how-to block ads
|
DaveL
join:2005-11-12
Urbana, IL
| Slowly dropping signal For the longest time (years) I had a signal strength of 82-83.
For the last 3 years the signal is slowly dropping.
It was in the 70s for a time and steadily dropping a little here and there. I am now at 61 for the last 2 weeks. Today is a clear day (no obstructions). I know that when it reaches 58 it will not function so it's time to get busy.
Not afraid to test and work on it. Been there as so many of us have.
Is this a transmitter thing?
What can I try?
Thanks
Dave
--
7000s, 99W, 1370, 77sig, pro plan
"Notice how clear the skies were after 9/11 due to lack of con trails?" | |
|  
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
| Re: Slowly dropping signal Two most common reasons: (a) branches creeping into line of sight, (b) drooping dish. Since you seem convinced (a) is not the cause, I suggest the push-pull test. Use the site search engine, it's been described here MANY times
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| | 
CM
Left Hip On Cattle
join:2002-04-15
Montana
2 edits | Re: Slowly dropping signal said by grohgreg  I suggest the push-pull test. Use the site search engine, it's been described here MANY times
Greg,
Did I 'invent' the push-pull test?
You were a long-time advisor when I arrived on this forum, and you are (apparently) still active. So, you 'were around' during my entire Dway Period.
I used to warp the dish to see if sig strength improved, and the idea came to me as an original thought. I began to suggest it to others who had similar issues.
Now I see the method is a 'searchable troubleshooting aid' with a name, and everything.
Am I the inventor, or just donning a legacy I don't deserve?
(I only stopped in today to see if people are still troubled by the same old gremlins...)
CM
--
Earthlink SRS (DW4000) | G16/970 | 4.2.1.10 beta3 | XP Pro SP2
ICS using 'phoneline' NICs with 1 client (XP Home SP2)
Norton Internet Security on host, XP SP2 ICF on client | |
| | |
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
| Re: Slowly dropping signal Dunno Charles. Quick search found 30 mentions on this site alone, dating back only two years. Been doing it for many more than that myself. But you're perfectly entitled to bask. I'm sure you recommended it to a good share of folks during your Hughes days as well.
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| DaveL
join:2005-11-12
Urbana, IL | Will give the ole push pull thing a try this weekend.
If not that, then what's next? | |
|
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY | One step at a time
//greg// | |
| laserfan
join:2005-01-14
Blanco, TX
| said by DaveL :For the last 3 years the signal is slowly dropping. Maybe an off-the-wall question, but The Chief would know: is there ANY CHANCE that misalignment over time can occur because of either satellite drift, or changes in the Earth's orbit?
I think for example the sun is setting higher in the West this year than it used to... 
--
HN7000S v5.8.0.49 Home Plan, G13/H1@127W, 1110MHz 33Msps, SigTyp 72, 10-20clients & 6+OSes | |
| | DaveL
join:2005-11-12
Urbana, IL
1 edit | Re: Slowly dropping signal I know you mean well but I got a chuckle out of the "setting higher". Kind of contradictory.
It's all Washington's fault anyway. Now they are messing with the universe.
--
7000s, 99W, 1370, 77sig, pro plan
"Notice how clear the skies were after 9/11 due to lack of con trails?" | |
|
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
2 edits | no.. well satellite drift maybe. But if that were the case, you'd already be hearing the screams of complaint from several hundred thousand other people. But "earth shift"? Not in my lifetime. Those satellites are essentiallyheld in place by earth's gravitational field. If/when we move - they move.
//greg// | |
| | laserfan
join:2005-01-14
Blanco, TX
1 edit | Re: Slowly dropping signal said by grohgreg :..."earth shift"? Not in my lifetime. Those satellites are essentiallyheld in place by earth's gravitational field. If/when we move - they move. Good point!
It is gratifying to realize too that even after our sun burns itself out in a few billion years, our satellite will still be there and our internet service will continue to function!
Thanks for indulging me.
@DaveL don't be surprised if Washington's answer to Global Warming will be to start a war with North Korea--nuclear winter becomes A Good Thing then, y'know... 
