Hi,

I have the following scenario where I need to beam 2Gbit to another location. Distance 500m.
I already had setup a pair of LHG60's which worked very well. I now want to attach another pair to double the bandwidth.

I've set the two AP's on different frequencies.

The existing AP was mounted without radome. For the new AP I thought it was a good idea to try out the radome. To protect from snow and dirt.

The first (already existing link) is doing just fine and always steady. The second link which I just set up has a strong signal and good rssi, but the error rate is fluctuating a lot. When idle it flips from 5% to 24% etc. When used the error rate reaches above 50%. Signal stays strong.
The first link stays stable, so it isn't the interference I guess...

So, before I try some things; remove the radome or maybe re-mount the antenna somewhere else, I'd like your opinion what causes this. Do you have any idea where to start? Thanks!

So, I've removed the radome for now on both sides.
Connection is strong, but error rate still very high. It's a little bit lower then before, but that might be due to better aiming.
Signal = 95
rssi = -52
MCS=8 (stable)

error rate = between 0% and 30% while idle -> goes up to 60% when transferring data

I'm using 62640 freq now on the second link. The first link is still very stable at 58320 freq.

Turn off the other one and see if that changes anything.

Which one works perfect the higher mounted one?

Turn off the other one and see if that changes anything.

That's not so easy I'm afraid since it is in production. I'll take it in consideration. Maybe I can find a way. Thanks

Which one works perfect the higher mounted one?

Yes, the higher mounted one works fine.

Which one works perfect the higher mounted one?

Yes, the higher mounted one works fine.

It is possible that the lower one is too close to the roof, especially when it is not on the edge (can't see that on the photo).
You need to keep a clear zone (fresnel zone), larger than the diameter of the dish itself.

Which one works perfect the higher mounted one?

Yes, the higher mounted one works fine.

It is possible that the lower one is too close to the roof, especially when it is not on the edge (can't see that on the photo).
You need to keep a clear zone (fresnel zone), larger than the diameter of the dish itself.

The pole is 2 meters high. The lower dish must be at around 1.50m high. It's on the edge of the roof so I don't think that's in the way.
There is however a building near that can be in the fresnel zone, maybe.

If it is fresnel, would this explain what I'm seeing? -> High signal, good rssi, MCS8, but a very high error rate? I would really like to know if this can explain my readings, so I can make the right choice to relocate the lower antenna. Thanks again!

If it is fresnel, would this explain what I'm seeing? -> High signal, good rssi, MCS8, but a very high error rate? I would really like to know if this can explain my readings, so I can make the right choice to relocate the lower antenna. Thanks again!

No not really, that is why I did not mention that in my first reply. I would say it is more likely that the receiver is overloaded by the local transmission from the other link.
When it works fine when the other link is off, you know that is why it is. If it remains bad, you can look more towards relocating the second dish (e.g. a second pole some distance away but with similar height, or a higher pole so the lowest of the two is where the highest is now)

No not really, that is why I did not mention that in my first reply. I would say it is more likely that the receiver is overloaded by the local transmission from the other link.
When it works fine when the other link is off, you know that is why it is. If it remains bad, you can look more towards relocating the second dish (e.g. a second pole some distance away but with similar height, or a higher pole so the lowest of the two is where the highest is now)

Agreed.
I would think the upper antenna would then also suffer from the lower one, but there is no observable change on that working antenna. I'll try to relocate the lower antenna on another pole and report back.

So, I've relocated both antennas on both sides. They are now about 10 meters apart.
The signals do cross each other but there is no other way, as you can see in the image. There are trees which are not blocking the LOS, but only in this configuration as you can see.
Both antennas are now at the same height, about 2 meters high on separate poles on the roof (lower part of the image).

Now, the error-rate seems less, but is fluctuating a lot for the 2nd link. I had just had almost no errors for half an hour, but now the error-rate seems to be as much as before. So strange! I can't get a grip on this.

Have you tried turning off the good link to see if the transmission errors go away on the troubled link? Would quickly tell you if the other radio is causing it or not. If its still high I would try swapping the radio / dish combinations as it could be bad geometry on one of the dishes.... Lots of AB testing to isolate the issues. Can Also link them up in other configurations such as if currently A<->B and Y<->Z then connect A<->Z, B<->Z... etc to isolate the issue. Time consuming but if you want to find the answer then controlled tests are what you need.