Advertisment: Broadband via satellite
Advertisment: Worldwide satellite services from Ground Control Company

www.satsig.net

Satellite Internet Forum.

Welcome, Guest.        Forum rules.
      Home            Login            Register          
Pages: 1

Marginal Eb/No and steady 80% raw BER on Shiron

(Read 10022 times)
chriswlan
Member
★★
Offline



Posts: 27
Apr 5th, 2013 at 9:33am  
Can't remember seeing this before: if a Ku modem gets too close to the signal strength threshold, it just eventually drops out, no?

I have several Shiron iRG's in the same area that often report (via telnet) a steady lock, as well as a steady marginal Eb/No (even a strong one once), and a very steady raw BER at .8 or so!

Am I mistaken or would this point to a LNB saturated/desensed, by an out of band strong pulsed signal, for example a local ship radar -at about 20% lower in frequency?  And the modem would claim to remain locked through the short X band pulses?

Any hints appreciated

Christian
Back to top
 
 
IP Logged
 
USN - Retired
YaBB Moderator
★★★★★
Offline



Posts: 837
Kentucky (USA)
Reply #1 - Apr 5th, 2013 at 12:20pm  
At first glance, a "very steady" BER would rule out RFI. That's assuming you mean error rates as reported by the users. I'd expect something like pulsed interference to cause a definable error pattern. How close to each other are the reporting users? Do they operate off the same electrical grid?

That said, it's possible for RFI to affect the TX only. Minor interference on the uplink can cause excessive retransmissions. This slows down modem throughput, without necessarily affecting user BER or lock. Assuming you're the hub, what do you see coming from the affected users?

//greg//
Back to top
 

USN (Ret)
 
IP Logged
 
Eric Johnston
Senior Member
★★★
Offline



Posts: 2109
Reply #2 - Apr 5th, 2013 at 6:42pm  
If you see rock steady, unchanging, values of Eb/No and BER at remote sites I would suspect that the measurements are faulty and that there is some software problem.  Normally such values vary slightly with time, and certainly so over a period of 24 hours. Plotting the graphs over 24 hours can pick up pointing errors.

Decrease and increase the outlink carrier from the hub by 0.5 dB and see what changes, in any, that the remotes report.

Were all these sites installed at a time when there was no carrier present on the opposite polarisation ?  Is the incidence of the problem at these sites associated with new cross-pol carrier being activated, perhaps temporarly ?  Possibly these sites are all badly set up for polarisation ?

Is the incidence of the problem at these sites associated with new carrier being activated on adjacent satellite, perhaps temporarly ?  Possibly these sites are all badly pointed ?

Best regards, Eric.
Back to top
 
 
IP Logged
 
chriswlan
Member
★★
Offline



Posts: 27
Reply #3 - Apr 6th, 2013 at 1:54am  

SOMETHING NEW TO CHEW ON:

Tks for msg, Greg: no I'm not the hub just another user, 10 km from the other "troubled" sites.

Just took the time to mess with my own dish, that has stable strong Eb of up to 7, and perhaps 2/3 of a dB less during the day   -sun noise I assume.

My solid 4ft dish has been sitting on a rotting palette for a couple years, so I just carefully levered a corner, AND NO WAY CAN I EVER GET CLOSE TO TELNET SAYING APPROX 80% BER, nor approx 50% etc, as at the other 2 sites.

Here, I NEVER NEVER see anything above 5% BER. and if I am close to the "threshold", FINELY tuning the screwdriver lever so to speak, I can only get about 4.2 Eb/No, locked with perhaps 1-3 % raw BER, and once in a while it will flip to a plain  immediate unlocked, as well as zero Eb, as well as 100% BER.

A very few percent, or a full 100%, nothing in between at this location. Whatever the sampling timing of the Shiron measurements might be...

And then, a moment later it will flip back to immediate locked, AND weak signal, AND a few percent BER.

On the other hand, at the 2 sick locations, the VSATs will remain locked and lo Eb and super high BER, passing no data, for what looks like hours, or once in a while a slightly better signal and Ber, and we have Internet for a few moments or whatever. Usually better at nite...


All 3 modems had been rotated tru all 3 locations a few days ago, and the super high ber is definately geography related.... Coincidence?

By the way Eric, yes this is a copol, looks just like the one on your 2 satsig articles, and my instruction sheet says Andrew Model 611678203, which Google cannot find. The 2 output flanges in the same flat plane. Might the diplexer Rx filter be mostly a low pass, rather than a tight bandpass?

Suggestions?

Christian
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
★★★
Offline



Posts: 2109
Reply #4 - Apr 6th, 2013 at 8:39am  
Does adjusting the polarisation angle have any effect ?

If there was no carrier present on the opposite polarisation when the dish was installed it is possible that the polarisation angle set was wrong.  It needs to be accurate to about +/-2 deg.   This might not be noticed until such time as a new carrier appeared on the other polarisation.

Try rotating the feed by the calculated adjustment amount.
See https://www.satsig.net/maps/satellite-tv-dish-pointing-australia-new-zealand.htm
A positive adjustment means clockwise as viewed while standing behind the dish and facing forwards towards the satellite in the sky.

The receve filter will be a low pass filter to stop the 14 GHz transmit signal getting into the LNB.  if this filter is not quite good enough the effect is to degrade the receive path whenever the BUC transmits a burst.

