BTT SELECT-78 770/850 MHz Public Safety Filter

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
To avoid confusion regarding the earlier comments about pre-selection prior to the LNA, I thought it might be good to provide a block diagram. One thing to point out is that the LNA has a very high OIP3. Also, keep in mind that IP3 specs for amplifiers like this are normally characterized with two-tone test with very close input frequencies. Just like the adjacent channels you would have in any mobile subscriber radio. The pass-band for the public safety systems have 800 to 1600 channels each. It would not be practical or realistic to say that all of those desired channels have to be filtered before the LNA in a handheld radio. In a base station, it would make sense to have tuned-cavity type filters for every channel used. Carrying around 50 pounds of gear for a portable doesn't make sense. Compromise is necessary sometimes. (most times). Again, the point of this filter is to give an improved front-end to most portable receivers monitoring combined 770/850 P25 systems.
 

Attachments

  • select78_block_diagram.pdf
    9.3 KB · Views: 22

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
If the UHF band pass filter before the LNA does not filter out the 718 to 768MHz range then it would be unusable at my location and the LNA will get driven into IMD land. I'm currently using a MiniCircuits ZQLSC-100 LNA with 35dBm IP3 and a high pass filter in front of it somewhere around 700MHz and its getting destroyed by 700MHz 5G. It doesn't make sense to filter after the LNA because its the thing you are protecting from overload. I've seen a lot of filter/LNAs in my 40yr radio career but I've never seen filters used after the LNA and for good reason.

As far as a box that only passes the two narrow frequency ranges, its only useful to people who only monitor the very specific 700/800 public service bands. If there is no bypass function in the box you end up disconnecting it when you want to receive something else then reconnecting, etc. I have a lot of other freqs in my scan list along with 700/800 stuff and would not want to put something in line just for that. Maybe I'm the odd one here and everyone else only monitors 700/800, no VHF no UHF, no VHF air, no UHF air?

I think it would be more useful to design a filter feeding a filter/LNA that covers 768MHz to 1300MHz or higher and removes strong offending frequencies before the LNA and with a bias Tee so it can be placed at the antenna. You could then have a companion box for the 118 through 512MHz range to dedicate to an appropriate antenna and combine the two for continuous reception of 118 through 1300MHz or beyond with all interfering bands and frequencies removed to protect the LNAs in each box.



To avoid confusion regarding the earlier comments about pre-selection prior to the LNA, I thought it might be good to provide a block diagram. One thing to point out is that the LNA has a very high OIP3. Also, keep in mind that IP3 specs for amplifiers like this are normally characterized with two-tone test with very close input frequencies. Just like the adjacent channels you would have in any mobile subscriber radio. The pass-band for the public safety systems have 800 to 1600 channels each. It would not be practical or realistic to say that all of those desired channels have to be filtered before the LNA in a handheld radio. In a base station, it would make sense to have tuned-cavity type filters for every channel used. Carrying around 50 pounds of gear for a portable doesn't make sense. Compromise is necessary sometimes. (most times). Again, the point of this filter is to give an improved front-end to most portable receivers monitoring combined 770/850 P25 systems.
 
Last edited:

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
If the UHF band pass filter before the LNA does not filter out the 718 to 768MHz range then it would be unusable at my location and the LNA will get driven into IMD land. It doesn't make sense to filter after the LNA because its the thing you are protecting from overload. I've seen a lot of filter/LNAs in my 40yr radio career but I've never seen filters used after the LNA and for good reason.

As far as a box that only passes the two narrow frequency ranges, its only useful to people who only monitor the very specific 700/800 public service bands. If there is no bypass function in the box you end up disconnecting it when you want to receive something else then reconnecting, etc. I have a lot of other freqs in my scan list along with 700/800 stuff and would not want to put something in line just for that. Maybe I'm the odd one here and everyone else only monitors 700/800, no VHF no UHF, no VHF air, no UHF air?

I think it would be more useful to design a filter feeding a filter/LNA that covers 768MHz to 1300MHz or higher and removes strong offending frequencies before the LNA and with a bias Tee so it can be placed at the antenna. You could then have a companion box for the 118 through 512MHz range to dedicate to an appropriate antenna and combine the two for continuous reception of 118 through 1300MHz or beyond with all interfering bands and frequencies removed to protect the LNAs in each box.
lol. ok. I'll wait for your design then.
 

