ETH6 Low SNR on two channels only

[JohnW248]

You'll notice that tuner #3 doesn't get any CCI information even though other tuners on channel 367 get CCI data. Looks like a possible problem with tuner communications.

[tzr916]

Yes, I noticed that several times when the very low SNR happens.

I was out today and had to re-schedule the comcast visit for tomorrow..

Ceton is responding quite quickly and we're moving toward RMA options for my ETH6... unless my comcast visit is a success.

[tzr916]

Comcast tech was here. He replaced my splitter/amp setup with just one splitter (no amp), cut off and re-terminated all the coax connectors on my property and at the box at the end of the driveway, and ran his spectrum analyzer at 135000 frequency for several minutes.

I was able to demonstrate the issue easily by setting just a few tuners to channel 367. He was really puzzled that my tuners showed 29dB, while his analyzer said 37dB. So everything does point at my ETH as being the problem. I will be taking the out of warranty RMA offer by Ceton for a refurbished unit. Hope they send me a good one!

[JohnW248]

Comcast tech was here. He replaced my splitter/amp setup with just one splitter (no amp), cut off and re-terminated all the coax connectors on my property and at the box at the end of the driveway, and ran his spectrum analyzer at 135000 frequency for several minutes.

I was able to demonstrate the issue easily by setting just a few tuners to channel 367. He was really puzzled that my tuners showed 29dB, while his analyzer said 37dB. So everything does point at my ETH as being the problem. I will be taking the out of warranty RMA offer by Ceton for a refurbished unit. Hope they send me a good one!

With a new tuner you'll have to both pair the cableCARD and THEN run TV Set-up again since the new tuner has a new unique ID for Windows as well.

[tzr916]

Thank you Ceton. Got refurbished ETH6 late yesterday. Waited until today to install...

Put on network with static IP
Updated firmware to v15.1.13.152
Ran WMC DCA
Ran WMC Ceton Tuner Setup
Ran WMC Setup Tv Signal
Called Comcast to activate/pair
Finished, all channels working

New SNR numbers look perfect (right now):



Keeping my fingers crossed that this is resolved.

[tzr916]

So BAD NEWS...

Watched the Ceton summary page and within 10 minutes four of the six tuners had SNR lower than 32dB.

Rebooted Ceton ETH6, re-tuned all six to the same two channels, SNR looked good for the first minute or two then started dropping and in less than 10 minutes again three of the six tuners were down below 32dB.


Guess I will be buying some SD tuners soon.

[JohnW248]

Before you do anything else, you might want to try this test:

Disconnect everything from the cable drop to your house. Then with a single coax connect from the drop to the ETH and see what you get. What this will do is rule out any internal wiring issue or intrusion so its important that the only thing on the drop is your Ceton for the test. If the SNR still drops the problem is either the tuner or some outside intrusion that is beating that frequency.

[tzr916]

Before you do anything else, you might want to try this test:

Disconnect everything from the cable drop to your house. Then with a single coax connect from the drop to the ETH and see what you get. What this will do is rule out any internal wiring issue or intrusion so its important that the only thing on the drop is your Ceton for the test. If the SNR still drops the problem is either the tuner or some outside intrusion that is beating that frequency.

Might be on to something. Tested with no splitter/no internet modem. Only 50ft from POE to ETH. Manually tuned up the two channels and let it sit for about 20 minutes. Then stopped the manual and set WMC to record the two channels 5x at once. Both tests the SNR stayed between 36.7-37.9 dB.

Re-connected splitter ONLY, ran the same tests and the SNR remained good.

Re-connect cable modem to splitter, ran the same test and the SNR dropped to 33dB. Also tried another coax patch cable.

So it does seem that introducing the cable modem into the line is squashing the 135000 frequency

What can I possibly do here? Buy a new modem? What if the new modem does the same thing? Why did my system work fine for almost two years and suddenly started crapping?

[STC]

Tried new COAX leads with new plugs? It could be ingress coming in somewhere.

[tzr916]

Tried new COAX leads with new plugs? It could be ingress coming in somewhere.

New 50ft from POE
New splitter.
2x new patches from splitter to modem+ETH.

Tuners work fine with 50ft from POE to splitter patched to ETH.
Connect modem to that splitter and tuners go to heck.

[Paw Paw]

You mention 2 new patch cables to the modem and Eth. That sound like the modem is located close to the ETH. If it is easy to do try moving the modem away from the ETH.

You also asked earlier what could change. Most cable companies and clearing more and more bandwidth for high speed internet so your modem might be bonded to different channels (frequencies) that is was before. This might cause your problem.

[tzr916]

Yes the modem and ETH are physically less than 12 inches from each other (as they have been for two years).
The coax patch cables are about 6ft long (as they have been for two years).

I just tried putting a short "ground jumper" between the modem and the ETH, but that did not have any good results. I guess I can try moving them apart physically but don't have much hope for success.

[JohnW248]

Paw Paw might have the answer. If they are upgrading internet speeds like there did here, the high speeds require new modems for max speed and they bind several channels together to get the bandwidth. What you would need (and they should provide) is a band filter for the modem line and one for the video line. I don't know where the internet spectrum is but you might be getting sideband harmonics that are killing the SNR. If they can't supply the filter you can 1) file a complaint with the FCC, 2) cancel internet service and get DSL or ???

If this is a problem (and it reminds me of the FiOS problem a few years back where they had a sideband MoCA problem) they should provide a solution, you should also tell Ceton about your test with the modem disconnected and they might have a suggestion as well.

