Norsat LNB Whats Wrong ? pics

[Rd100]

I do not know if it will solve your problem can you get any nilesat horizontals at the moment? The best set up to have for nilesat would be a invacom c120 quad or a norsat 4106a with rg11 or wf125 for low signal loss, That is it or get a bigger dish

Ok RD100 your point is made clear about the RX/TX OMT versus the RX only. So no more LNB/BUC dual flanges.
The 611-6062-07 and 611-6062-10 seem to use the same OMT.
The 07 is wide band and the 10 is (11.7 to 12.2 ghz).
Only the actual feedhorn is varied.
For Showtime on Nilesat the use of two 4000A LNB's with the 611-6062-10 will solve the problem.
Could you confirm?

 

[canofan]

I get all horizontals now that it is not overcast.
I use an invacom quattro.
I shall now order a 611-6062-10 to use with my norsats with a low loss multiswitch I should be able to have the same result as you had achieved with this same setup.
I could also upgrade to rg11. besides as the intermediate frequency is lower there will be again less loss in the cables.

 

[canofan]

Originally Posted by Rd100

For maximum signal from the norsat lnb it should be bolted direct to the feedhorn then skew it by hand for horizontal or vertical that is it there is no point in trying to add more lnbs to the feed as you will get to much signal loss and you then might as well use a invacom.​

Do you mean the loss of the signal getting divided between the two LNBs?

The dual flange acts as a filter. It impedes the horizontals in the vertical channel and the same to the verticals in the horizontal channel. As what is impeded does not go through, it remains available for the other channel (filter theory). So practically there would be no loss and the reception gain will remain virtually the same. The VSWR is the same for both channels when used in the same band. If one channel is used for the higher band 12.2 to 12.75 ghz the VSWR will go up slightly because the flange is optimised for the 11.7 to 12.2 ghz.

 

[keano]

Guys just to let you know, that I am around and will try the norsat later this week.
Rd100 got the 4708B of ebay, in fact got a shed load of them for next to nothing and brand new

If there is a massive difference then I'll look at the 4106A but at the mo Nilesat is coming in nice and strong and OK with c120 Invacom quad.

 

[keano]

Just to let you guys know had a quick play with norsat lnb ( at last ) and I am shocked at the difference.

Having a B version I was only able to try above 12.20, so I tried 7.0w 12226H.
With the invacom quad c120 I was getting 61% on the db, switched to the norsat and this went to 67%.

I was also even able to tune in the V transponders too, which is weird because I thought these lnb's were single only.

At the mo showtime is coming in load and clear on the invacom so not to bothered but come winter and rain time it might be good to have one.

Cheers guys

 

[canofan]

Bravo Keano,
I assume that you used the corrugated wideband 611-6060-07 feedhorn.
With the results you obtained, I can infer that, in the lower band (11.7 to 12.2 ghz), with the appropriate (611-6060-10) narrower band feedhorn and the norsat 4206A, one should expect even more miracles.
I am still waiting for the proper OMT to materialise. It may be possible to obtain the two flanged Andrew RX OMT separately.

 

[keano]

OK m8, now I though the norsat 4106A would be better than the 4206A.

Now the thing is trying to find a dealer who sells them, in Europe I can't find anyone at all.

Oh and yes I did use the recommended feedhorn with the norsat so I will also try the narrow band feed soon.

 

[ouagadougou]

Its difficult to make a dual polarised solution using these rectangular flanged LNB's and maitain the low noise performance because the LNB's have Rectangular WaveguideRectangular75 waveguide ports. Single polarised low noise systems will work better.

If the feedhorn has a WR75 flange then it is not designed for dual polarisation so forget dual pol with the 611-6060-07. The 611-6062-07 feed has a Circular120 flange before the dual pol WR75 spliter/transition. The dual WR75 transition is a splitter which will introduce some loss before the LNB inputs which defeats the improvements gained by using these low noise LNB's.

Dont use an OMT for this solution the 2nd port is not designed for the receive frequency passband it is designed for the higher uplink frequency range at 14GHz.

Based on keano's results it seems we can trust the Norsat specifications which is a nice change from the mass market LNB's.

Cheers.

 

[canofan]

keano,
The 4000 series incorporates filtering because it is mainly used in conjunction with a nearby BUC.
This filtering makes it more expensive than its little brother the 7000 series.
Also, in our case we do not need a more expensive precision local oscillator frequency.
A stability like that of Invacom should do: "1Mhz".
So a 7706A (750mhz)or 7906A (900mhz) should do if you do not have cable lengths problems (the 7000 have less LNB IF gain than the 4000).
I would use the best noise figure for the Horizontal transponders.
I also investigated the availablity of the cheaper japanese NJR2744H single DRO lnbs. This has specs comparable to the 7000 series from Norsat but cannot deliver a good noise figure (only 0.8db). So I would forget about them unless they are used only for the vertical transponders.

