Troubleshooting Signal Loss With 5G Bandpass Filters Attached

[el bandido]

I have a pair of Norsat 8000RI lnbs attached to Eagle stage 2 5G bandpass filters. These lnbs are attached to an old Radio Shack multiswitch to feed the fta receivers. Recently I lost about 3-4 db(snr) on the horizontal side for no apparent reason. Finding the cause of the signal loss is pretty straightforward, but is complicated a bit with the 5G filters installed in the system.

Adding the 5G filters creates a loss that cannot be recovered in the range of 1-1.5 db(snr). The 5G filter loss is hust something that you have to live with. A 5G bandpass filter is basically a piece of metal that allows certain signals to pass while rejecting other signals, hence the name bandpass. To go further, a bandpass filter could be considered to be a tuned piece of metal, so it also has to be checked anytime there is an unexplained loss of signal in the antenna system.

For signal loss, start with the basics of the coax feed. We can almost immediately rule out the main run of coax because the vertical side of the antenna system seems to be working correctly. Installing a barrel connector to eliminate the Radio Shack multiswitch shows individual testing of both horizontal and vertical polarities to be about the same with the multiswitch installed. This test sort of eliminates the main run of coax and the multiswitch.

There are two short runs of coax between the lnbs and the multiswitch. These two runs were swapped on the lnbs and the two polarities rested individually again with no apparent changes and with the horizontal loss still present. At this point, we can almost 100 percent rule out a coax or connector problem.

Current draw will usually fall or become noticeably higher when a lnb is sick or failing. A voltage/milliamp meter was installed to test voltages from the receiver and current draw of the lnbs. The vertical and horizontal voltages are correct, with a little over 18 volts for horizontal and a little over 13 volts for vertical.

The vertical lnb showed 111 milliamps of current draw and the horizontal lnb showed 110 milliamps. The receiver was rebooted in both milliamp tests, with the vertical lnb holding at 111 milliamps and the horizontal lnb dropping to 85.3 milliamps. Both lnbs were tested on the same 18 volt side, which is horizontal in the receiver. At this point, it seems the horizontal lnb is sick, but you have 5G filters installed which means more work is needed.

Testing at this point would be complete and we could say with a certainty the horizontal lnb needs to be replaced. But what if the lnb isn't bad and the problem is really the filter? What will need to happen next is the lnbs will need to be removed from the dish and swapped. Move the horizontal lnb to the vertical lnb location and move the vertical lnb to the horizontal lnb location. The 5G filters should remain "as is" and not be moved. Then the lnb assembly needs to be installed back in its original position and retested.

Once the lnbs are swapped, the signal loss should follow the lnb (if it is bad) with the vertical side of the antenna system now having the signal loss and the horizontal side of the antenna system operating normally. If nothing changes after swapping the lnbs, and the horizontal side still has signal loss, then the 5G filter on the horizontal side most likely has a problem. Then and only then should the 5G filters be swapped and testing should show the signal loss following the location of the sick 5G filter which is now on the vertical side of the antenna system.

So at this point I am down to swapping the lnbs, which means removing the feed from the dish, unbolting and swapping the lnbs then reinstalling everything as it was. It is very important to mark things before disassembly or removal lest you get confused about the original location of the parts Finding the source of the signal loss becomes more complicated with the 5G bandpass filters installed, but it is doable by the process of elimination.

This is a busy time of year for me, but hopefully I can get the lnbs swapped and tested in the next day or two. Extended band lnbs are useless with 5G filters installed because only 200 MHz of band is usable. These lnbs have a l.o. of 5750 which means I will need to buy 2 lnbs if one is bad, or buy another basically useless 8000RI.

 

[el bandido]

I am not used to having dish problems, and fortunately this little problem solved fairly quick.

The feed on the horizontal side has a little compartment or cavity for the probe. This little cavity filled with water and cause the signal loss. As another test, I decided that the feed should be turned 180 degrees which would make the vertical lnb horizontal and the horizontal lnb vertical. Water started dripping out of the port hole when the lnb was rotated and the port hole was facing straight down. Problem found!


The feed was removed, then the horizontal bandpass/lnb was removed. Some water was still in the bottom of the cavity.


A small hole was drilled in the bottom of the cavity on each side of the probe. At least the water or moisture will have an exit if this happens again. The small porthole was filled with gasket sealer which may or may not stop the leak, but it should at least slow the leak down in a future storm.


We had a bunch of rain from Helene along with a bit of wind. Obviously the feed is damaged a bit else it would not leak! Maybe the feed will be replaced at some point in the future. Signal in db(snr) on the horizontal side went back up after this repair. The dish and feed could stand some adjusting as it has been a while since anything has been moved.

Horizontal signal Today on 101w Before repairs:



Horizontal signal Today on 101w After repairs:



I am very fortunate as this is the only thing I could find wrong after a devastating Hurricane for parts of the South!

