Voltage backfeeding from CPE

[Log In to View]

I mostly handle escalation calls and I have to say about 75 percent of our repeat calls for intermittent hsd is from voltage backfeeding from CPE.  Either 2 prong outlets or open ground outlets.  To fix this problem we had to test which lines were backfeeding.  Then place a ground block on that line BEFORE it went to the output of the splitter. The voltage was causing resistance in the splitter which then caused the HSD to keep cutting out. Alot of the new Tvs are doing this too 50-70V.  30V seems to be the limit that starts the problem.  After we grounded it this way I have never heard back from the 100s of subs again. 
Has anybody else ran into this problem?  I think it's a pretty common problem that everybody is running into. You have to think though how many of our installers and service techs are actually checking for voltage though?   

[Log In to View]

I have seen this happen more so with the actual computer that is hooked up to the modem causing an issue resulting in constant resets. Even had a router reset the modem whenever data was pushed through it.

TVs however I havent heard of or ran into yet.
Is it random cut outs or is it unusable in any way?

[Log In to View]

It causes the modem to completely lose sync for hours on end even days.  Then it will come up briefly, it's totally random.  Signal will show 57TX -10Rf on modem, even when you unhook from modem and plug into sig meter until the resistance is released long enough then it will jump back to a 45TX, +2Rf.   Some houses especially older ones you may have 4 groundblocks coming off a 4way splitter all using one ground wire stripped back far enough for them all to fit on.  It just HAS to be before the output of splitter.  Apartment buildings was the first places we started noticing it. They backfed to the MDU which wasnt grounded and was shutting everybody off.  Just hooking a groundwire to a splitter wasnt working.

[Log In to View]

This is news to me!!

[Log In to View]

I have seen this happen more so with the actual computer that is hooked up to the modem causing an issue resulting in constant resets. Even had a router reset the modem whenever data was pushed through it.

TVs however I havent heard of or ran into yet.
Is it random cut outs or is it unusable in any way?

I am seeing this more and more though I am not sure its the TV, I sometimes wonder if its our boxes, yesterday 2 of 5 outlets running over 40 V, but here is the weird thing, with a VOM:


Black lead on shield, red on center conductor no voltage, black lead on ground block 40 V, at this point I can put the red on the shield or center conductor and I am getting 40+V.


I knew it had voltage because I got that lovely tingly/poking feeling every time the back of my hand grazed the fitting. Other cases I have been able to get voltage [up to 93V AC so far yay me] using the black on shield, red on stinger. So I am guessing when you have to touch the ground block its a bad ground, and when you can do it on the cable alone its a bad neutral? anyone know more about this?

[Log In to View]

GBs before the splitter isnt really a cure....yes it might solve your issue then, but as the customer sees it, unless it IMPACTS THEM...most dont care.....around here we disconnect the device/line...and tell them they NEED to replace the device, or in the case of a bad outlet, call an electrician....then they call us back (direct or 1800) and we hook it back up....

my boss got me some ac ground loop filters...screw on type....going to see if it helps any......most times, when we find it (maint) the customer isnt home......and again, if it isnt causing them a noticable problem, they dont care.....

also NO normal service tech is checking voltage on even the first repeat......i have to beg and plead for them to take out the VOM , if they even have one.....

[Log In to View]

GBs before the splitter isnt really a cure....yes it might solve your issue then, but as the customer sees it, unless it IMPACTS THEM...most dont care.....around here we disconnect the device/line...and tell them they NEED to replace the device, or in the case of a bad outlet, call an electrician....then they call us back (direct or 1800) and we hook it back up....

my boss got me some ac ground loop filters...screw on type....going to see if it helps any......most times, when we find it (maint) the customer isnt home......and again, if it isnt causing them a noticable problem, they dont care.....

also NO normal service tech is checking voltage on even the first repeat......i have to beg and plead for them to take out the VOM , if they even have one.....

I barely ever bust out my VOM unless i get shocked touching things.
Had a PS3 feedback 120v (like SPARKBANG couple amps) to the coax ground via HDMI shielding-TVfame-RF output. a/c outlets there did not have ground prongs so it went out the only way it could.

[Log In to View]

GBs before the splitter isnt really a cure....yes it might solve your issue then, but as the customer sees it, unless it IMPACTS THEM...most dont care.....around here we disconnect the device/line...and tell them they NEED to replace the device, or in the case of a bad outlet, call an electrician....then they call us back (direct or 1800) and we hook it back up....

