Social Feed

I picked up a KBSL1 LNBFs to evaluate its use on a couple of my toroid dishes. Thus this will be a sometimes narrow look at the unit and doesn't really qualify as a full review.

To date I have been bringing into service two Wave Frontier's T90 dishes to cover most of the NA Ku birds. While they work fine for my purposes, they do lose gain the farther off-center one gets, and if one wants two degree spacings, you either need to overlap dishes or find feeds that can be squeezed in. So far I have been using refugee LNBs and feeds from Dish Network Superdishes. Being bandstacked, they also reduce the switching demands by a factor of two, which when you have 25 toroid LNBs can be helpful.

The DN 121W feeds can be separated from their Siamese twin 119W feeds with a hacksaw as has been frequently demonstrated on this forum, and are very narrow widthwise. The 105W feeds are wider, but one can alternate the 121 and 105 feeds to easily get two degrees between each LNB. This has worked well for me mechanically, but the DN LNBs are not up to modern standards in terms of performance. Coupled with the loss of gain on the extremes of the T90, I have too much of a variation in performance across the satellites to be satisfied. At the moment I'm considering a third T90 to center the LNBs better, but I also thought it might be interesting to see how Sadoun's new LNBF compared performance-wise. There are very few alternatives for bandstacked Ku LNBFs in NA.

Out-of-the-box the KBSL1 fits the T90 LNBF holders perfectly. But the plastic shell quickly gets in the way for tight spacings on the T90 rail. So I liberated the unit from its shell as a first step (see pictures). While this makes mounting a little more challenging, the diameter of the waveguide is pretty much the same as for the DN FSS feeds, and I already had solutions for those. The shell is not watertight, and in my mind is more of cosmetic than practical use. The feedhorn mouth on the Sadoun unit is larger than the DN 105W feeds, but it is still possible to alternate it with the DN 121W feeds to get two degree spacings.

I wasn't interested in the on-center performance differences between the bandstacked LNBFs, because the DN units already work fine on the T90 in those positions. The extremes are where I have problems. On the other hand I did not want to get into an extensive toroid realignment exercise. These two goals turned out to be fairly mutual, because I decided to go for rough alignments on the T90 and see how well the DN units did against Sadoun's.

The first test was with RTV on 83W. I set the T90 so this bird hit about 10 degrees off-axis. The T90 acts like about a 75-80cm dish at this point. I tweaked the dish for the best performance on a DN 121W feed, which isn't very good. I used one of my new Prof 7301 tuners and got a solid lock at 25% SQ with no video breakups. I know from an earlier test that the 7301 will hold lock on RTV down to about 20% SQ. I swapped in the Sadoun LNBF and the SQ jumped to about 45% SQ. Much better, but somewhat expected.

Next up was 97W as I had some numbers from another forum member using his DVB World 2104 tuner on a 90 cm dish with the KBSL1. I decided to go out to about 15 degrees off-axis on the T90 and compare to a DN 105W feed as that has a higher performing feed than the 121W version. At this point the T90 is performing below 70 cm dishes so it is a stressful test. Looking across all the transponders, the Sadoun unit was mostly in the high 40s and low 50s in SQ on my DVB World 2104 (not the Prof) with the 105W unit performing 10-15 points lower. My friend's 97W numbers were mostly in the mid 80s for SQ with the same tuner model, but centered on a 90 cm offset.

I did run my Prof 7301 against the Sadown LNBF, but nearly all the SQs were at 100%. But this is where a fly in the ointment appeared. Here are the DVB World 2104 numbers (SL/SQ) for the highest frequency transponders on 97W with the KBSL1:

12122 H 20000 91/51
12146 V 22000 91/42
12152 H 20000 91/52
12177 V 23000 91/34

These are the Prof 7301 numbers from the same cable:

12122 H 20000 100/100
12146 V 22000 100/100
12152 H 20000 100/100
12177 V 23000 97/90

Whoops! What's going on with those V transponders? My friend had first highlighted a similar SQ drop by 20-30 points on these, so mine was not a single bad LNBF. Off popped the LNBF feed cover to reveal what is likely going on. In the picture of the feedhorn mouth, one can see the probes, but it's a little hard to note the relations between them. I scratched my head to figure out a better way of showing this, and was able to get my camera to get an odd but useful photo. The H probe is near the back of the waveguide, which is normally the best place to put a probe. But it can be tough to squeeze in probes for both polarities, so the designers for the Sadoun LNBF chose to move the vertical probe forward, and put a vertical post behind it to simulate a back plate. I've seen other cases like this and this design compromises the performance of the forward probe. I'm pretty sure this is one if not the main reason why the V performance falls at the highest Ku frequencies.

What bothers me in particular is where the spectrums fall. The KBSL1 has LOs of 10750 for H and 10100 for V. That means 11700-12200 MHz runs from 950-1450 MHz for H and 1600-2100 for V. This puts the poorest performance of the V polarization at the very top of the L-band spectrum coming down the cable to your receiver. This is where one would expect the most signal degradation, and the degradations will combine.

By comparison Dish used LOs of 10750 for V and 13850 for H. The means 11700-12200 MHz runs from 950-1450 MHz for V and 2150-1650 for H (note inverted spectrum). The DN LNBs have PCB probes at the same optimal distance from the waveguide and do not show a similar degradation. But any stressful performance problems at the top of the H range are passed mid-L-band where distribution systems are likely to be benign. This choice would have helped the KBSL1, and another alternative would be to switch the LOs so the more stressed polarization came down the cable at 950-1450 MHz instead of 1600-2100 MHz.

After painting the picture a little on the gloomy side, it is important to note some mitigating information. First I suspect the DVB World 2104 tuners provide very misleading SQ measurements at the top of L-band, particularly when the noise floor is not flat across the spectrum. I was surprised to discover this a couple of months ago chasing down an unrelated problem. Both my friend and I see similar drops in SQ for the top V transponders when using the 2104s. But look at the Pro 7301 numbers: the SQ goes down from 100% to a still respectable 90% and only for the highest frequency V transponder.

This is largely confirmed by actual demodulation performance. Normally my 2104s start losing lock when the SQ falls a few points below 40%. This is when the noise floor is flat, or I'm only using the lower 500 MHz on the cable. But at a 34% SQ, the DVB World was locking perfectly and wasn't even close to losing lock (I did a quick test). The Prof 7301 normally loses lock for me just below 20%. It was coasting here at 90%. This makes me fairly certain that the 2104 uses a SQ algorithm that yields meaningless numbers with a high-end rolloff in L-band. The 7301 is more reliable.

The reduction in performance on the high V transponders is real. The 2104 exaggerates it, but it is measurable on the 7301. But both units still receive fine under these conditions. While it would be nice to see this improved, overall the KBSL1 is a definite winner in performance over the bandstacked DN units.

From a personal perspective, it would have been nice if the Sadoun unit had used the established DN de facto standard for LOs, because a certain number of receivers and PC software already have those LOs included. It's certainly possible to ignore the fact that the LNBF is bandstacked and tell the receiver to search from 11700-12800 MHz, but then one is forever translating frequencies back and forth. On top of this DN made a rather nice Dish Pro adapter for their legacy receivers that provides the receiver with 950-1450 MHz for both H & V off one of their bandstacked units through normal voltage selection. This makes the LNBF appear to be a standard switching LNB to receivers that don't get along with two LOs. Unfortunately with a different spectrum map, you can't do this with the KBSL1. Nevertheless this is more of a nit for a unit that performs very well. I'm very thankful to Sadoun for offering a bandstacked LNBF to the NA market, because while these can be very useful items, they have been hard to find.