CABL Forum - Intermittent raised noise floor...

Hey guys, need a little help...

We've got the dreaded intermittent raised noise floor in one of my nodes. The floor will rise 10 dB, creating extremely low upstream SNR issues for all the modems in the node. Sometimes it will last a few hours. Sometimes it will last only a few minutes. It's hit or miss and it has nothing to do with wind, moisture, or temperature, as I've seen it occur and not occur in every one of those conditions. We thought we caught it Saturday night, but instead we were left a little puzzled...

GI 870 MHz equipment:

Went to the node and found the noise coming off port 3. Went to the first active (GI MB with internal 2way) and found the raised floor coming off ports 3 and 4. I stayed at the amp and a coworker went to the next amp in cascade on port 4 (trunk run to the amp) and found nothing. He then went to the first amp in cascade on port 3 and found nothing. The port 3 amp was fed by 4 taps however, so he backtracked from 10 tap to 14 tap to 17 tap to 20 tap to see if this issue was caused by a drop or bad port terminator. He found nothing, until he got to the 20 tap. At the input of the 20 tap, the cable was kinked in two different spots right were the expansion loop was located. He broke apart the connector at the 20 tap input as I was watching at the first active and the noise disappeared. We thought we finally found the cause of the issue. We cut the cable, straight-spliced it, and cut in a new tap at the location. I went back to the original MB to confirm and I was pissed to find the raised floor came back!

Angered at the situation, I took out the internal splitter and replaced it with a suitcase. I was sure the noise would disappear because we disconnected forward signal, but I was surprised that it was still there... on port 4 (the original trunk run)! My coworker immediately went to the port 4 amp... again, and checked for noise: nothing. I told him to loosen the seizure screw at the input of the amp just to make sure it wasn't coming from it. He did and the raised floor remained, but suddenly dropped a little. Instead of 10 dB up, it was now only 3-5 dB up. We knew we were on to something because resistance was obviously relieved. He re-tightened the seizure screw at his amp: no change (still the same 3-5 dB raised floor).

Now here's the weird part:

I decided to loosen the output seizure screw of my amp just to see if anything would happen. The second my screwdriver touched the screw, the noise came back up. Solid 10 dB raised floor! I took it off and the noise disappeared. Just to see why this happened, I did the same to port #3: no change. Port #2: no change. Input seizure screw: 10 dB raised floor.

Conclusion:

When I touch the port #4 seizure screw with the screwdriver the noise floor rises 10 dB. When I touch the input seizure screw the noise floor rises 10 dB. No change with any of the other ports.

Question:

Is this raised floor a product of a possible bad pin on both the input and output of the amp? Is this raised floor the product of a bad amp itself? Or... is the amp reflecting an issue we haven't found further down in cascade between the first active and the second active?

Thanks,

J_r0kk

INSTALLATION

CONSTRUCTION / TECH

OFFICE / OTHER

Intermittent raised noise floor...