These are my first impressions with the new R5H, pls add to the topic your own experience with this new piece of hardware.

All right, i just received the new R5H which in my personal opinion is the best wireless mini-pci ever produced by mk.
Check it out the improvements:

Much better finishing - just by grabbing it you'll able see the project was very well designed and it has much better looking than the previous R52 and R52H. You can compare it to the ubiquity's.

Much needed mmcx connector - the ufls are too small and i have a hard time with the ufl pigtails and now i can use the same XR5's pigtails, so i can replace the ubiquity's with this new card without much work (and vice versa).

Much better sensibility. Techinal spec says: -96db@6mb compare to -90@6mb on the R52H and on paper it has better sensitivity than the XR5 (-94 dbm - to be proved).

Improved Heatsink - For those using the R52H at it's highest power know these cards usually have a problem with overheat and this makes some routerboards to lock. I think this problem is gone with this new heatsink.

I did some lab tests but i didn't have enough time yet to put them to work in the real world, what i plan to do this week. At the lab i saw just a little improvement in the signal strenght -1/-3 dbm but we're able to connect at higher speed comparing to the R52H.

At the MUM in Brazil i made a presentation about a 80 kmb link using XR5 and R52(as a backup link). This week i intend to exchange the R52's by the R5H and i'll put my findings here.

Below are 2 pictures of the new R5H comparing the the R52H (picts are just so so).

thanks for your review. Its been helpful
Looking forward for some tests on signal rates and strength

Has anyone seen a power output chart for these yet? They are listed simply as:

Output Power Up to 25dB (320mW)

I'm assuming that 25dB is only for 24 Mbps and lower data rates like most other cards, what is the power output @36Mbps, @48Mbps and @54Mbps?

24 dBm from 6 to 36 Mbps
21 dBm @ 48 Mbps
19 dBm @ 54 Mbps

I think it's a remarked Wistron Neweb DCMA-82 (CM-11), layout, chip code and power is the same, the only difference is the single MMCX connector instead of double MMCX or U.FL

I have many of them, i started testing them last week and for now i can say that performance-wise SR5 is better, have to compare against DCMA-82 yet but it would be no suprise if they perform similar.

Regards,
Giorgio

Well, here are the 1st results of the real world tests.

I did exchange the R52H for the R2H on my backup link. Just to remember those not present at the MUM in Brazil, this is a 76km link with a very bad fresnel, i have a telco tower just in front of my antennas. I'll try to take a good picture of this site just to illustrate how bad it is. I tought was impossible to link with this tower sitting just in front of.

Main link is on XR5 doing just great.
Backup is now running on the new R5H.

I noticied a -8/-10dbm signal increase, but no signal to noise increase (i'll try some different channels later on).
The good news is CCQ have increase from average 60% to 90%. I'm not sure yet if is only related with the R5H or a group of factors (pictails, etc). I'm still using the same antenna.

Here is the 1st pict with the R5H running.

I'll try some different tests later.

Thanks for the review. This looks like a nice radio. I am looking forward to receiving my order.

I believe the reason for no improvement in signal to noise ratio is that both the transmit power and receive sensitivity have been improved equally. So besides transmitting louder the radio also hears the noise better.

The stronger transmit signal probably improved the CCQ as it is hearing fewer bad (distorted) packets from the other end.

Just a guess but I have convinced myself.

Tom

Roc Noc, i totally agree with you. But i'll have more details soon. My team is still working with this link.
When you going to get your R5Hs?
Pls let me know when you start using them and your results. Thx

http://www.mikrotik.com/pdf/R5H.pdf

Thanks Normis... the PDF has been updated since I last looked at it, none of that info was in there before

I like the fact there is almost no power loss @36Mbps...

I like the fact there is almost no power loss @36Mbps...

Don't forget the specs states a +/-2db, the power also varies across the band.

Have you tried to put 3 of these into a 433 or 493 board? You can only get 2 of the XR5's into one, and I'm wondering with the new heatsink if this is a problem. Thanks! john

John, i did not try to put 3 of these cards in on single 433, but i can do this test for you. I'll try to do this during this week.
I did ask MK people about card limitation on the RB600 and the wrote back saying i cannot put more than 4 R5H on the RB600, so besides my test you should ask Normis and get a official answer from them.
About the heatsink, no problem at all.

Don't forget the specs states a +/-2db, the power also varies across the band.

nearly all radio's have +/-2db and varry across the band, that's pretty standard. What's NOT standard is that MT actually posts the variences across the band and not just the peak power level of the radio like most vendors do. That is very usefull when calculating links to determine what frequency / radio / antenna is best used for a given application.

What's NOT standard is that MT actually posts the variences across the band and not just the peak power level of the radio like most vendors do. That is very usefull when calculating links to determine what frequency / radio / antenna is best used for a given application.

I agree. For links where every dB counts, this information is very helpful.