BTW if they are making these changes, its a good time to RENT the modem from the company and not buy since the new schemes are very specific as to what modems work with it for max speed and maybe the new frequencies are causing new back feeding from your modem. If its rented now, replace it, if not consider getting a modem from the cable company and see if that works.

[STC]

...sideband harmonics...

My wife gives me that. Usually in my right ear, sometimes in the left.

[tzr916]

Post #1 shows my modem bonding channels and frequencies. I've had this same 100M service for months, and had 50M service for more than a year before that. I have no idea if or when the bonding channels & frequencies every might have changed.

When the comcast guy was here, with his frequency analyzer set to 135000, I showed him the low SNR on my Tv screen thru WMC and the SNR was normal on his analyzer. He was puzzled.

PLUS, how in the world can some tuners in the same ETH be 10dB lower than others at the same instance in time? If something were "feeding back" and dampening that frequency then every tuner in my ETH should be 10dB lower. My experiments show otherwise. One example- five tuners (all set to channel 367 or 369), two tuners have very low SNR, three tuners are normal:



I've ordered an HDHR Prime and going to get a second cable card. Planning to setup and test these two channels on it.

[JohnW248]

PLUS, how in the world can some tuners in the same ETH be 10dB lower than others at the same instance in time? If something were "feeding back" and dampening that frequency then every tuner in my ETH should be 10dB lower. My experiments show otherwise. One example- five tuners (all set to channel 367 or 369), two tuners have very low SNR, three tuners are normal:

I've ordered an HDHR Prime and going to get a second cable card. Planning to setup and test these two channels on it.

Same problem that Capt Kirk had on the Enterprise---Shields. RF is funny stuff and if your modem is radiating a harmonic it could be shielding, placement, etc. Sometimes rotation of a few degrees might fix the problem. But try for a practical solution (try moving that modem as far as you can or put it in a Faraday Cage and see if that helps.

Once you find the solution, you'll know the problem and the explanation will be clear---until we can isolate all variables and make it work, our theories are no better that guesses.

I have lots of RF stories including a dead unplugged garage door opening receiver that was creating a null point around the new garage door operer. In that case something had gone bad in the non used receiver---something might have gone bad in your modem (electrolytic caps are notorious) that might cause the problem and not yet impact its performance. Again all guesses.

[tzr916]

...try moving that modem as far as you can...

I moved the modem from 12 inches to 6 feet, and no change in the problem.

What I also find interesting is the time it takes for the problem to creep in. After I moved the modem, I rebooted the ETH and ran some tests. The SNR looked perfect. I kept watching the numbers, and they stayed good for about 8-10 minutes. Then the SNR started dropping gradually until it got down to 32dB.

..a band filter for the modem line and one for the video line..

http://www.wiredathome.com/cables-and-c ... ss-filter/ Would this filter work? Where do I put it? on the modem patch cable or tuner patch cable?

[JohnW248]

That might do the trick, put it on the input to the tuner which should block any back intrusion coming from the cable modem. For fun you can try wrapping the cable modem in alum foil (poor man's Faraday Cage) and see who long it takes. But I think you've found your culprit, now its a matter of getting it to perform properly. I wouldn't discount that something in the modem might have gone bad (filters are usually either a RC (resistor/capacitor) or LC (coil capacitor) and if either value changes then it impacts how the filter work. Generally resistors go up in value over time and capacitors down or open. If the modem is properly designed I'd suspect that a cap is open or there is a cold solder joint somewhere. With current construction with wave soldering and surface mount components, repair and rework is very costly and most companies just replace and replace until it works. If you haven't swapped out the modem, you should try that.

[Paw Paw]

Since relocating the modem 6 feet seemed to decrease the rate at which your S/N drops I would be leaning to radiated RF not back feed up the cable. The foil trick that JohnW suggested would be my next test. Just be sure the modem has sufficient cooling. If you have some longer cables or butt splices so you can extend the modem further and time the deration of the S/N that might prove interesting too. If either test displays improvement in your problem the I think a new modem is the solution. However, at least with Comcast, you can pick up a modem at their store front test it and then return it and it would probably be one month's rent or less. That would absolutely prove where the problem is originating.

[tzr916]

Since relocating the modem 6 feet seemed to decrease the rate at which your S/N drops I would be leaning to radiated RF not back feed up the cable....

But it did not decrease the rate. It behaved the same as when the modem was 12 inches away. Moving the modem away from the tuner had no affect from what I have observed. I was simply making a note of how strange it is that the SNR looks ok for several minutes then starts to drop.

That might do the trick, put it on the input to the tuner which should block any back intrusion coming from the cable modem. For fun you can try wrapping the cable modem in alum foil (poor man's Faraday Cage) and see who long it takes. But I think you've found your culprit, now its a matter of getting it to perform properly. I wouldn't discount that something in the modem might have gone bad (filters are usually either a RC (resistor/capacitor) or LC (coil capacitor) and if either value changes then it impacts how the filter work. Generally resistors go up in value over time and capacitors down or open. If the modem is properly designed I'd suspect that a cap is open or there is a cold solder joint somewhere. With current construction with wave soldering and surface mount components, repair and rework is very costly and most companies just replace and replace until it works. If you haven't swapped out the modem, you should try that.

I'll try the filter before considering replacing the modem. Leaning much more toward an inline coax issue rather than radiant RF.

Also have an SD HDHR Prime coming and very curious to see if it will have the same problem...