--
HN7000S v5.8.0.49 Home Plan, G13/H1@127W, 1110MHz 33Msps, SigTyp 72, 10-20clients & 6+OSes | |
| ddysart1
join:2009-07-10
Paola, KS | What is your dish mounted on? Perhaps the house or pole is gradually settling.
I had that problem 3 years ago when we had some foundation work done on the house. | |
| horse7
join:2008-09-10
Staatsburg, NY
| Not to hijack the thread, but I have observed the same sort of thing.
In the Spring the signal finally dropped to a value that was unusable (DW7000 signal 40 or less won't work). The signal was probably dropping before that, but I wasn't watching-- at installation the signal wasn't very good compared to what everyone else seems to have, maybe mid-50s 3-4 years ago.
Went after the trees first-- original issue seemed to show up when the leaves came out (and, coincidentally, when the temps were higher). No help.
Next azimuth/elevation. Tiny help (I only have one person so did this the hard way, not pushpull). Dropping elevation about a degree helped by a point or three, with the odd effect that the "lows" were lower while the "highs" were only a bit higher (30-43 after, 38-42 before, but the signal is usable for more of the day, perhaps 5hr vs. 3hr.).
Replaced LNB. No help.
Replaced outside cabling and used Stuf(r). A little better, about 46 is the high now, and sometimes the signal us usable until 10AM or 11AM. (yes, my time here for today is about over!)
I now have a replacement trans/LNB asm, but not installed yet. I usually make one change and observe for a while, to see if local conditions (like the horrid weather) are causing what I am observing vs. the change I made.
If the signal is very marginal but usable, I have noticed that increased use will degrade the signal more. This strengthens my impression that the problem is partly thermal and not, say, water ingress in the cables or atmospheric effects.
Is it possible that the issue is with the DW7000 modem? This modem is pretty old... soon the only original parts will be the physical dish, the inside cables (all exposed and undamaged), the outside dual barrel connection/ground/arrestor, and the modem.
I am in the process of trying to install cable (this is a long story in and of itself!) but can't until the ground dries up. Thus I am reduced to fixing the current infrastructure, mostly by throwing parts at the problem.
Thanks.
Mike | |
| |
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
1 edit | Re: Slowly dropping signal Assuming that you know what you're buying, a new TRIA and cable will help I'm sure. But make sure that approved and compatible connectors (to include the ground block) are part of the job.
Without the proper equipment though, pointing is at best a haphazard venture. I strongly recommend borrowing (or investing in) an OPI. If you can't do that, keep in mind that the modem takes a while to digest the fact that you're out there moving the dish. If you don't wait long enough, you're moving the dish again before the modem tells you what it saw in the last position.
And don't neglect ACP. Once you peak SQF, the job is only half done. At that point you peak ACP, then go back and see if there's any more ACP to be squeezed out of a final fine tuning of the Az and El
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| horse7
join:2008-09-10
Staatsburg, NY
| The TRIA is nearly the same rev, and the cables are CommScope 3GHz w/T&B (but plenum, which is good enough for this purpose for a few months-- go direct burial if they are needed for longer). The LNB replacement was a bit questionable used unit but made no difference good or bad.
I don't have a problem with buying an OPI except that I hope not to be on satellite much longer-- the work I do (requiring x-server connects, VPNs, proxy servers that don't seem to like the NOC for some reason, and using problematic tools like LNotes) is much more difficult with the high latency sat link. Unfortunately the ground is saturated-- "underwater" is a better term-- and has been for months making cable (actually UTP) burial infeasible.
One can argue that a used Birdog or some such can be purchased, used for as long as necessary (months? years?), and resold for not much loss I guess. Then I would have essentially all the tools necessary for debug.
My question about the modem has been answered by implication-- it sounds like the modem is an unlikely source of failure so I should focus on the other components. | |
| |
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
1 edit | Re: Slowly dropping signal said by horse7 :Then I would have essentially all the tools necessary for debug. No, the BirdDog and like meters cannot replicate the ACP test - which is actually more important than the received signal level (SQF). The OPI does both. But if you have a laptop, you can optimize both by using the modem user interface. That's basically all the OPI does anyway - is just remote real time numbers from the modem. My problem with laptops (besides the expense) is the display - and hence the numbers sent by the modem - are difficult to see at best, impossible on a bright day.
