Thank you all for the insight so far. I read your posts last night and came in early today. I inherited this install. I noticed some connectors that were not crimped correctly. I ended up replacing 6 connectors. Here is a list of the lengths.
I used a Snap Shot to get the lengths. http://t3innovation.com/snapshot
Here is the result of the test you told me to run. The N/A is -43 when done from another receiver. I did the test on all 24's and 25's and nothing varied more than 5 in either direction.
All the DECA's are still orange BUT everything is working and I have not had any drop out's since.
All Empty ports on the 8-ways have Terminators.
OK, you've got some work here.
You will have reduced networking when close to -45, so "Here's what I would do":
1) move the H25 245'-0", H21 330'-0", HR21 242'-0", and HR22 140'-0" or HR-21 138'-0" to a 4-way splitter. This will reduce the loss by 3 dB and keep the longest runs on the same splitter so the loss between them is the least.
2) move the HR22 140'-0" or HR-21 138'-0", H24 197'-0", H23 180'-0", H24 150'-0", H25 179'-0", H23 178'-0", and H25 175'-0" to the same 8-way splitter. The shorter lengths will help with the loss of the 8-way.
3) Connect the broadband DECA with a 2-way splitter to the line for the HR22 or 21 on the 4-way as this is the shortest run and the PI to the -16 power port.
4) get two of these: http://www.solidsign...sing-(std-9501)
Add them between the splitters and the SWiM-16 outputs and connect the OTA ports to each other. This will bridge the DECA between the two outputs instead of using the internal bridge of the -16 and decrease the loss by about 5 dB.
This should bring the losses down to less than 40 and get the network within specs.
After these changes, I'd check the LEDs and both test screens. The levels should be under -40 and the mesh "should be" around 240+.
If there still is a problem with the LEDs, the mesh screen should also show it with low numbers.
Before going into any further steps, please post the results of the two screens.