Frankenstein Triple Ortho or Dual Ortho Universal

[pendragon]

This thread has been a long time in in the making, but thanks to a kick from zamar23 and today being completely unproductive with school starting for all the kids, I threw in the towel and took some pictures on the roof.

Long ago I started with a 3m Winegard perf and decided to put on a dual orthomode feed with Norsat 8115s and 2707s as LNBs. As it was my only FTA dish everything worked fine, or so I thought. I could lock just about everything on C and Ku with good SQ. But the family traffic got a little heavy on the 3m, so I started adding more dishes. I wasn't very happy, at least initially, to find that a Geosatpro 1.2m with an Invacom QPH-031 LNBF (quad Ku with linear and circular polarizations) pretty well equaled or bettered Ku on the 3m. Then I put up a 1.8m solid for C-band only, and while the 3m stomped it, the difference wasn't as much as I calculated that it should be.

This started me down the road of feed modifications. I wrote a thread some time ago about hacking up the scalar rings on a dual ortho to better match a deep dish like the Winegard. That got my C-band performance up on the 3m to where it made sense given the 1.8m performance. This demonstrated the dual ortho was initially under-illuminating the Winegard's f/D of 0.278. While I could fix C-band, there was not much I could do to get the Ku part of the feed to see more of the 3m.

A 2.3m mesh arrived next with a f/D of 0.35. I figured maybe this would be a better match for the dual ortho because of the higher f/D. But Ku performance was horrible. My conclusion was the surface was simply too inaccurate for Ku, although worked fine for C.

The final stop for the dual ortho was my homely 1.8m. I had since put single ortho (C-band only) feeds on the larger dishes and measured a slight improvement in performance. I was nearly ready to dump the dual ortho, but I had picked up an Invacom QTF-031 quatro universal Ku LNB with a prime focus ADF-120 feed and was trying to decide between matching the ADF-120 to an extra Geosatpro 1.2m that was collecting dust or a second 1.8m prime focus that would have to be acquired. The Geosatpro would have been an interesting experiment, as it has a fairly low f/D of 0.5. I wondered whether a carefully modified ADF-120 would be a better match than an offset feed designed for say f/Ds of 0.6 and above. Under-illumination again.

I modified the ADF-120 but never tried it on the 1.2m because the first test was the ADF-120 + QTF-031 combo on the 1.8m. I had a Ku clamp for this dish, but it was lousy and I wasted far too much time trying to get the assembly mounted and centered. Performance was good but mechanically awful.

Then the light clicked while looking at the soon-to-be-sledge-hammered dual ortho feed sitting next to the 1.8m. That was mechanically rock solid on the 1.8m, and perhaps I could pull off the Norsat Ku LNBs and substitute the Invacom QTF-031. This would save buying another 1.8m and I had a friend who was interested in my extra 1.2m. Because there isn't a lot of low band Ku universal available in Colorado, the Invacom would be a good operational match to the primarily C-band only use of the 1.8m.

So off with the Ku waveguide and Ku LNB mounts (see first picture). Note on the right side the tubing has three successive diameters; this slides into the back of my ADL RP3-2-2B feed and is held with set screws. The Chaparral Bullseye looks very similar and can likely be similarly modified.

Next came an otherwise useless flange leftover from a Superdish 121 feed assembly (next picture). I had liberated many 119 and 121 feeds from each other with a hacksaw to use the 121 LNBs and feeds on my toroid. That left me with a box of 119 feeds, flanges and LNBs to stare at. The flange is a C120, which is the same used by the Invacom, and the waveguide diameter was only a bit larger than the dual ortho port in the back. Off with its head! (see next picture).

The rest was easy. One can either mill or file the Superdish tube to fit the dual ortho feed. It works best with two steps as in the original Ku waveguide. I was on the roof so a file seemed the more appropriate weapon, and in a couple of minutes everything fit perfectly. Down came the lousy Ku feed holder on the 1.8m (and straight to the sledge-hammer) and up went the Frankenstein dual ortho C/Ku-universal assembly. It performed so well that I pardoned the dual ortho and it became my permanent choice for the 1.8m. The only headache was dragging a total of six coax lines back to the switch matrix from my farthest dish.

Another alternative would be to mount an Invacom QPF-031 LNB (not the far more common QPH-031) in a similar manner. Then you would have a triple ortho with linear C, linear Ku and circular Ku. I have resisted the urge to build a quad ortho so far, but am working on options for linear and circular C on the same dish.