Your 3GHz plenum cable, I hope the center conductor is not copper coated steel (CCS). To get proper voltage/amperage to the transmitter, bare copper (BC) is mandatory.
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| horse7
join:2008-09-10
Staatsburg, NY
| The 50 feet of outside cables were swapped with CommScope quad shield CCS mostly because the replacements were prefab and copper can get kinked in the foam and cause non-obvious signal problems. I had not bought from this supplier before... also, my main purpose is debugging the problem. The point about copper and the power supply is a valid one though.
The TRIA was just replaced with a used / working unit, with no advantageous or deleterious effect.
I'm down to the dual barrel / arrestor, the inside cables, the sat modem, and the power supply. And maybe the black helicopter between my dish and 93West... | |
| | 
One more too
@direcway.com
| Re: Slowly dropping signal said by horse7 :I'm down to the dual barrel / arrestor, the inside cables, the sat modem, and the power supply. And maybe the black helicopter between my dish and 93West... In describing your issues, until now you had not mentioned that you are on 93 W. You may well have some equipment issues on your end, but you should also know that, for the last year and a bit more, 93 W has become the hell hole of the Hughes system. Problems started with a major power failure on the satellite. Since then, Hughes has been unable to get any kind of consistent performance for 93 W users. One of the things they've done is to switch people to different transponders as they tried to snatch up any transponders on the satellite that were still working. A few months ago, the moved another satellite in to replace the one that had been at 93 W, and that satellite is not in great shape either. Since all of this started, others have reported significant signal level drops, probably caused by the changes in transponders. Others have reported consistently very slow speeds because the transponders being used don't have the same capacity as the original ones. In any case, problems for 93 W users have been so bad and performance has been so inconsistent that anyone who complains at an even moderately loud level gets offered a no cost move to a new 9000 system on Spaceway.
Again, you may have equipment issues on your end, but, even if you resolve them, you will still most likely have to deal with the issues related to a very troubled satellite. Thus, your best move might be to try to get reassigned to another Ku satellite or to get Hughes to offer you the move to Spaceway. | |
| | | horse7
join:2008-09-10
Staatsburg, NY
| Re: Slowly dropping signal 93West...
Yes, I was there when the satellite had its failure.
The performance never did really recover to what it was previous to the incident, although I figured it was due to load. There weren't any consistent outages like I have now, although the system was more sensitive to weather conditions and had higher latency.
Now, having thrown hundreds of dollars of parts at the problem, I am reasonably confident that the issue is not at my end. It *could* be, but really the remaining parts are ones that don't generally fail (cables mostly indoors and the sat modem). Essentially I have been able to squeeze out a little more margin above the noise floor but have not found anything that would be described as "broken"-- now I have 2 TRIAs and an extra LNB, a bunch of old RG-6, plus firewood from cutting down trees... and more knowledge of how things work which is always useful.
As for satellite performance, it must be obvious at the NOC which receivers are consistently losing link. There is a big difference between idle receivers and receivers falling off the network, although the latter happens due to atmospheric conditions it won't be consistent every single day (and if one were truly competent, the correlation could be made to surface weather conditions).
I am unimpressed with the Hughesnet business model-- my case is a perfect instance where they should be proactive in trying to ensure good quality service based on observed link performance. Instead they are simply going to react when I contact them...
| |
| |
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
2 edits | Well, I'm glad at least that you acknowledge the potential for problems with CCS. But the fact that it's new can also present a whole new problem. I know all this stuff is basic 75 ohms hardware, but that doesn't tell the whole story. It's important that
-all cable segments be the same cable type,
-all connectors are frequency swept to match the cable
-the cable ground block is also swept to match cable and connectors.
Failing this renders the IF signal vulnerable to distortion resulting from an impedance mismatch.