 

[Terryl]

??? One question...How did you measure the DC current draw without disturbing the signals from the LNB???

When I worked in the microwave world we had to use a special device to measure the DC current on the coax going out to the antennas, this to keep from affecting the signals.

 

[ArloG]

Let's see what EB comes up with. For me I just cut back the outer shield and made a pigtail out of it a bit short of the F connector. Stuck the stinger in to make contact with the lnb connector center pin. And measured current draw with my multimeter from the braid to the outer F connector barrel.
I didn't care about signal. Just current draw. But surprisingly (expected too) I got a signal to the receiver. Because the stinger is the RF carrying element in an unbalanced cable scenario.
I was in a kick to use a power inserter to supply the lnb's after a diseqc/multiswitch to keep receiver power demands low. Measuring using the same test jig, receiver current draw was in the 50 mA. range (at the switches) with the lnb's (power inserter supplying the pixies at a constant 19 VDC) needing substantially more. One being a PLL and the other a DRO.
A cool receiver is a happy one.

 

[Terryl]

Let's see what EB comes up with. For me I just cut back the outer shield and made a pigtail out of it a bit short of the F connector. Stuck the stinger in to make contact with the lnb connector center pin. And measured current draw with my multimeter from the braid to the outer F connector barrel.
I didn't care about signal. Just current draw. But surprisingly (expected too) I got a signal to the receiver. Because the stinger is the RF carrying element in an unbalanced cable scenario.
I was in a kick to use a power inserter to supply the lnb's after a diseqc/multiswitch to keep receiver power demands low. Measuring using the same test jig, receiver current draw was in the 50 mA. range (at the switches) with the lnb's (power inserter supplying the pixies at a constant 19 VDC) needing substantially more. One being a PLL and the other a DRO.
A cool receiver is a happy one.

Thats not a real measurement for DC current on an active RF circuit, I used a modified receiver, I found the DC voltage injection circuit in it and cut the trace, I then bridged that circuit with a precision 1 ohm 1 watt resistor, then used an RF millivolt meter to measure the DC voltage drop across that resistor, ohms law provides the rest.

 

[ArloG]

Thats not a real measurement for DC current on an active RF circuit, I used a modified receiver, I found the DC voltage injection circuit in it and cut the trace, I then bridged that circuit with a precision 1 ohm 1 watt resistor, then used an RF millivolt meter to measure the DC voltage drop across that resistor, ohms law provides the rest.

I used the easy method. I would have said to power the lnb with a fixed voltage/current source. Measure OCV at the coax into the lnb. Then measure again with it connected using a jumper exposing the stinger and braid. Check voltage drop. Then Ohms is your uncle.
Then again. Do we really care about every single microamp? Doing so at the receiver would include drop over the coax run. DeVry 101.
Looking for pure steady state current draw of the lnb regulator and RF stages. Not superimposed included RF.

 

[el bandido]

You may do a current test of fta gear many different ways. We are not building a nuclear bomb and most of us do not have thousands of dollars in test equipment that would be found in a radio shop. The idea is to get a decent idea of how much current is being drawn, then if the current changes much, there is usually some sort of problem.

I purchased a cheap combination voltage/ammeter some years ago and wired it into a short jumper. The little meter is excellent for finding bad lnb and switch ports. Plus it helps give an idea of how everything is running. I think the meter I got was around 2 dollars when I got it, but they are more than that now. Here is one Ebay example:


Here is a picture of the meter crudely wired into a short cable:


The thing may not be 100% accurate and I don't need it to be. But I have a good idea of how much current my lnbs and switches draw. I also have a good idea of how much current the entire antenna system needs to operate. This little meter has been a great tool for a number of years.

 

[el bandido]

Another challenge with 5G filters installed is Tuning. I find it much easier to fine tune a pair of lnbs as compared to a pair of filtered lnbs. The The lnbs, filters, and feed weigh close to ten pounds which makes tuning awkward. Anyway, I found another db in signal and can probably get a tad bit more if time is spent tuning. Maybe I can borrow a bucket truck sometime soon and do the tuning job properly.

 

[ArloG]

Cool. I take it you mean physical tuning on the dish and not the individual slugs in the filter?
You could always go Iron Man! lol.

 

[el bandido]

Just tuning by moving the lnbs on the dish. Tuning the filters would require equipment that is normally found in a Motorola type radio shop.

 

[ArloG]

Figured so. My current setup let's me get the elusive 115W which I never was able to snag. Knew it was 'there', just never could get the slightest hint of signal.
I like your way of using the tinysa. Never really thought of using my own ancient HP network analyzer. Probably due to being too focused on a 75>50 ohm impedance match.
I've been playing around with my SDRPlay RSP2 and the spectrum analyzer application for L Band satellite aero stuff. The refresh rate is pretty decent when extending the bandwidth.
Oh. I'm on the very fringes of 115W looking at footprint maps. Anyway. Good luck with the fine tuning there Ironman. Or getting that bucket truck.