybe it's coincidence. Maybe not. Sure thing, for me, in some cases, unfortunately, china does it better. I have boxes untouched for more than 30 months. With both MT and china cards. Think that r52 is anyway, younger.

Anyway, 4 of 4, for r52h, is a terrific percentage of deaf cards. 19-22 dbs difference, in same box, wireless survey. Facing _anything_ else.

Well, it seems like just a list of failures, with nothing of use, for repairing this.
If it is indeed the diversity switch on the card that fails, is it anyone who has succeeded in replacing/disabling it ?
I mean, can you use the card in anything but indoors for 50 meters ?
If anybody did it, post here.
Complaining without doing nothing, and mt telling to check ground and sh.... is no use.
Sorry, i'm just sad....

Well, it seems like just a list of failures, with nothing of use, for repairing this.
If it is indeed the diversity switch on the card that fails, is it anyone who has succeeded in replacing/disabling it ?
I mean, can you use the card in anything but indoors for 50 meters ?
If anybody did it, post here.
Complaining without doing nothing, and mt telling to check ground and sh.... is no use.
Sorry, i'm just sad....

its either the diversity switch or the pa that goes.. infant mortality seems a big problem.

they can be repaired if you have the reflow machine to remove/replace the parts (we have) but its really not worth the hassle.

going to see if we can source sr5's instead, have a choice of connector then too.

why not XR5 s kijomabb ? what are your experiences with them ?

Same thing happening to me here:

I have over 400 units in the field and used sr5 until a few months ago... switched to the R52H because of reasonable cost. NEVER and I mean NEVER had a failure on an SR5 out of the box.... or at an installation... 15%+ of all our R52H's are failing out of the box OR a few hours after deployment.... symptoms are dead TX or 10+db loss...

Mikrotik needs to do something about the QA on these... it is a big problem, and hassle.

I use DC grounded antennas, RB133c's and RB411's properly grounded with an 18V power supply.

Same Here!

First it was the SR2’s, I swapped them out for R52H’s hopping that would take care of the issue. Now I’m having the R52H’s fail! … NOT GOOD!!!

I’ve started to use a ground wire as described in (http://forum.mikrotik.com/viewtopic.php?f=7&t=9734) So far so good! But I shouldn’t have to do this!

What’s MikroTik's position on this issue?

RMA?

If not I will have to abandon MikroTik cards.

This has been a VERY informative thread. On buying my MikroTik CPE's, my supplier populated the MPCI slots with Compex CX-WLM54AG. I was biased towards the MikroTik brand and asked for an upgrade to R52H. At the same time, I came across this thread and instead asked for the Compex CX-WLM54AG23... no stock! So ended up with some nice rock solid R52's. All's well that ends well.

Same thing...never had an 8602 or a sr2 or xr2 or r52 fail.


Out of 5 R52Hs i just deployed 2 have falied with in a month.

The issue on these two are very irratic TX...no one (laptop users) can connect, and we were seeing 10-20 users steady, before the card change (antenna is Routerboard are the same).

any meaningful reply from MT? with out a decent reply you guys are loosing tons and tons of card business.

what about the routerboard Crossroads? does it use the same as the R52H?

tks

we have crossroads used for WiFi indoor AP's and no failures yet. real pain they arrive region/power locked though making them illegal to use here until we netinstall em!

Without wanting to go off-topic, what's with netinstall on crossroads then?

Lost yet another R52H this morning due to thunder storm! It went deaf!

We have continued failures, with no thunder storms.... seems to be random. I have a stack of 30+ cards to send back.

Mikrotik needs to address this issue.

Since MikroTik has not addressed this issue I have stopped using the R52H cards. I have even stopped using their boards since the ones I want to use wont fit inside the Pac wireless enclosures that I have. I can believe that they retired the 532A. They should at lease have come out with a replacement that had similar board size.

I actually am very pleased that the 133, 411, 333 and 433's all have the same hole spacing.. it has made things very neat for me.

hytanium,

What are you using for an outdoor enclosure?