For example; let's say your new CommScope is frequency swept to 3GHz. That means if your cable block - and/or cables from there to the modem - and/or coaxial connectors - are less than 3GHz rated, you're "strangling" the signal. Hughes minimum spec is 2.2GHz twin-shield hardware on BC cable between the modem and the dish. But if you exceed minimum in one segment, you must match it in all.
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| | | horse7
join:2008-09-10
Staatsburg, NY
| Re: Slowly dropping signal My original intent with the cabling was to rule out the possibility that the cables were the issue-- and knowing of the potential problems with copper [kinking from abuse by vendor fab, and not being obvious], I preferred to have CCS. Since the cables don't seem to be the issue, the longer term solution is to go with direct burial quad shield [although you state dual or tri shield should meet spec so those are options] copper swept to 3GHz multiple shield since the RG-6 cable pair will be next to each other, or possibly siamese. The test cables being outdoors and PVC plenum rated won't last too long anyway.
The original cable was Perfect Vision 18AWG copper E232510 3GHz swept siamese... I think it is quad shield.
Connectors are all snap and seal type, I've been using Thomas and Betts but I don't know what the originals are... they look more like PPCs from what I recall.
I get really ambitious I'd go to 3GHz RG-11... or maybe Heliax.... but the satellite latency is a huge issue with the work that I do (not helped if 93West is really the root of the problem). | |
| | | |
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
3 edits | Re: Slowly dropping signal said by horse7 :My original intent with the cabling was to rule out the possibility that the cables were the issue-- Yeah, but in swapping out only part of the cable path, you could have introduced a whole new problem that can skew the results. I can't personally approve of what you've done so far, simply because the CCS cable segments represent a red herring. Your perception of "problems" with BC don't hold water, unless the guy pulling the cable is a total incompetent.
T&B make a wide range of connectors. Just because they're Snap&Seal doesn't mean they're frequency/impedance compatible with your cable path.
RG11/heliax are overkill - money wasted - unless your dish is more than 276 cable feet from the modem. And neither will have ANY impact on latency whatsoever.
//greg//
--
HN7000S/98cm Prodelin/2w Osiris/ProPlus - G16/1250H/Germantown - NAT 66.82.187.152/Gateway 66.82.25.10/DNS 66.82.4.12 and 66.82.4.8 - Firefox 3 - AV/Firewalled by NIS2009 | |
| horse7
join:2008-09-10
Staatsburg, NY
| CCS vs. copper...
CCS, 18AWG (CommScope) at 0.0163 ohm/ft,
Copper, 18AWG, at 0.0075 ohm/ft.
For 100', at 2A draw, the drop is
CCS 3.3v drop
Cu 1.5v drop
This ignores non-linear effects and presumes that only the center conductor is used for current carrying (not the shield). In my case there is 50' of cable to and from the TRIA for a total of 100', with perhaps another 60-70' indoors (round trip 120-140) of original copper.
I don't remember the brick voltage output since I am not colocated with the satellite modem at the moment (I called Hughesnet Level 1 support, with the net result after an hour that now I have dropped from 40% availability to 1% availability.... really, they need to train level1 support better, there is only so much that can be done from a script)-- anyway, I sort of recall the voltage is 30-40v, which implies that 100' (round trip) of CCS adds 4-8% additional drop.
This probably affects transmit much more than receive, receive maybe more in noise margin reduction. I don't see this as being sufficient to make or break the link quality if the signal were adequate though-- partly this is because the original installer would go up to 200' outdoors with the cable he was using, 4x my length, and that apparently was functional (with copper).
If I ever get the blasted thing working again it would be interesting to experiment with swapping cable sections and observe the results. At the moment further experimentation is pointless.
Yes, I know, the latency is mostly geosync and unless Einstein was wrong that isn't going to change.... | |
|
grohgreg
Dunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
3 edits | Horse7: You're correct that Cu is more important on the TX than on the RX. But it's important that both runs present identical potential to the modem. Plus, you never know when it may be come necessary to swap Tx and Rx cables around.
Do you understand skin-effect? It's a problem with CCS. With CU, there's no "skin".
//greg// | |
| | | |
|