 

[Cadsulfide]

Pendragon;

Very Cool, isn't it great when the light bulb blinks on? I have a pile of unused SD 121 waveguides and you have found an excellent solution for subverting them to something useful again. Great Post!

 

[zamar23]

Pendragon

Very interesting solution. Am I correct, you cut a piece of a standard Superdish flange to a calculated length (how calculated?), then inserted the flange into ADL RP3-2-2B feed's Ku-band port, and mounted Invacom Universal Quatro LNB on that flange? How did you fix the flange inside the feedhorn with screws - are they included in the feedhorn assembly? Was your Scalar Ring included with the feedhorn? What makes Norsat C-band LNB better than others in your view?

 

[pendragon]

Am I correct, you cut a piece of a standard Superdish flange to a calculated length (how calculated?), then inserted the flange into ADL RP3-2-2B feed's Ku-band port, and mounted Invacom Universal Quatro LNB on that flange?

Yup. The Superdish flange is acting as a waveguide so the length isn't very important. I chose a length that would give easy access to all the components and not cause anything to bump into anything else.

How did you fix the flange inside the feedhorn with screws - are they included in the feedhorn assembly?

The ADL feed has a couple of set screws to hold the standard Ku waveguide and Ku flanges (first picture). These lock my alternate waveguide just as well.

Was your Scalar Ring included with the feedhorn?

Like most feeds, the ADL comes with a scalar ring. To be perfectly honest, I butchered the original scalar to match the ADL to the 3m where it was first installed because that dish is very deep. The 1.8m has a more normal f/D and doesn't require this modification. As I had an extra scalar lying around, I used that instead of the hacked one. It looks nicer.

What makes Norsat C-band LNB better than others in your view?

The phase noise is very low.

It doesn't hurt that the frequency stability over the temperature range is tight, meaning that I can read out frequencies directly on my spectrum analyzer and know they are accurate. This isn't important for most people. The Norsats also have a low equivalent noise spec that is probably accurate, unlike many of its cheaper competitors.

 

[zamar23]

I butchered the original scalar to match the ADL to the 3m where it was first installed because that dish is very deep. The 1.8m has a more normal f/D and doesn't require this modification.

How would one know, what scalar ring size and profile matches type and spec of a given dish? Is there a way to calculate for an FTA fan (as opposed to a scalar ring design engineer), how to modify a scalar ring to match a given dish, and to find out if any such modification is at all required?

 

[pendragon]

How would one know, what scalar ring size and profile matches type and spec of a given dish? Is there a way to calculate for an FTA fan (as opposed to a scalar ring design engineer), how to modify a scalar ring to match a given dish, and to find out if any such modification is at all required?

Just use the one that comes with the feed you buy. They are pretty much all the same. The important aspect is setting f/D, which is usually nothing more than adjusting how much the feedhorn mouth extends beyond the scalar rings. For example the Fortec 1.8m (the one I have) has a f/D of 0.38. For this dish the mouth needs to extend by about 0.4" beyond the scalar rings. Once this is set it is normally best to leave it alone. Unfortunately some people adjust the focal length distance for their dishes by sliding the feed in and out of the scalar rings. Unless they're lucky, they will set the wrong f/D.

 

[zamar23]

For example the Fortec 1.8m (the one I have) has a f/D of 0.38. For this dish the mouth needs to extend by about 0.4" beyond the scalar rings.

How did you find that 0.4" number - by measuring distances or any calculation?

 

[pendragon]

How did you find that 0.4" number - by measuring distances or any calculation?

By modeling (calculation). There is not a closed solution, but there are approximations. If you look at Invacom's ADF-120 specs, they have a curve for the feed mouth extension vs. f/D, but that's Ku.

 

[zamar23]

There seems to be a number C/Ku band prime focus dish LNBFs on the market, and more advanced specs of newer models are popping up periodically. Yet people who tried them claim mediocre results for Ku-band reception regardless of C-band dish size. You are using a completely different setup, and claim to achieve pretty good results in both C and Ku bands with it on a prime focus 1.8 m dish. How would you explain that? Is it a lower qualification of some CN manufacturers offering these products, lack of user end software and precise recommendations to mount and adjust these C/Ku-band LNBFs properly, narrower specs of military grade US made sat products, or...what? Could you suggest a handy way for an average Joe to achieve good results in both C and Ku band reception with one of consumer grade C/Ku-band LNBFs on a prime focus C-band dish?