Generally i tell them I recommend they get an electrician to look at it, and don't bother with much form there, However, if a ground block fixes it, wouldn't grounding the splitter/amp do the same thing then?

[Log In to View]

If you ground from the splitter, it still effects the other outlets.  You place the GB between the splitter at the device on the leg where the voltage comes from and the voltage never makes it to another outlet.

[Log In to View]

thanks for feedback guys.  Generally when I get to a sub Im the 3rd or 5th guy there so I HAVE to try to make it work. We inform them that they need an electrician or unhook the tv from the cable.  The problem with this is some areas we work in they are barely getting by.  I'm all about taking care of customers and corporate would rather us do that to keep the sub rather than them drop us.  Especially after the way our Video subs are dropping. 
 
 

[Log In to View]

  I'm all about taking care of customers and corporate would rather us do that to keep the sub rather than them drop us.  Especially after the way our Video subs are dropping. 
 
 

Great attitude! I agree with that 100%
 As much as they can drive us insane- Subs are our money. Its a scary time lately. Im sure we will all pull it out, we always have. Things are changing though.

[Log In to View]

too bad market health scores from comcast dont have a low/mid/high income level toggle....

[Log In to View]

If you ground from the splitter, it still effects the other outlets.  You place the GB between the splitter at the device on the leg where the voltage comes from and the voltage never makes it to another outlet.

I fail to see how though. The splitter is essentially multiple  F-81's with a shared metal casing, a GB is essentially one F-81 with a metal casing, connecting either metal casing to a ground wire should work the same except for the splitter grounds 2 - 8 outlets instead of one unless the ground is not the least path of resistance..

[Log In to View]

I've run into this same issue, but have never correlated it to CPE voltage feed. It manifests as high TX particularly, eventually knocking the modem offline. Can pull the line at the modem and run a DOCSIS test, will see the high TX, if you even get lock. Work your way back toward the ground block, and you'll see the high TX up to a point (assuming more than one splitter between GB and modem -- I mean, there's NEVER more than one splitter!!!) until suddenly it tests good. Put it all back together and now your TX is great at the modem.

I had a supervisor once explain it as a capacitance building up in the splitter, usually due to poor/no bonding. When you pull a cable from that splitter and replace, it dissipates, but start building again and the time bomb starts ticking. But I've also run into the issue in a case there the bond appeared fine. This CPE voltage is an interesting new avenue to pursue.

Previously I had been chalking it up to CPD since usually some very old, oxidized fittings were in place. Also because to me, as a service tech, CPD is magical, mystical stuff since I can't test for it and was never really taught too much about it. Most of what I know about CPD I've inferred myself, probably incorrectly. My typical attack was to change everything outside, all fittings, all splitter, GB, ground wire and ground clamp.

[Log In to View]

I've run into this same issue, but have never correlated it to CPE voltage feed. It manifests as high TX particularly, eventually knocking the modem offline. Can pull the line at the modem and run a DOCSIS test, will see the high TX, if you even get lock. Work your way back toward the ground block, and you'll see the high TX up to a point (assuming more than one splitter between GB and modem -- I mean, there's NEVER more than one splitter!!!) until suddenly it tests good. Put it all back together and now your TX is great at the modem.

I had a supervisor once explain it as a capacitance building up in the splitter, usually due to poor/no bonding. When you pull a cable from that splitter and replace, it dissipates, but start building again and the time bomb starts ticking. But I've also run into the issue in a case there the bond appeared fine. This CPE voltage is an interesting new avenue to pursue.

Previously I had been chalking it up to CPD since usually some very old, oxidized fittings were in place. Also because to me, as a service tech, CPD is magical, mystical stuff since I can't test for it and was never really taught too much about it. Most of what I know about CPD I've inferred myself, probably incorrectly. My typical attack was to change everything outside, all fittings, all splitter, GB, ground wire and ground clamp.

I see this a lot as resistance mismatch, when you disconnect the corroded fitting etc the circuit opens back up 'resetting' the impedence.  I have also seen a weird thing, probably reflections, where if you test the line by itself, it's AOK, but then you add any passive device and you get weird MER issues.

Never thought about bonding a problem outlet by itself before.  Does sound band-aidish, like putting the HPF behind the elderly couple's 1970 TV chock full of ingress that they refuse to replace.