We use a custom enclosure imported from China.

Here is an UPDATE:
We were running short of cards, so I decided to test every one of the R52H's (40 in total) we had ready for RMA just to make sure our techs didn't slip in some good ones.

We tested every card on the same RB411 and external antenna set-up.

Out of all 40 cards we found 13 that would associate and seemed to have the same RX/TX levels. which means 27 were still bad. With the bad ones we jiggled the ufl to see if it was maybe a bad connection. This did not seem to help.

We think that maybe some of the cards that tested good may quit after running them for a few hours. We have 3 of them in boards and powered up to test this theory. I'll let you know how we made out.

Regardless these dead cards cost us alot of truck rolls whick means $$ down the toilet.

KNOCK on WOOD, I haven't had any dead r52h's yet or any that have failed me in the field.

strange failures of R52Hs for us as well. No rain or thunders

Hytanium, any clue about you enclosures?? i would like to check something cheap.

Post your mail address and I will send you some photo's of what I use in the field.

BTW... the retested cards that I let burn for a day stayed associated... the R52H issue seems to be totally random.

Pac Wireless will be coming out with a new enclosure in the 3rd quarter! It will allow for enough room to use the 433!

No R52H failures here yet either. We are using most of our R52H for 2.4 frequency though. Very few for 5.x.

hytanium: gujack20@hotmail.com

I have had 4 out of 10 doa. No more of those cards for me. I have never spent so much to have that many doa. Then one died within an hour of use. No overpowering. Every other card has worked fine in my routers, I just wanted a less noisy card. Oh well. I will stick with Ubis now.

r52h are fine for me, we have 100's of them all running 2.4-10mhz with default power in a rb with 12v psu
All mounted in rootenna

we have some fail but no more than others (cm9 etc)

sometimes we get a few that are not so powerfull or bad rx (Doa) I just put them in close customers.



P.S I say these things not for brag but incase my setup helps someone.

To add my 2p worth : we have a mountain top site with 3x R52H. The radio servicing the backhaul to the site has failed three times this summer. Either completely broken, or output signal down 20dB. Each time after a thunderstorm (any lightning within a few miles of the site seems to blow the R52). The site has run with older 2.4 radios for years with no problems due to EMP.

Hi Guys

We have had similar situations.
R52H running at fixed rates of 20dBm, due to our limitations, was running in AP mode on 5.8GHz on a sector antenna. We then put it into scan mode to see what other networks we pick up and perhaps change to a channel with less noise and traffic. After the scan, no clients registered. We ended up using this radio for the 2.4GHz band instead as it still transmits and receives on that band. The only reason why we got the R52H in the first place is due to the fact that our local distributer, as usual, didn't have stock on the R52 at the time.

I have recently done a install with a R52. I tested it at home on 2.4GHz all fine. I basically took it as is, in it's enclosure to the installation site where we simply used a 5.8GHz grid instead of the 2.4GHz grid. We can receive at -69dB but the other end doesn't see us at all. Not even at -95.

Last week, a friend installed his R52 after finally upgrading from his D-Link to Routerboard and was very happy to obtain good speeds on a stable -75dB both sides link. 2 Days later the signal went down to -85dB from his side and it seems to fluctuate with a few dB's (-83 to -87)

These are just some of our experiences.... beleive me I can go on and on and on and on...

What is noticible though, is that only Mikrotik's radios seem to be doing this. We get the Compex cards at slightly cheaper and now only use the R52 when our supplier, as usual, doesn't have stock on the compex.

Personally, I'd rather wait a week for the stock to arrive than have to buy 2 items and having to do the installation twice on the same location. Some of our sites are extremely remote, meaning that if these things fail, the links will be down for at least a week or two.

In order for me to buy the Mikrotik boards again as a first choice, they need to show that they have re-developed their cards entirely. I want a new board layout before I'll beleive they fixed their crap. And when the first one fails, they'll move straight down to the bottom of my list again.