Also, you mentioned that somewhat similar results can be achieved with a lot cheaper consumer grade products on the market today. Can you give a low cost equipment models list example that in your view looks like a good cheaper replacement for your described setup?

 

[pendragon]

There seems to be a number C/Ku band prime focus dish LNBFs on the market, and more advanced specs of newer models are popping up periodically. Yet people who tried them claim mediocre results for Ku-band reception regardless of C-band dish size. You are using a completely different setup, and claim to achieve pretty good results in both C and Ku bands with it on a prime focus 1.8 m dish. How would you explain that?

My other C-band dishes are larger (2.3, 3 and 3.2m) and they all use single ortho feeds and Norsat 8115s. Those are by far my best performers. The 1.8m has a dual ortho that necessarily has some losses by comparison. Nevertheless it gets considerable use because it is the first dish to be committed. I hold the others in reserve for feeds it cannot do, but those are more the exception than the rule.

Ku is easy on a 1.8m as long as the surface deviations are within reason. My 1.2m has enough gain on Ku to do just about anything, and an extra 2-3 dB with the 1.8m is simply icing on the cake.

Putting the dual ortho C/Ku-universal on the 1.8m was a winning move for me because I added a second high gain Ku dish and Ku universal capability without having to put up another dish. The only cost was a slight loss in C-band performance compared to the single ortho feed that had been there before. This is probably a good compromise for those who can only put up one dish, and certainly much better than a mini-BUD. For those without such constraints, I would still recommend a larger BUD dedicated to C-band and something like a 1.2m dedicated to Ku.

My 1.8m results can be chalked up to good parts and a lot of attention to little details that add up. One of the less obvious points is I only use the dual ortho on the smallest dish because I don't want ANY losses on the bigger ones.

Is it a lower qualification of some CN manufacturers offering these products, lack of user end software and precise recommendations to mount and adjust these C/Ku-band LNBFs properly, narrower specs of military grade US made sat products, or...what?

Better parts cost more to make and one often pays a premium because fewer people buy them. Norsat 8115s are not military grade. In fact they aren't even commercial grade. But they are good performers for the price and mine were made in Korea, for whatever that is worth.

Could you suggest a handy way for an average Joe to achieve good results in both C and Ku band reception with one of consumer grade C/Ku-band LNBFs on a prime focus C-band dish?

Also, you mentioned that somewhat similar results can be achieved with a lot cheaper consumer grade products on the market today. Can you give a low cost equipment models list example that in your view looks like a good cheaper replacement for your described setup?

This forum is a good place for people to provide information on the whole spectrum of FTA. I am only one data point and probably an outlier at that. I am a cheapskate in many respects, but not on quality. When I invest time into something like this, I expect it to last at least 10-20 years. Even though my time is 'free', something that saves me a few bucks today but keeps taking my time over and over again is no bargain.

I can describe and recommend components and approaches I have tried, but I am reluctant to even make suggestions on equipment I have not tested, particularly when I have concerns about its performance. Rather than having me speculate, you need to get that information from others with direct experience. Knowing your requirements AND budget will allow us to identify what might work best for you, rather than simply describe what each of us did. I'm not trying to sell you on duplicating my system.

 

[zamar23]

Thanks. Well, no hardware models list is also a statement of sort.

 

[pendragon]

I think we're going a little off-topic for this particular thread on a higher-end mod. Doing this on the cheap with completely different parts might be possible, but the approach would not be the same. I could offer some armchair-quarterback-type advice if I saw some pictures and people were willing to butcher their components into possible oblivion, but as I already have a working system, I won't be pursuing this on my own.

In terms of configuring a cost-effective C/Ku combined system, that's probably more apropos to your current thread. It might be helpful for you to specify exactly what you want to accomplish, your current concept and maybe an idea of what level of budget you are working with. As I last understood, you wanted one dish to do C-linear and Ku-linear for FTA and Ku-circular for subscription, but I'm kind of lost as to the receiver configuration and switching you want to do. I might be completely off-base as your thinking may have evolved over the variety of possibilities that have been discussed.

 

[zamar23]

Yah, just trying to gather thoughts on this very topic. I have an interesting quality of attempting to understand a subject better than average, and then offering a totally different approach to a relevant (but not necessarily the same) task at hand.