 

[canofan]

Its difficult to make a dual polarised solution using these rectangular flanged LNB's and maitain the low noise performance because the LNB's have Rectangular WaveguideRectangular75 waveguide ports. Single polarised low noise systems will work better.

If the feedhorn has a WR75 flange then it is not designed for dual polarisation so forget dual pol with the 611-6060-07. The 611-6062-07 feed has a Circular120 flange before the dual pol WR75 spliter/transition. The dual WR75 transition is a splitter which will introduce some loss before the LNB inputs which defeats the improvements gained by using these low noise LNB's.

Dont use an OMT for this solution the 2nd port is not designed for the receive frequency passband it is designed for the higher uplink frequency range at 14GHz.

Based on keano's results it seems we can trust the Norsat specifications which is a nice change from the mass market LNB's.

Cheers.

Ok about the OMT

The Junction block for the dual polarity RX only waveguide is the same for 611-6062-10 and 611-6062-07.

The VSWR is unchanged in the two channels.

I repeat what I said earlier:
The dual flange acts as a filter. It impedes the horizontals in the vertical channel and the same to the verticals in the horizontal channel. As what is impeded does not go through, it remains available for the other channel (filter theory). So practically there would be no loss and the reception gain will remain virtually the same. The VSWR is the same for both channels when used in the same band. If one channel is used for the higher band 12.2 to 12.75 ghz the VSWR will go up slightly because the flange is optimised for the 11.7 to 12.2 ghz.

 

[ouagadougou]

Ok about the OMT

The Junction block for the dual polarity RX only waveguide is the same for 611-6062-10 and 611-6062-07.

The VSWR is unchanged in the two channels..

That may well be (dont have 1 in front of me) but the dual pol transition has more insertion loss than a straight connection to a single LNB. VSWR does not not necessarily have any direct relationship to insertion loss and the dual pol transition definately has insertion loss. This insertion loss will always mean that the dual pol solution will perform worse than a single pol solution alone. Insertion loss before the LNB increases the effective system noise temperature more significantly than the value of the insertion loss itself because the system noise temperature is increased disproportionately to the insertion loss itself.

I repeat what I said earlier:
The dual flange acts as a filter. It impedes the horizontals in the vertical channel and the same to the verticals in the horizontal channel. As what is impeded does not go through, it remains available for the other channel (filter theory). So practically there would be no loss and the reception gain will remain virtually the same. The VSWR is the same for both channels when used in the same band. If one channel is used for the higher band 12.2 to 12.75 ghz the VSWR will go up slightly because the flange is optimised for the 11.7 to 12.2 ghz.

True the dual flange/dual pol transition acts as a polarisation filter but its the insertion loss it introduces that results in reduced overall sensitivity. I wouldnt worry too much about VSWR in a receive only system as system noise temperature is the overiding concern and any loss before the LNB input will result in lower receive sensitivity. None of these rx waveguide systems have tuning screws in the receive branches so none of them are VSWR optimised.

 

[canofan]

There are on the market some very expensive RX only OMTs with screws to tune. The Andrew junction blocks are optimised for this application only and do not need the screws.
Qualitatively you are right about insertion loss. Nothing to insert is always better. But how much quantitatively does this insertion worsen the signal? Perhaps is it marginal and the benefit/comfort of using one may be worth it.
If you told me that the insertion loss may reduce SNR by 1.5db I would be unhappy about it; but 0.2 db I am ready to accept.
It seems to me that only by testing such a rig could we find out.

 

[ouagadougou]

There are on the market some very expensive RX only OMTs with screws to tune..

The whole purpose of the OMT is for dual pol transmit/receive applications so I still doubt you will find something suitable for a dual pol receive only application. Even dual pol LNB's are a compromise due to the dual probes in the waveguide to coax transition in the LNB.

The Andrew junction blocks are optimised for this application only and do not need the screws..

The Andrew devices are passive with no optimisation at all and will introduce insertion loss as all inserted transitions/splitters do.

Qualitatively you are right about insertion loss. Nothing to insert is always better. But how much quantitatively does this insertion worsen the signal? Perhaps is it marginal and the benefit/comfort of using one may be worth it.
If you told me that the insertion loss may reduce SNR by 1.5db I would be unhappy about it; but 0.2 db I am ready to accept.
It seems to me that only by testing such a rig could we find out.

Every 0.1dB of insertion loss before the LNB/LNA input adds at least 7.5 degrees kelvin to the system noise temperature.

