Just installed a 900Mhz solution, 150' tower, Pac Wireless 11dbi Omni with 1 degree electronic downtilt, Clint rig is using 13dbi Pac Wireless Yagi. I'm getting very erratic behavior out of it.. The 2.4ghz solution is doing way better than the 900mhz as far as signal strength etc.. 2.4 Has better range also.. I'm totally stumped as to the problem. Very Rural area, not much in the way of interference as far as I can tell. What things do I need to look for with this?

Thanks, Connor

Come on guys.. Any ideas?

what is the noise floor at? when you do a freq. scan what do you get (i dont know if this works on the 900 mhz stuff yet.)

why not use 2.4ghz?

We do have 2.4Ghz, but, we also want the 900mhz for NLOS. The Freq scan doesn't show much on the spectrum.. But, I'm not too show how well that works.

Noise floor is around -99

but 900MHz is not free...

regards,
C. G.

We are also finding that the SR9 cards are not very stable. We had a stable link with them using another OS but Mikrotik does not seem to have the same stability. Could it be a driver issue?

Cheers,
Dan

Typically, you should not put 900Mhz so high up on a tower, at 150', it will pick up too much noise, it's usually recommened to stay around the tree line and it will perform much better. You should also never put two 900Mhz radios near each other unless you have REALLY good seperation.

i dont know much about these cards, but is 750mw too much for these boards to drive and still be stable? what type of power supply?

They actually draw less power than a SR2 or SR5, anyway's I think I found the problem.. It looks as if I was able to get a signal from over 1.7 miles away with the dang thing using the wrong antenna port!! I switched it from Antenna A to Antenna B, and it started working at around -65 @ 1.7 miles instead of -89 or so.. As for mounting the 900Mhz up at 150' I this area.. it's what's needed.. Way too many hills and such to mount it lower..

Connor, I think your "dumb" mistake, just proved to everyone how powerful these cards are! Thanx for that, you gave me half the answer of my big question, how good is SR9?!

Has anyone ever had the opertunity to compare SR9 or 900MHz with Wimax/Pre Wimax?

I would also like to know how many non-overlapping channels there are on 900MHz, and how bad overlapping channels will affect each other. My experience taught me that 2.4 is much more affected by interference on overlapping channels than 5.8, is 900MHz even better?

Perhaps a few of us could get together and setup a simple page with all the basic info everyone needs to know. Eg the fact that 900MHz shouldnt be too high, that your isolation should be excelent etc etc. Any ideas?

Re,G

SR9 rb112 and 8km link with no line of site. signal strength is 85-87dBm but we have no link actually. the station and bridge sit connected all time, but ping is bad. noise floor 97, 2427 2GHz-10MHz, antennas 15dBi Yagi.

best regards,
C. G.

i dont know much about these cards, but is 750mw too much for these boards to drive and still be stable? what type of power supply?

we consider stable behaviour of one sr9 card on RB112 with any poe.
we replace rb with pc and results we see were almost the same.

regards,
C. G.

Don't trust the noise floor readings. MT 2.9 is not optimzed for an SR9 and you have to take the noise floor as an "indication" rather than the abolute truth.

We have a lot of SR9s in the field, they work well on 112s and 532s. We NEVER, EVER install one with less than a -75 signal, stronger than that if the area is noisy.

Also, you are unlikely to be able to use more than the 5MHz channel. Its remotely possible, but unlikely as 10MHz or 20MHz will see too much noise.

How high you can mount depends where you are and what your noise environment is.

I have one Vpol omni at 300' that works fine, we have a 9dB horizontal yagi at 20' that causes intermittent problems. Depends completely where we are.

You should mount as low as possible, and sectorize as much as possible. Plan your network for an increasing noise floor in 900, just like there is going to be an increasing noise floor everywhere else. Use horiztontal as much as possible as most 900 noise we've seen is vertical. YMMV

Hint for the day: a PacWireless 9dB yagi makes a heck of a 120* horizontal sector for ten percent of the usual price. Attach one to a RB112 with an SR9 and you just built a REALLY cheap micro-pop that can handle 20-25 customers.