The main reason to ask these questions instead of "nice dish" exclaims is, I feel you might be able to answer them (!), and setting up a good working C/Ku-band set does require understanding of these concepts to avoid pointlessly moving randomly picked up parts around to no avail, as often reported. Going back to your "higher-end mod", can you offer a thought, why suggested as an alternative QPF-031 LNB appear to be "L" type packaged compare to "I" type QTF-031 LNB? Also, your LNBF assembly looks a bit heavy for that dish LNBF support struts set - is it? How big part of the dish surface it shadows?

 

[pendragon]

Going back to your "higher-end mod", can you offer a thought, why suggested as an alternative QPF-031 LNB appear to be "L" type packaged compare to "I" type QTF-031 LNB?

In purely mechanical terms, the QTF-031 fits this application very well. It's small and it hides completely behind the scalar, thus causing no additional shadowing. It does not interfere with the C-band LNBs at all. I'm sure its plastic case is only cosmetic; not that it matters much, but I'd prefer a rugged metal box like the Norsats with the LNB outputs on the back.

The QPF-031 appears to be based on the QPH-031 and therefore probably has a larger case. I have a QPH-031 on my 1.2m and I don't see any reason why the QPF-031 wouldn't work as an alternative to the QTF-031. The C-band LNB mounting on the dual ortho is such that the QPF-031 would simply jut out in the opposite direction. Unlike the QTF-031, it will extend beyond the scalar's shadow (but see below).

Also, your LNBF assembly looks a bit heavy for that dish LNBF support struts set - is it?

The Fortec arms look flimsy, but once bolted to a scalar, they are quite rigid and strong. I previously had a single ortho on this dish and ended up using the same scalar ring for the dual ortho. No doubt the dual ortho is heavier, either with the original Ku waveguide and two LNBs or the QTF-031 (which is lighter than the original). I have a homemade laser jig for aligning C-band feeds and the additional weight had no effect. I've swapped the single and dual orthos in the scalar and they both end up pointing exactly the same without any adjustments.

How big part of the dish surface it shadows?

My single and dual ortho feeds, with or without the mod, essentially only shadow the dish with the scalar ring. This causes a negligible loss, even on a 1.8m. How much is negligible? Well, a 1.8m dish has a surface are of about 2.5 sq meters. A typical scalar ring has a surface area of 0.02 sq meters. Scalar shadowing will cause a loss of only 0.035 dB. Any prime focus dish of this size will suffer at least this same loss because of the scalar, regardless of the feed.

I'm estimating the less than ideal extension of the QPF-031 will occlude an additional 0.008 sq meters, causing a further loss of 0.014 dB. Even a tiny misalignment error, that would require fairly sophisticated instrumentation to measure, will result in a larger loss than this.

 

[zamar23]

I have a homemade laser jig for aligning C-band feeds and the additional weight had no effect.

I'm estimating the less than ideal extension of the QPF-031 will occlude an additional 0.008 sq meters, causing a further loss of 0.014 dB. Even a tiny misalignment error, that would require fairly sophisticated instrumentation to measure, will result in a larger loss than this.

Thanks. Could you be so kind to share your laser jig design and the method to use it, including a couple of pics? As far as assessing shadowing the dish, I take it, you just subscribe estimated dish and component surfaces, assuming signal from a sat to be parallel to, and shadowing components coaxial with the dish axis? How you then translate it to signal loss estimates - using what formula?

 

[pendragon]

Thanks. Could you be so kind to share your laser jig design and the method to use it, including a couple of pics?

Nothing fancy. I simply drilled some holes in a block of wood that fits over a C-band feedhorn mouth and can rotate. On the other side I drilled a hole for a cheapo laser pointer. Laser pointers don't shine precisely on-axis, so the hole was drilled to accommodate shimming. While rotating the block on a feed I shimmed the pointer so the spot stayed in the same place. That's called calibration

As far as assessing shadowing the dish, I take it, you just subscribe estimated dish and component surfaces, assuming signal from a sat to be parallel to, and shadowing components coaxial with the dish axis? How you then translate it to signal loss estimates - using what formula?

From 22,000 miles away, the rays from the bird are going to be pretty close to parallel. To calculate a gain or loss for two surface areas A & B and express it in dB, just calculate 10 * log10 (A/B). In this case B would be the area of the dish and A would be the area of the dish minus the area of the shadow.