iDirect Rx Saturation Level

When you say "saturation"...are you speaking to Rx_power on the line cards and remotes? If so, I believe the optimum value you are looking for is approx -36dBM (+/- 5 dB is my rule of thumb).

Is in the Manual of the iDirect Infiniti 300:
RF Power Range
Transmit:-35 dBm to +7 dBm
Receive:-65 dBm to +0 dBm composite
and in the trainig of the IOM presentation
'Network Operations, Basic Acquisition & Troubleshooting
is from -65 to -5 dBm.

Receive level at a remote
The total composite power into the remote L band input socket must not exceed -5 dBm or 0dBm (in above examples) as this would saturate the L band tuner of the modem.  This power comes from the LNB.  The circumstances of excessive power are when the dish is abnormally large (2.4 - 13m diameter) and the satellite is also broadcasting 30 TV carriers, each 27.5MHz wide and intended for 30cm dia dishes, and short LNB cable.   Such high composite power from a satellte can overload the LNB output stages and you need special low gain LNBs for use in large dishes.  To measure the composite power you need an Agilent (HP) Power meter and N type power meter head (diode or thermocouple) or similar.
In a VSAT don't worry about exceeding the max power into the remote.
As far as the lower limit is concerned, e.g. -65 dBm, the problem could arise with excessive length cable loss between the LNB and the modem.  Read the wanted carrier level in the modem software and compare with the specification.  The specification might say something like Desired carrier: min -130+10log(Symbol rate) dBm, max -90+10log(Symbol rate) dBm.    
Alternatively note the receivce C/N or Eb/No readout.  Then add a length of extra coax to the receive cable to add extra loss of say 6 dB.  This should make the level of the wanted signal go down by 6 dB but should make no difference to the C/N or Eb/No, which should stay the same. (You can't use a 6 dB 75ohm attenuator pad to make this test as the DC would burn it out).  If the C/N or Eb/No does goes down you are clearly approaching the noise floor of the modem input itself. You need a shorter LNB cable, lower loss cable, in-line L band amplifier or higher gain LNB.  Remove or redesign any splitters in the LNB cable.  Note that poor connection in the LNB cable results is low DC supply volts to the LNB and increased L band losses, leading to low signal levels.  At commissioning record levels for reference so you can check remotes from time to time to proactively identify trouble before the customers have an outage.

Transmit level at a remote
When you set the transmit level at a remote site you must start with a very low level, like -35dBm and increase it only slowly in 1 or 2 dB steps, while watching the CW signal at the hub. You will see it increase in 1 or 2 dB steps.   Plot the values on a graph if you wish.  As you increase the remote power you will find that the BUC starts to go non-linear when the measured level lags 1 dB behind the power settings at the remote then BUC is saturated and you must not go any further.  Record the P-1dB remote transmit power setting in the hub database as the level never to be exceeded in service.  

The uplink power control sytem will reduce the power by several dB to achieve the normal clear sky operating point.

So a 2 watt BUC might normally operate at 0.5 watt output in clear sky and increase during heavy rain to 2 watt (6 dB increase due to UPPC).
It is very important the the UPPC system does not try to increae abve the 2W.  That is why you measure the P-1dB point and then know what modem output power corresponds to the 2W (or whatever is the nominal BUC power).

Ideally, all TDMA bursts arriving at the hub should be at the same level at the line card (+/-1 dB) and all with the same C/N at the hub line card.

If you put too much power into a BUC, even for a moment, you risk damaging the BUC slightly or catastrophically as the output transistors are partially or totally burned out.  Prolonged operation above saturation with the modulation on is bad for the heath of your BUC !

The picture below shows the transfer characteristic of an amplifier with specified power rating of 1.25W.

This graph refers to an amplifer with 22dB linear gain.  Typical BUCs have 50 to 60 dB gain and have 1, 2, 3, 4 or 5 watts rated output.

Best regards, Eric.

I don't understand how you are altering the level.  

If you alter the receive level into the remote modem by swapping in/out various lengths of coax cable in the LNB to modem cable run then that will vary the level of all signals into the remote modem, including the noise floor.  Only if you have exceptionally long cable and thus low levels will you get problems.   The noise floor of the modem itself will have an effect and your outlink C/N and Eb/No will deteriorate.

If you really think that your composite power into the remote modem is exceeding -5 dBm then make the LNB cable much longer.   79 equal power carriers all from the same satellite, each at -24 dBm would make a total of -5 dBm.

If you alter the receive carrier level into the remote modem by adjusting the transmit power at the teleport then you are affecting the C/N and Eb/No directly of that site and all the other sides.  You need to set the nominal level as agreed with your satellite operator.  You pay for a particular power and bandwidth from the satellite.

If increasing the transmit power of the outlink carrier at the teleport causes errors at the remotes then you are likely overloading the BUC or HPA at the teleport hub.  This may also cause errors on the TDMA return links to the teleport due to interference caused to adjacent frequencies either side of the outlink carrier.