FreqNout

Member
Joined
Jul 23, 2021
Messages
296
Location
Chicagoland
As far as a box that only passes the two narrow frequency ranges, its only useful to people who only monitor the very specific 700/800 public service bands.
Thus the name of this filter. Select 78. The pass band is clear and has been said multiple times.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
Once I find a very sharp cutoff 768MHz high pass filter to 1300MHz I'll retrofit this box on my tower and life will be good again. It has a 30-80MHz input, a 118 to 512MHz input and currently 700 to over 1300MHz input with very high IP3 amplifiers per band, the best 88-108 trap I've found and all combined onto one feedline. There are also bypass relays that will select the filter/LNA path or straight through. It worked well until 5G went on line over the last year or so and some those levels at my antenna are in the-35dBm range.

1674276263913.png

lol. ok. I'll wait for your design then.
 

kb5udf

Member
Premium Subscriber
Joined
Apr 7, 2004
Messages
801
Location
Louisiana
BTT my immediate application would be to dedicate an Omni antenna to this filter, followed by a small minicircuits amp and a splitter for 700/800, with the goal of being able to monitor multiple/outlying LWIN sites I used to be able to hear before the 768mhz cell band ramped up.
I may also use to to allow my p25-RX to go mobile again. I was about to attempt yagi’s pointed at each site of interest, but your product may save a lot of effort and 4 more coax runs.
 

FreqNout

Member
Joined
Jul 23, 2021
Messages
296
Location
Chicagoland
I used to be able to hear before the 768mhz cell band ramped up.
I may also use to to allow my p25-RX to go mobile again. I was about to attempt yagi’s pointed at each site of interest, but your product may save a lot of effort and 4 more coax runs.
The LTE is a real problem for sure. Hard to say a yagi alone would solve your issue. The yagi would be increasing the LTE signal in that direction as well. Plenty of cell towers in all directions. Before LTE, what signal strength were you getting on these sites?
 

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
BTT my immediate application would be to dedicate an Omni antenna to this filter, followed by a small minicircuits amp and a splitter for 700/800, with the goal of being able to monitor multiple/outlying LWIN sites I used to be able to hear before the 768mhz cell band ramped up.
I may also use to to allow my p25-RX to go mobile again. I was about to attempt yagi’s pointed at each site of interest, but your product may save a lot of effort and 4 more coax runs.
Please contact me via email.
 

kb5udf

Member
Premium Subscriber
Joined
Apr 7, 2004
Messages
801
Location
Louisiana
It has now been several years FreqNout since the outlying LWIN sites immediately surrounding the Scott, LA site were usable at my QTH. I never took exact measurements other than that I can say long ago before 768 mhz was a problem I could receive these adjacent sites sufficiently well with cheap SDR dongles and decoding with dsdplus. Since I was getting good digital decodes, I never much paid attention to the signal level. At some point reception of these outlying sites ceased and I just wrongly assumed they had changed patterns or decreased power. Wasn’t till much later I look at the spectrum on an air spy with SDR sharp and saw the problem. Also, I did test the yagi pointed at my main site of interest and it was beneficial, because the main cell tower troubling me is off to the side of the yagi, and this helped with reception of some of the lower frequency channels on this particular site that had been garbled at times.

BTT I am happy to email you but I don’t have your address. PM sent to your with mine. Ok disregard I found email on your website, email sent shortly.
 
Last edited:

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
It worked well until 5G went on line over the last year or so and some those levels at my antenna are in the-35dBm range.
At those levels, I wouldn't expect issues with high IP3 amps. The LNA used in the BTT SELECT-78 has an OIP3 spec of +42.5 dBm with two-tone, delta freq measurements of 1 MHz and an output power level of +3 dBm. Third order distortion for an input level of -35 dBm input at 750 MHz would be -133.5 dBm. With a 12.5 kHz channel, that is almost exactly the thermal noise floor ( 10log10(12.5e3)-174 ) = -133 dBm.