George

Don't trust the noise floor readings. MT 2.9 is not optimzed for an SR9 and you have to take the noise floor as an "indication" rather than the abolute truth.

We have a lot of SR9s in the field, they work well on 112s and 532s. We NEVER, EVER install one with less than a -75 signal, stronger than that if the area is noisy.

Also, you are unlikely to be able to use more than the 5MHz channel. Its remotely possible, but unlikely as 10MHz or 20MHz will see too much noise.

How high you can mount depends where you are and what your noise environment is.

I have one Vpol omni at 300' that works fine, we have a 9dB horizontal yagi at 20' that causes intermittent problems. Depends completely where we are.

You should mount as low as possible, and sectorize as much as possible. Plan your network for an increasing noise floor in 900, just like there is going to be an increasing noise floor everywhere else. Use horiztontal as much as possible as most 900 noise we've seen is vertical. YMMV

Hint for the day: a PacWireless 9dB yagi makes a heck of a 120* horizontal sector for ten percent of the usual price. Attach one to a RB112 with an SR9 and you just built a REALLY cheap micro-pop that can handle 20-25 customers.

George

George, Very well said!! As I read your post, and thought about my experience I related exactly to what you said. 900Mhz is a whole different set of rules that can often be confusing. Would like to chat with you about this subject! Thanks for the tip of the day as well! Whats the next tip!!....

why are the horizontal sectors 2-4x the price of a vertical sector ? Are they that much harder to make ? Is it worth the price difference ?

why are the horizontal sectors 2-4x the price of a vertical sector ? Are they that much harder to make ? Is it worth the price difference ?

Less energy going into earth and space, more goes toward horizon. But ghmorris makes an excellent suggestion; if you can't spare US $400-$900 for an "official" 900mhz horizontal sector, start with something affordable. Just make sure that if/when you do change to something with a tighter vertical pattern you don't end up degrading customers who happened to fall into the looser pattern of the less expensive antenna...

why are the horizontal sectors 2-4x the price of a vertical sector ? Are they that much harder to make ? Is it worth the price difference ?

Less energy going into earth and space, more goes toward horizon. But ghmorris makes an excellent suggestion; if you can't spare US $400-$900 for an "official" 900mhz horizontal sector, start with something affordable. Just make sure that if/when you do change to something with a tighter vertical pattern you don't end up degrading customers who happened to fall into the looser pattern of the less expensive antenna...

I understand that the horizontal ones are better, just can't figure out why they cost 2-4x as much ... is it super high grade gold inside or something? Does the mfg process take twice as much time ? To me it seems like an artificial markup of price because they perform better.

I understand that the horizontal ones are better, just can't figure out why they cost 2-4x as much ... is it super high grade gold inside or something? Does the mfg process take twice as much time ? To me it seems like an artificial markup of price because they perform better.

One possible reason is that not as many are manufactured and sold (which is just as well; horizontal is a good refuge in the face of clueless, uncoordinated colleagues). I suppose it's also possible that there are fewer designs for these in the public domain, so maybe some licensing fees are being paid for designs? Sheer speculation.

What kind of obstacles will 900Mhz fail on? We have a site down in a valley that our 2.4GHz overshoots, we were thinking of putting two 12dB 900 MHz Roo2s and SR9s to create a point to point link of approx. 1½ miles, however from the tower to the valley there are perhaps 10-20 buildings in the way as well as foliage. The big problem is getting down to the valley - there isn't that much of a drop off, maybe 20-30ft, but it would mean we would have to aim the tower antenna down towards the ground slightly and most likely through some of those buildings too. Does this sound like it has any chance at all of working?

...
We NEVER, EVER install one with less than a -75 signal, stronger than that if the area is noisy.
...

We too have found that at least -75dB or better signal is key for stable bandwidth and latency.