The set up and calibration of the transmit power at the teleport should include finding the P-1dB compression point of the hub teleport BUC, particularly if the hub BUC is only a few watts.  In this case connect the output of BUC to a dummy load and measure the output power with a cross-waveguide coupler and Agilent power meter.   If the teleport hub has an HPA it probably has an inbuilt power meter (say 0 - 200 watts).  Use the reading.   Check the output spectrum for spurii and spectral regrowth and intermodulation products if you are transmitting multiple carriers.  I have seen 150W TWTA HPAs that saturate at 10W due to deterioration of the cathode emission.  Record the voltages and currents on the meter every few  months.   Don't test hub HPAs while the big teleport dish is pointed at a satellite.

Best regards, Eric.

The limit of -5 dBm in the manual refers to the applied composite total of all received carriers from the satellite.  The only practical way of measuring (apart from measuring each carrier seperately and adding them all up) is to insert a DC block and provide the LNB power via a side cable.  Then connect a power meter sensor head (of the correct power sensing range or using an attenuator) to the end of the DC isolated L band coax cable. You may need a 75/50 ohm adaptor.  Use an HP Agilent (or similar) power meter to measure the total power - typically using a thermocouple.  Be careful, the test gear is very expensive.

79 equal power carriers all from the same satellite, each at -24 dBm would make a total of -5 dBm.

Overload of the modem receiver tends to occur with very high power satellites or large receive dishes. With both combined you can even overload the LNB output stages, so you need special low gain LNA/LNBs or high power output LNA/LNBs.

Are you receiving from some abnormally high power satellite or group of co-located high power satellites ?

Can you see other high power carriers ? What power and how many of them ?

If the spec says -5 to -60 dBm then your modem might receive a -15dBm carrier if that was one carrier out of 10 similar carriers on the satellite (total power -5dBm).  Alternatively it might receive a -40 dBm carrier, being one out of 3100 similar carriers (total power -5dBm).  I would not advise operating any amplifier or modem at its extreme spec limits.  Always try to set up signal levels through amplifiers working well clear of the noise floor and well below saturation and intermodulation problems.  The problem most commonly occurs in L band and 70 MHz IF distribution systems in large earth stations.

Best regards, Eric.

SCPC errors and TDM loss are certainly indicative of disruption of the downstream carrier.  Either locally or as it is uplinked from the hub.  In your case it is locally.  I will let Eric chime in, but short of relocating that dish, your options may be limited.  What kind of downstream SNR are you seeing to that system (rx snr)?

Local interference from car ignition, neon signs, switches, air conditioners, radar etc can cause interference into the LNB.  C band interference from wireless broadband is common.

The problem may be reduced, but not necessarily eliminated, by earth station site shielding.

Interference is also possible due to direct injection of L band (and lower frequency) interference into the cable between the LNB and the modem.  In this case, adjacent cables, cell phones (both base stations and mobiles) and any nearby wireless transmitters are possible sources.  The problem will be aggravated by poor termination of the coax cable screen at the ends or an earth loop. It should be possible, in theory, to eliminate the problem with high quality 100% sheathed cables.

If the antenna base is earthed (often by law) then try a really thick earth cable between the antenna and the indoor rack, routed close alongside the LNB coax.  Earthing the cable sheath near the building entry (often by law) helps to reduce interference and damage if lightning hits.

Best regards, Eric.

Quote:

The thick cable earth idea was only intended as a possible solution to earth loop type interference - which might happen if the antenna base is earthed and the modem is earthed elsewhere.

With the LNB/feed in your hand you can point the feed directly at possible nearby interfering sources while watching the analyser.

Quote:

Does this mean the different netmodem was not affected by the interference ?  Or was the different netmodem affected much worse by the interference ?. Is the interference getting into the netmodem via the lid joints, via power supply or via the ethernet cable ?

Look at the screen in the cable.  If there is a solid copper or aluminium foil wrapped round and overlapped, that is good.  Mesh wire braid with spaces between the wires is poor.  Note that some cables have plastic foil with aluminium coating (Rf screening) plus wire braiding (for the DC supply).  With moisture, the aluminium coating can turn to a white powder (Aluminium oxide) and become quite useless, while the DC supply happily goes up a widely spaced wire braid wires.

The best screen cable, Andrew Heliax, with solid corrugated copper pipe as the outer screen is too expensive for customer installations.

Best regards, Eric.

Quote:

I would try to find a location for the antenna so that it is not visible to the road.  As long at it can still see the satellite, the antenna is fine at ground level surrounded by walls or on a flat roof surrounded by low walls.

Quote:

Worth a try but don't raise your hopes too much. Interference from car ignition is wideband and some of the energy will be at the wanted frequency. If that is causing the errors then changing the LNB will not help.  The total power of the interference may be the problem, in which case the LNB is saturating.  A different LNB with higher output power rating may help.  Invacom VSAT LNBs are particularly well designed. Borrow the LNB if possible so you can take it back if there is no improvement.

Quote:

Not recommended due to high cost and low probability of success in this case.  If pointing the feed (or dish) at the traffic makes things worse then the interference is in Ku band and is getting in via the feed horn, not via the cable sheath.

Best regards, Eric.

Adrian,
We had those problems too and if the noise is coming only from the cars you have two solutions as Mr. Eric said, first make a interference study to find another place or and earth station site shielding (Cage) Trying different LNB or cables is not a solution for my based in the problem that you have.