PTAWUG is a free public wireless network.
http://www.ptawug.co.za
http://www.wug.za.net

hi,

its sad to hear but the r52h's are still an issue. One supplier we have only does r52h, no ubuiqiti so if we need the full band c ERP capability in a hurry we have to get them.

We test each one in a radio before they go out, 25% of them are DoA with low output power or bad rx/tx, some of the ones that get through this fail in transit after this screening! (i.e. not powered but inside a built unit in a static free van.).

If you can get them in service without these failures they will usually go on fine (until lightning gets near them).

I am sure MT will say "we sell a lot of R52H's so it isn't a problem", what they probably don't appreciate is a good 30% of these purchases end up in a bin or early retired to 2.4 ghz use when the 5ghz doesn't work.

So sales are higher than they should be as we always buy 50% more than we need to allow for the duff ones.

how about a dedicated 5ghz card MT, remove the 2.4 crap, fit a quarter wave ground short on the output to deal with lightning proximity, stop pushing the little power amps to death and fit a meatier one (sirenza) . Ubiquitu can so why not MT?

they have done it...

check R5H
http://www.mikrotik.com/pdf/R5H.pdf?md5=1226606604

We have purchase 20 pieces of these cards about 2 months ago,
every week we have one or two cards failing. mainly they can Receive
signal but no Tx (almost zero)

any idea what should we do? send to reseller or MT? We're in Malaysia and
bought them from US.


Thanks

We have purchase 20 pieces of these cards about 2 months ago,
every week we have one or two cards failing. mainly they can Receive
signal but no Tx (almost zero)

any idea what should we do? send to reseller or MT? We're in Malaysia and
bought them from US.


Thanks

which cards now? R52H or R5H?
Always send to reseller, the seller has obligations to his customers.

We have purchase 20 pieces of these cards about 2 months ago,
every week we have one or two cards failing. mainly they can Receive
signal but no Tx (almost zero)

any idea what should we do? send to reseller or MT? We're in Malaysia and
bought them from US.


Thanks

which cards now? R52H or R5H?
Always send to reseller, the seller has obligations to his customers.

What if you can't get in hold of reseller? (they may be out of business).
i've only had one bad R52 that can't Tx but Rx seemes to work.

I have four R52 and two R52H, gotta check the stock to count again.

Is R5H unreliable in same way? we have many downtime in network in critical locations.

no, R5H is a very good and reliable card. watch http://www.tiktube.com/?video=196

When you buy something, the seller probably gives you some kind of warranty. you must contact them in case of problems. I cannot help you if they are not cooperative. If I buy a TV in a local store, I will bring my broken TV to them, I would never even think of sending it to Sony headquarters in Japan.

I have been using the R52H’s for quite awhile now, and what I have learned is that they work great in a CPE “Rotenna”, but the failure rate is too high when used as AP. I have also seen that there is a problem trying to return defective ones. Ubiquiti Will exchange no questions asked! I use the SR2 and XR2’s in my AP’s. AL-LSXM-ME Lightning Arrestors are a must when attaching to an omni antenna.

no, R5H is a very good and reliable card. watch http://www.tiktube.com/?video=196

When you buy something, the seller probably gives you some kind of warranty. you must contact them in case of problems. I cannot help you if they are not cooperative. If I buy a TV in a local store, I will bring my broken TV to them, I would never even think of sending it to Sony headquarters in Japan.

having a support line is not relate to headquarters.
perhaps you can check Dell support system to find out how
they provide support for IT related equipments.

a reseller is usually a shop which does not have staff or facility
to check whether your tv is broken by or bad quality.

Thanks

Well, we already know that the R52H has some issues which MK should fix it, but compare this model to the new R5H is not fair. R5H is a complete different kind of wireless card and i can assure you it does't have any issues at all. By the way i think it's better than the XR5 and it has a better buck/performance.