Have you tried to put 3 of these into a 433 or 493 board? You can only get 2 of the XR5's into one, and I'm wondering with the new heatsink if this is a problem. Thanks! john

yea - there are no problems to put 3 cards in mine RB433.

Yes, you can put 3 R5H on the R433.
I did the test myself (as others already reported) and it's ok.
I also got the official answer from MT stating it's ok to do so.

I think it's a remarked Wistron Neweb DCMA-82 (CM-11), layout, chip code and power is the same, the only difference is the single MMCX connector instead of double MMCX or U.FL

Don't think so as the max outputs are not quite the same, plus the huge hit for me was the fact that the R5H is 5GHz only where the DCMA-82 is 2.4/5Ghz.

Also I have been apprehensive about UBNT cards throughput after discovering an issue on a 2.5km (little antenna required and high noise area) connection where we were only seeing 7M traffic on an XR5. Then I found a whitepaper on their site confirming that. (http://www.ubnt.com/downloads/XR5_Mtik_PtP.pdf). I have been skiddish to work with the SR5 as by UBNT's own words, its just a better ampped version of the SR5.

Then I found a whitepaper on their site confirming that. (http://www.ubnt.com/downloads/XR5_Mtik_PtP.pdf).

Confirming what?
The limiting factor in that test is the old/weak routerboard. The XR5 is a great radio card.

I think nobody is doubting the XR5's greatness but talking about a new cost effective competitor. I have several links running on the XR5s but i have no plans to add more, instead i'm using the R5Hs with no major issues at all. Comparing price/performance, definely the R5H is a killer.

I'm more concerned about the R5H resistance to thunderstorms. We already know that XR5s are pretty good for ESD resistance. Saving a few dollars on purchase, but costing a tower climb after a storm would be a bad bargain.

I like the idea that the diversity chip is gone as a really good start!!

Anyone who lives in a stormy area have an opinion about R5H ESD survivability yet?

George

Well, as far as i concerned, Brazil as any tropical country is hit almost every day by huge thunderstorms. I can tell by my own experience (i lived in the midwest USA, for 4 years) that we have down here a lot more daily thunderstorms than any region in the northern hemisphere (even Florida). My R5Hs are running as good as my old XR5s and cost a lot less. I had no single problem concerning really bad weather (rain and thunderstomrs) and the R5Hs, actually i had no single concern about them, they are running just perfectly. As i said before the R52H ...i don't like them (to say the minimum) but this R5H is a very different kind of hardware, buy it!

I just did some research (with google help) and i found a text about lightning strikes and Brazil: "Brazil is the world's champion in lighitning strikes with around 60 milion strikes evey year". I can't image a worst environment to check them out. And you won't see much of lightnings without a very bad thunderstom around (correct me if i'm wrong) and i hope since you live in Canada you not talking about snow storms.

Confirming what?
The limiting factor in that test is the old/weak routerboard. The XR5 is a great radio card.

Also a little old CM9 was able to pass 20M on a RB112.

We were able to reproduce the same effect on an RB600 with XR5s. We emailed UBNT and they confirmed it and then sent us the whitepaper all again.

Thanks Juliano for posting here. I was watching your presentation in MUM for this card vs XR5, so I'm impressed. We are making link specifications to make very stable link with much more bandwidth in 74km in Serbia, location Valjevo - Belgrade with offset antennas 45dBi and these two cards or XR5s, so we decided to use those cards after reading this post. We hope, we'll make success. I will post our tests.

Thank you.

mjoksimovic, just to update you, we are now running a 80 mbp/s very stable link using turbo mode. Note that the R5H has reached EOL, MK is urging every one to switch to the new R5Hn, card which i had not tested yet.
Please posts your results here.

just have done link @8.6km whitout LOS ( there are in the way some trees, then should be clear,but i dont see another point).
Link is up with R5H cards @24/12mbps 5ghz-turbo*nstream. Real test ~3.2mb/s one way.


!!! Signal to noise -2 - 2db !!!
and -78/-80dbm

wonderfull, i think we'll make some alligment with antennas and get about 5-6db stable signal to noise. When's there no traffic link goes down, after reboot could appear, but works just under load pings 1-2ms whithout load, all is perferct to me

Wireless link on 5GHz without LOS it's wasted time and nerve. Check this link when trees goes green )

Wireless link on 5GHz without LOS it's wasted time and nerve. Check this link when trees goes green )

I do not mean that trees )) they are all the time GREEN. antenn is suited @ ~23m from ground

sorry for my english )





For these link used quality antennas: SP60/54 Svenska antenni spetsialisten 28dbi 5.4ghz each side.
I think, we'll iprove some quality, becouse at the we do not have made alligment , so i think we have ~5" @ each side wrong turn

P.S. it'll be LOS, if we could up antenna for 2meters it's not made throw forest


link is up for ~12 hours, sometime goes to calibration and lost for 1-2minutes, but'snt so scarry