In sensitivity terms it looks like this:

LNB Noise Figure = 0.6dB same as best Norsat
LNB Noise Temp = 43 Degrees Kelvin
Noise Bandwidth = 28MHz Signal Bandwidth
Sensitivity Noise Floor = -167.79dBm

If we add a filter or splitter with 0.1dB insertion loss which adds at least 7.5 degrees kelvin to system temp then:

Resulting Noise Temp = 50.5 degrees Kelvin
Resulting Noise Figure = 0.7dB
Noise Bandwidth = 28MHz same as before
Resulting Sensitivity Noise Floor = -167.10dBm

So in this case 0.1dB insertion loss = at least +7.5kelvin to system temp = 0.7dB reduced sensitivity. So not directly related.

If you have an insertion loss of 0.2dB = at least 15 degrees kelvin is added to system temp and then sensitivity is reduced by at least 1.3dB so it begins to get silly.

Never ever insert anything before the LNB input if it introduces insertion loss because the degradation in sensitivity does not directly correlate to the amount of loss inserted. This is purely and simply due to the laws of physics and because the receive systems we are talking about are at the extremes of sensitivity. Terrestrial radio systems dont exhibit the same degradation in practical applications because the system temperatures are determined by the warm earth at 290kelvin(ish).

In the case of horizontal transponders on 7w, myslef, keano and others who have tried from NW England are right on the fringe so every little bit helps and 1.3dB sensitivity reduction is huge.

 

[canofan]

Very nice, thank you.
Pardon me if I am trying to plck your brain.
But I want to get back to reevaluating the Invacom Quad QPF31.
_www.satellites.co.uk/satellite/archive/t-113826.html
_http://www.invacom.com/products/qpf_031.htm
_http://www.invacom.com/products/documents/2ndgenQuadpolardatasheetwhite.pdf
This has 4 outputs 2 of which concern us for Nilesat.
With its 0.3db noise figure.
and its two bands 11.7 to 12.2 and 12.2 to 12.7 voltage switchable for polarity and 22khz switchable for band selection.
Qualitatively we know that the switching circuits introduce insertion loss.
Quantitavely how much insertion loss do they introduce?
Do the losses offset the 0.3 db advantage so that they become less interesting than the Norsat?

 

[ouagadougou]

With its 0.3db noise figure.

The Invacom has a claimed Noise Figure of 0.3dB in reality this is just sales talk. It is much more likely to be between 0.6 & 0.8dB for an ambient operating temperature in air with no assisted cooling.

Qualitatively we know that the switching circuits introduce insertion loss.
Quantitavely how much insertion loss do they introduce?
Do the losses offset the 0.3 db advantage so that they become less interesting than the Norsat?

The losses are at IF so they do not have a direct affect on the LNB input noise figure. What you have in fact is 2 LNB chains operating in parallel, 1 for vertical and 1 for horizontal receiving both polarisations simultaneously from idividual probes in the LNB waveguide section. These parallel amplification chains are downconverted to IF and then split and fed to 4 IF switches for H/V selection. I guess one reason for the better performance with the quad LNB is that it has a bit more IF gain because of the need to feed 4 ports, just a guess because I havent measured it.

1 chain will be more sensitive than the other in practice just as any 2 individual LNB's are never the same. If you find that the H or V performance is better then rotate the LNB 90 degrees and use the better performing H or V side to receive the weakest signal and then adjust the receiver to change the polarisation setting.

I wish I had as much time to play as you guys. Hope your wife is ok with the dish now keano.

Cheers.

 

[canofan]

The Invacom has a claimed Noise Figure of 0.3dB in reality this is just sales talk. It is much more likely to be between 0.6 & 0.8dB for an ambient operating temperature in air with no assisted cooling.

How sure are you about this statement? have any measurements been made or experiments in that direction to challenge the Invacom 0.3db noise figure?

The losses are at IF so they do not have a direct affect on the LNB input noise figure. What you have in fact is 2 LNB chains operating in parallel, 1 for vertical and 1 for horizontal receiving both polarisations simultaneously from idividual probes in the LNB waveguide section. These parallel amplification chains are downconverted to IF and then split and fed to 4 IF switches for H/V selection. I guess one reason for the better performance with the quad LNB is that it has a bit more IF gain because of the need to feed 4 ports, just a guess because I havent measured it.

1 chain will be more sensitive than the other in practice just as any 2 individual LNB's are never the same. If you find that the H or V performance is better then rotate the LNB 90 degrees and use the better performing H or V side to receive the weakest signal and then adjust the receiver to change the polarisation setting.