Are you using a multicoupler in your setup? If so, you might want to try and bypass it. Some of these devices have poor IP3 specs and might be an issue at those levels.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
Its more complicated than that and the insane amount of 5G signals around here combined is a huge level. I measured one hot carrier around -35dBm but to take an average I measured 105 carriers between 718 and 722MHz at around -49dBm each and combined the spectrum would be around -29dBm. Then there are 48 carriers from 729.24 to 733.6Mhz at -39dBm each and the combined spectrum is around -22dBm. Then 94 carriers between 734.49 and 743MHz at -49dBm for a combined level of -29dBm, then 80 carriers from 743.6 to 755.5Mhz at a level of -41dBm each for a combined level of -22dBm then 100 carriers between 758.48 and 767.5MHZ at a level of -52dBm each for a combined level of -32dBm.

Plug that into your IP3 calculator and you'll see there is probably no LNA available with a high enough IP3 rating that will not produce lots of destructive IMD with that spectrum hitting it. This is why all filters should go in front of an LNA to protect the LNA. I'm using two passive diplexers to combine the 30-88, 118 to 512 and 750MHz and up ranges. All my IMD is from the 700MHz and up preamp by itself.

At those levels, I wouldn't expect issues with high IP3 amps. The LNA used in the BTT SELECT-78 has an OIP3 spec of +42.5 dBm with two-tone, delta freq measurements of 1 MHz and an output power level of +3 dBm. Third order distortion for an input level of -35 dBm input at 750 MHz would be -133.5 dBm. With a 12.5 kHz channel, that is almost exactly the thermal noise floor ( 10log10(12.5e3)-174 ) = -133 dBm.

Are you using a multicoupler in your setup? If so, you might want to try and bypass it. Some of these devices have poor IP3 specs and might be an issue at those levels.
 

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
Its more complicated than that and the insane amount of 5G signals around here combined is a huge level. I measured one hot carrier around -35dBm but to take an average I measured 105 carriers between 718 and 722MHz at around -49dBm each and combined the spectrum would be around -29dBm. Then there are 48 carriers from 729.24 to 733.6Mhz at -39dBm each and the combined spectrum is around -22dBm. Then 94 carriers between 734.49 and 743MHz at -49dBm for a combined level of -29dBm, then 80 carriers from 743.6 to 755.5Mhz at a level of -41dBm each for a combined level of -22dBm then 100 carriers between 758.48 and 767.5MHZ at a level of -52dBm each for a combined level of -32dBm.

Plug that into your IP3 calculator and you'll see there is probably no LNA available with a high enough IP3 rating that will not produce lots of destructive IMD with that spectrum hitting it. This is why all filters should go in front of an LNA to protect the LNA. I'm using two passive diplexers to combine the 30-88, 118 to 512 and 750MHz and up ranges. All my IMD is from the 700MHz and up preamp by itself.
You might want to ask yourself how modern WiFi or LTE receivers work in these conditions. They have very wide channels of 5, 10, 20, 40 MHz, or more. An old school WiFi receiver is 20 MHz wide channels with 64 sub-carriers. I've designed several industrial high-speed, high-power transceivers with channel bandwidths up to 40 MHz and hundreds of sub-carriers. In these types of devices, for example, an EVM of above 3% with QAM-64 modulation on 48 sub-carriers will result in degraded communications. I designed a 902-928 MHz, high power transceiver with a 20 MHz channel bandwidth capable of 72 Mbps data rates (64 sub-channels, 48 data channels). It works fine with a 28 MHz wide filter before the first LNA. Thanks for the input and sharing a picture of your design. I hope you figure out the issues with LTE.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
As a follow up and to simplify what my LNA is getting hit with I just stuck an HP power meter on the coax from the bypassed tower top amp but with 700MHz high pass filter in line and average power is around -20dBm in the shack. With about 3.3dB feedline loss at 825Mhz from the shack to the tower top filter/amp that would be more than 17dBm of aggregate power hitting the preamp with hundreds and hundreds of carriers and the some of worst offenders are in the 718 to 767.5MHz range. How will your BTT SELECT-78 handle that with what sounds like a broad front end filter that does not seem to address the 718 to 767.5MHz range?

