Hi,
I placed the antennas in the city at different physical locations 1km away while in the village I placed both antennas on the same tube and the antennas are very close to only 10-15cm distance between them. The capacity of individual and both breaths is about 200mb + 200mb, but when I turn on both wireless, their capacity is halved and they only have 100 + 100mb.
whether it is necessary to place the breathing in the village at a greater distance so as not to interfere with each other and whether 7-10 meters of physical distance would help?


Links have great signal
First -50db
Second --47db
I think the problem is the same location on one side.i think i must split this antenas minimum 7-10m for best frensel zone.What do you think?
Thanks

For starters:
- On both of those Dyna Dish systems, use frequencies as far apart as possible.

You might try something like this:
On the lowest frequency , try eeeC
On the highest frequency , try Ceee

fyi - that pipe sticking out through the roof looks like a water leak waiting to happen.

First dish have 5920 Ceee
Second have 5615 eCee
I try on second lowest 5220 ceee but its the same problem .when i turn on second wireless i have half capacity on both. i thinki must separete it

Now, when i separate Dish all works fine

I was really hoping I could find an authoritative answer for referencing how much distance you need between dishes. The ITU has this recommendation in ITU-R M.2244 where it says, "Vertical isolation can be computed by the following equation (isolation in dB) = 28+40(log(distance in meters/wavelength in meters))" I want to be clear that this may not be a perfect application to your situation because it only applies to vertical dipoles (e.g. omnidirectional whip antennas) where the distance is greater than 10 times the wavelength. In a 2.655ghz system, in order to get 30dB of isolation, you should have 0.13 meters of vertical separation --between whip antennas of course--, but again with dishes the side lobes may cause issues, so more separation may be necessary. For a more complete answer in your application, check the referenced ITU document. In my own practice, I've generally shortened this document down to 10 times the wavelength is the recommended minimum separation.

Edit: for a base isolation(dB) number to try to gain, you ultimately need to treat these two antennas like a system that exists in a Friis equation (https://www.pasternack.com/t-calculator-friis.aspx) and try to get as much separation so that the dishes do not sense each other. The dynadish 5 model has a receive sensitivity of -72dBm and a US legal eirp of 36dBm meaning the transmitter legally should only be running at 11dBm. Ultimately now I need the radiation pattern to complete the equation, but this is not available. If the dynadish behaved like the RF Elements Ultradish 24,I can use the gain elements from the radiation pattern to determine the adequate separation (I'll guess from looking at the PDF and use -25dbi of gain in the vertical stacked orientation, leaving me with a remainder of 0dbi of gain from the dish). Using the 10*wavelength rule (.6 meters), I have a Friis equation result of -31 dbm (not good enough, too much signal), but to build a more reliable system, I need to add more separation. To get below the MCS9 threshold of -72dbm, I need to use 67 meters of separation. Thankfully, Mikrotik hardware has noise and interference filtering on the RF side and PHY frame filtering on the datalink side, so 67 meters isn't entirely necessary.

Using friis equation in this case is not exactly correct, because it doesn't take into account that both systems are using separate frequencies. So there's another thing to consider: filter in Tx which limits amount of stray power emited in adjacent frequency band. Rx filter doesn't do any good here as both systems operate in same (wide) frequency band and Rx pre-amplifier in cheap systems amplifies whole supported band, AGC triggers on highest signal level which isn't necessarily the signal we want to receive (which effectively reduces receiving sensitivity).
So in reality the necessary distance between antennae is smaller by quite some margin.

Which holds true if both systems operate on different frequencies. If they both use same frequency (of partially overlaping frequency bands), then friis equation it is.

The ITU recomendation mentioned in post by @qamtester is very handy when placing antennae for MIMO (multiple radio chains). Antennae used for same MIMO channel need separation and realistically there are two ways of achieving it: spatial separation described in mentioned post and polarization separation (dipole antennae oriented at 90° angle). With MIMO 2x2 one can get away with polarization separation alone, with higher MIMO ranks it is necessary to use spatial separation as well. And the rule of a few lambda (the more the better) applies.
The trick in such use is that both (all) antennae are in same mode (Tx or Rx) at any time and the huge Tx power of adjacent antenna saturating receiver doesn't happen.