I appreciate very much this quantitative analysis. It is not mathematical but almost. As a result of the foregoing analysis:

1-apart from a possible experiment with the Invacom QPF-031 coupled to a 611-6060-10 feedhorn, and in case this experiment fails to give the expected results.
2-and apart from tryng to implement a chaparal corotor by finding a way to manually or automatically rotate the LNB to the proper polarity.

One should consider using two smaller dishes for the Nilesat: one for horizontals with 611-6060-10 and Norsat4206A and the other for verticals with 611-6060-10 and Norsat7707A or NJR2744H. It is possible to find used smaller dishes, so that should reduce the outlay and the whole thing will be less expensive than a bigger dish.

 

[ouagadougou]

How sure are you about this statement? have any measurements been made or experiments in that direction to challenge the Invacom 0.3db noise figure?.

I am absolutely sure. The best transistors on the market have 0.3 to 0.35dB Noise Figure at room temp but you will not achieve this in a working LNA/LNB because of input losses before the transistor. Most LNB's use a waveguide coupled to a microstrip input to the 1st transistor and this has measured loss of around 0.3dB hence lowest achievable noise figures of 0.6 to 0.65dB for the best room temp LNA/LNB's. I have verified this with friends who design low noise amplifiers. Lower noise figure requires a lower loss input circuit or amplifier cooling.

I appreciate very much this quantitative analysis. It is not mathematical but almost. As a result of the foregoing analysis:

1-apart from a possible experiment with the Invacom QPF-031 coupled to a 611-6060-10 feedhorn, and in case this experiment fails to give the expected results.
2-and apart from tryng to implement a chaparal corotor by finding a way to manually or automatically rotate the LNB to the proper polarity.

One should consider using two smaller dishes for the Nilesat: one for horizontals with 611-6060-10 and Norsat4206A and the other for verticals with 611-6060-10 and Norsat7707A or NJR2744H. It is possible to find used smaller dishes, so that should reduce the outlay and the whole thing will be less expensive than a bigger dish.

Forget the Chaparal Corotor as its too lossy for real low noise use and forget reduced dish size as that only makes your problems worse. I just live with the fact that I have to change manually when I want to change to C-Band and I would do the same with a single polarised solution for Nilesat if it meant I could get the H transponders. But my dish is ground mounted so no problem for me.

 

[canofan]

I am absolutely sure. The best transistors on the market have 0.3 to 0.35dB Noise Figure at room temp but you will not achieve this in a working LNA/LNB because of input losses before the transistor. Most LNB's use a waveguide coupled to a microstrip input to the 1st transistor and this has measured loss of around 0.3dB hence lowest achievable noise figures of 0.6 to 0.65dB for the best room temp LNA/LNB's. I have verified this with friends who design low noise amplifiers. Lower noise figure requires a lower loss input circuit or amplifier cooling.

How about many low rbbs in parallel? they consume a bit more but generate proportionately less noise, but may be they just do that in LNBs.

Anyway it seems that Invacom need to have a sales talk because they address the consumer market, while Norsat addresses the more knowledgeable industrial market.

Forget the Chaparal Corotor as its too lossy for real low noise use

Yes, I have come to the same conclusion just after I submitted the post

...and forget reduced dish size as that only makes your problems worse. I just live with the fact that I have to change manually when I want to change to C-Band and I would do the same with a single polarised solution for Nilesat if it meant I could get the H transponders. But my dish is ground mounted so no problem for me.

Every time I would need to change polarity I should go up to the roof 8 floors up. But an extra dish for the less demanding verticals would help when the original dish is optimised for the horizontals.

 

[canofan]

I now use two dishes for Nilesat.
I have set up the CM120 for Horizontals with a Norsat7706A and an Andrews 611-6060-10.
I obtained 10% increase on my DB from 67 to 77%.
For Verticals I use the Triax140 Prime Focus with also a Norsat 7706A but with an Invacom feedhorn. Here I got only 6% increase from 73 to 80% even after best tweaking of the length of the horn.
I bought my Norsats and Andrew feedhorn from the US. The feedhorn assembly is complete with a transition o-ring and cover. The Norsats I bought off US E b a y.
I feed the Norsats through a multiswitch and select them through the diseQc of the receiver.
I am very pleased with the result. But the weather has been favourable and I keep wondering, for the verticals, if a Triax110 with the same Norsat but a 611-6060-10 instead of an Invacom feedhorn would not have been just as satisfactory.

 

[keano]

m8 I got a 4708A Norsat of ebay USA and it came yesterday but the results were not good.
It performed 4% less than the invacom.

I thought the 4000 range of norsats were better than the 7000 range.

I am glad at your results, so you are able to watch showtime OK now ?