These high RF levels can't be isolated to just my neighborhood and I suspect its higher at other locations all over the country.

You might want to ask yourself how modern WiFi or LTE receivers work in these conditions. They have very wide channels of 5, 10, 20, 40 MHz, or more. An old school WiFi receiver is 20 MHz wide channels with 64 sub-carriers. I've designed several industrial high-speed, high-power transceivers with channel bandwidths up to 40 MHz and hundreds of sub-carriers. In these types of devices, for example, an EVM of above 3% with QAM-64 modulation on 48 sub-carriers will result in degraded communications. I designed a 902-928 MHz, high power transceiver with a 20 MHz channel bandwidth capable of 72 Mbps data rates (64 sub-channels, 48 data channels). It works fine with a 28 MHz wide filter before the first LNA. Thanks for the input and sharing a picture of your design. I hope you figure out the issues with LTE.
 

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
How will your BTT SELECT-78 handle that with what sounds like a broad front end filter that does not seem to address the 718 to 767.5MHz range?

These high RF levels can't be isolated to just my neighborhood and I suspect its higher at other locations all over the country.
I guess we are going to find out. It probably isn't as good as this passive filter: Comba Telecom, Inc. - Public Safety 700/800MHz Filter - FP-78-IN2 - Tessco, but it will cost a lot less and it is quite a bit more compact. I'm going to have a 4th tester in a bad location try it out in various situations. We'll see what happens there and I will report back.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
It would be nice if your testers have a spectrum analyzer. Not a dongle SDR with analyzer program but an actual spectrum analyzer that can measure carrier to noise ratio and noise floor. It would be good to know if your product improves or degrades the signal to noise ratio or raises the noise floor.

In areas where interference is not a problem I would guess it lowers the signal to noise slightly due to loss in front of the LNA and the LNA noise figure. In high RF areas it could improve reception via its filtering or further degrade it depending on the level of interference and if its driving the LNA into producing high levels of IMD.

I guess we are going to find out. It probably isn't as good as this passive filter: Comba Telecom, Inc. - Public Safety 700/800MHz Filter - FP-78-IN2 - Tessco, but it will cost a lot less and it is quite a bit more compact. I'm going to have a 4th tester in a bad location try it out in various situations. We'll see what happens there and I will report back.
 

btt

Banned
Banned
Joined
Mar 11, 2020
Messages
2,585
Location
Wa State
In areas where interference is not a problem I would guess it lowers the signal to noise slightly due to loss in front of the LNA and the LNA noise figure. In high RF areas it could improve reception via its filtering or further degrade it depending on the level of interference and if its driving the LNA into producing high levels of IMD.
I don't have absolute numbers on that for you yet. I know what it is on paper, but I would rather measure before stating it. It depends on what you are comparing it to. I suspect it will improve sensitivity on some receivers. It is the same amplifier used in the original P25RX product. It is a sensitive receiver.
 

FreqNout

Member
Joined
Jul 23, 2021
Messages
296
Location
Chicagoland
the insane amount of 5G signals around here
Could you please share what cell carrier is using real 5G on 700MHz? What is your source of that info? Not that it matters... RF energy is RF energy. You just have a lot of it at your location.
 

prcguy

Member
Joined
Jun 30, 2006
Messages
15,228
Location
So Cal - Richardson, TX - Tewksbury, MA
I would have to DF the site but its probably one I can see at 263yds from my tower as measured with a laser range finder. I will have to see if there is a company sign on the building. I'm assuming its 5G as the site has been there for 30yrs and the 700MHz stuff just went on the air within the last year or so.

Could you please share what cell carrier is using real 5G on 700MHz? What is your source of that info? Not that it matters... RF energy is RF energy. You just have a lot of it at your location.
 

FreqNout

Member
Joined
Jul 23, 2021
Messages
296
Location
Chicagoland
Both ATT and VZW use 850 for 5G nationwide (plus other higher bands). The 700 MHz Band 14 is 4G LTE. Again, just pointing out it is an RF energy issue not a 4G or 5G issue.
 
Top