Very low frequencies, into the antenna, will be blocked by the small circular wavegide in the narrow throat of the feed horn.
wxw
Best regards, Eric.
Back to top
 
 
IP Logged
 
chriswlan
Member
★★
Offline



Posts: 27
Reply #5 - Apr 6th, 2013 at 10:36am  

Polarization didn't seem to matter in the past.

Not sure about the other pol being used, here at 11646:
https://www.lyngsat.com/Eutelsat-172A.html

But I'll give it a try in a couple days. And triple check the pointing once again (not easy while .6 to .8 Ber). Might even take my good entire feed sytem and try it at the sick location...

Would radar pulses at 9500 Meg have no chance to get to the LNB at 11646? Ships are anchored offshore up to a mile away for months at a time.

When I tried another LNB, I did notice some dusty crud on half of the flange cover "contact surface" of the diplexer (all outside the O-ring); how best to clean that? Scotch pad? Almost looked liked the bolts on the side that is hard to reach were not quite tight enough. Should check for water too!

C
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
★★★
Offline



Posts: 2109
Reply #6 - Apr 6th, 2013 at 12:41pm  
Polarisation adjustment error of several degrees would not be noticable if there was no carrier present on the other polarisation. An error of 45 deg would only produce a 3 dB drop in signal level.

Once carriers are activated on the other polarisation transponder there are serious cross-pol interference issues both ways. You into them and them into you.

Try adjusting the polarisation and get it within +/-2 deg of the null.  The teleport hub should be able to help you over the phone and make your terminal temporarily transmit a CW carrier, on a carefully selected and agreed frequency, while they look on an analyser.

The resulting polarisation adjustment angle amount should be close to the calculated value (see above link).  Try placing an inclinometer sideways across some convenient surface on the LNB or BUC.

best regards, Eric.
Back to top
 
 
IP Logged
 
chriswlan
Member
★★
Offline



Posts: 27
Reply #7 - Apr 9th, 2013 at 2:07am  

SOLVED!!!!!!!!!

Went back to one site yesterday with my own entire feed assy.

Very interestingly it didn't show any difference, and still thinking of RFI we just dropped the entire dish and mount, put it on the ground, propped, pointed, used roof tin -no difference. Dragged the whole thing around the building, as far the back as the coaxes allowed, and still no real difference.

And as I had already tried my own modem a few days back, the coaxes testing OK, no mice bites etc, I got to wonder...

After plenty of cold rain, here by 2 degree latitude (see below...), and the signal was just slightly stronger and not showing any monster BER, just  0 to 5 % or so. So a meaningful true pointing was possible today, and still no dice.

Gee: only the reflector is left unchecked! Right away the strings showed a gap almost half an inch at the crossing point!!!!! Wonderful! Dish in tropical sun for only a year n a half! No software bug, no overheating fancy electronics, no no no! A near new piece of inert material is dramatically  faulty!

Solved: dish on smallish size to begin with, out of shape and the Eb hovers around the threshold point.

Still can't explain the strange "monster BERs", but I'll just give up on that part.


And today I did a quick string trick on the other site reflector, a Andrew dish, and here the gap is measuring at 5/8 at the crossing -so well over a full inch at the bottom leg location!!!!

And there is still several locations reporting intermittent Internet, so the potential for a "tropical epidemic" is still a possibility.

From now on a ziplock bag full of strings and heavy nuts will be part of my kit!

Thanks to everybody. See my other thread (s) re dish quality.

Christian
Back to top
 
 
IP Logged
 
Eric Johnston
Senior Member
★★★
Offline



Posts: 2109
Reply #8 - Apr 9th, 2013 at 9:00am  
Some dishes have rear side struts attached to the az/el adjustment assembly.  When assembling the dish the bolts at the ends of these rear side struts may be tightened while the edges of the dish are either pulled backwards or pushed forwards to prevent the weight of the BUC/LNB feed assembly distorting the dish. It helps to have three people working simultaneously.
See https://www.satsig.net/pointing/rear-side-struts.htm
wxw
With a larger 1.8m dish, without rear side struts and with very heavy BUC, I have improvised using a long metal girder sideways behind the dish with the ends of the girder secured to the side attachment points with adjustable steel cored cable ties.

Beware also, when the rear end of the feed support arm is attached to a steel spine behind the dish.  The joint between feed support arm and the steel back spine allows small fore and aft adjustment.  Adjust so that the hole in the lower edge of the dish is aligned, before a bolt is gently dropped into that hole.  If you put a bolt in the hole first the lower edge of the dish may be pushed forwards or backwards.

Some larger dishes have an optional extra part comprising a push rod attached towards the bottom of the main upright mount pole and pointed at the rear end of the feed support arm.  Once everything is set up and pointed, this is adjusted to push the rear end of the feed support arm forwards and reduce the stress on the dish caused by the heavy weight of the BUC/LNB feed assembly.

The stress in the dish can go other way. At low elevation the sides of the dish can get pulled forwards. At very high elevations the sides may be pushed back.  There may be some intermediate elevation angle when the dish is exactly balanced in elevation.  In this situation stiffness of the joints is important to prevent rocking in the wind, as there is no weight bias to hold the elevation steady.

Best regards, Eric.
Back to top
 
 
IP Logged
 
Pages: 1