SGQ-450D Duplexer - Tune-up and Review

[marcspaz]

Hey folks!  For people looking for an affordable duplexer that works well, I may have a solution for you.

 

For those of you who have been following the thread, you may be aware that our friend @LeoG had a problem with a duplexer in his B-Tech repeater.  They exchanged it for him, but I offered to take a look at the duplexer, hoping that we could get a spare available, or a second for another machine.  I have to say, I am impressed.

 

The duplexer is a SGQ-450D.  It is a traditional six cavity mobile duplexer, made in China, and the cost is about $115-$120.  When I first got the duplexer, I connected it to a VNA.  The unit looked like it was tuned perfectly, the notch and SWR were great... it was just on the wrong frequency, preventing it from working on GMRS.  I connected it to a radio and a dummy load, checking both sets of filters, and it worked perfectly.

 

The downfall with duplexers setup like the ones installed in the B-Tech repeaters isn't the duplexer itself, but rather the tuning.  What companies do to make these wideband (and what I found with this duplexer), they tune the 3 receive cavities to slightly overlap instead of perfectly align. Also, the 3 transmit cavities to slightly overlap instead of perfectly align.  What you end up with is about 500KHz-700KHz of filtering at about 45dB-50dB on each side, for a max of about 100dB of isolation.  While not terrible, it leaves a lot of room for improvement, which is easily accomplished if you align all 3 cavities to work on the same frequency.

 

@LeoG asked to have the duplexer aligned to 462.600/467.600.  On the high side, I was able to get a minimum of -85dB notch centered at 462.600MHz.  On the low side, I was able to get a minimum of -87.4dB notch centered at 467.600MHz.  That is better than -172dB of isolation... almost double what the previous tune was.

 

Here are the final tune-up results.

Isolation = > -172dB

Transmit SWR = 1.08:1

Receive SWR = 1.18:1

Loss due to SWR = 0.0064dB

Total Insertion Loss is <1dB

 

For a test, I hooked it up to my portable repeater while at my house.  I got the same range as with my personal duplexer.  For comparison, these specs are on par with my SinTech/Sinclair duplexer, which is priced 4 times higher than the SGQ-450D.  The duplexer will be shipped back to Leo next week.  I am looking forward to his real-world feedback after the tune-up.

 

Below are a few images of the duplexer and the tuning results.  

 

 

 

 

 

 

 

 

 

[WSIK532]

@marcspaz Great Job! i know @LeoG will be excited!

[WSIK532]

@marcspaz How long did it take to tune it?

 

[marcspaz]

@WSIK532 the entire process from setup and initial testing to the finalized range test and cleanup was about 2 hours. The actual tuning itself only takes a few minutes. 

[LeoG]

Yep, I can't wait to do a real world test when I put it back in the RPT50 Btech.  Right now I have a wide band tuned duplexer in it that they sent as a replacement.  And this duplexer worked much better than the one I gave Marc to tune.  As soon as I swapped out I noticed a marked improvement in reception.  Not sure how it's tuned, if it's the same tuning as the one I gave to Marc or something else.  It's probably tuned better than the original just from the fact I could get better reception out of it.

I've done a lot of real world testing with my current setup.  I know where I can hit it and where I can't and where it's iffy.  It's the iffy spots I'm hoping to see better results.

And Marc I just want to say it so the people here see it too.  Thanks for doing this.  I appreciate your effort and I'm looking forward to putting it to good use.  I know you just wanted to play with your toys, but it's hopefully going to be a real world difference for my repeater.  This way we can both have our fun.  And just think, now I'll have that other duplexer.....

[WRYZ926]

Outstanding! 

We look forward to your test results. As mentioned by many of us, a duplexer tuned for all of the GMRS repeater channels will not work as well as one tuned for just one channel.

[LeoG]

Never expected it to.  It's absolutely a compromise.  But what it does is allow you to figure out what channel might work best for you in your area.  And when you figure that out you can commit to a tuned duplexer.

Of course when I first started out I didn't know squat about any of this stuff.  And that was a little bit more than a year ago when I got a pair of HTs.  But I'm a quick learner and generally don't need to ask questions because I can find answers myself.  But duplexers are simple and complex at the same time and it took many vids before I understood how they worked.  I always knew what they did.

I'm excited to see if it'll improve my reception.  We have 2 repeaters nearby.  Commercial repeaters all gussied up with the best of the best.  And you can tell by how they receive, very sensitive.  While mine being much closer wasn't that sensitive.  Kinda frustrating.  Hoping for a similar sensitivity with the better tuned duplexer.  Of course the commercial ones have the advantage of being 700ft+ in the air while mine currently sits at 40ft.

[marcspaz]

9 hours ago, LeoG said:

And Marc I just want to say it so the people here see it too.  Thanks for doing this.  I appreciate your effort and I'm looking forward to putting it to good use.  I know you just wanted to play with your toys, but it's hopefully going to be a real world difference for my repeater.  This way we can both have our fun.  And just think, now I'll have that other duplexer.....

 

Thank you for the kind words.  It was my pleasure!  I'm glad I was able to help.

[LeoG]

Any chance you can enlighten us on the equipment you used and the techniques to tune it in this particular instance?  Details like brand names and models of the equipment so I can looke them up and see them type of details.  Unless you don't want to give up your secrets LOL.

[marcspaz]

Sure! On the way out right now, but I'll share what im using for this job. Its easy to do and the tools are very affordable. 

[LeoG]

Hmmm just thinking out loud.  If the claim of .25µV 12dB SINAD with the wide band duplexer is true is there going to be any gain on those number with the single band tuning on the new duplexer?  And any guesses on how much?

All I can do is excitedly speculate until it gets back into my possession. 

[WRYZ926]

7 hours ago, LeoG said:

Never expected it to.  It's absolutely a compromise.  But what it does is allow you to figure out what channel might work best for you in your area.  And when you figure that out you can commit to a tuned duplexer.

 

Nothing wrong with that plan at all. 

[LeoG]

I doubt Btech thinks that's a plan.  They should mention it and might be able to sell an extra duplexer with each unit.

[marcspaz]

3 hours ago, LeoG said:

Hmmm just thinking out loud.  If the claim of .25µV 12dB SINAD with the wide band duplexer is true is there going to be any gain on those number with the single band tuning on the new duplexer?  And any guesses on how much?

All I can do is excitedly speculate until it gets back into my possession. 

 

That's a good question, but not the way receivers get rated.  The sensitivity is the 'receiver' sensitivity. 

 

A couple points about how the duplexer work and why the receive sensitivity isn't impacted.  The duplexer filters out the receive frequency on the transmit side, and the transmit frequency on the receive side.  There is no direct impact to the signal due to tuning.  The only real 'reduction' would be the overall reduction in signal from insertion loss.  However, you may see a difference/improvement due to the filtering for another reason.

 

The cavity filters have a bit of a floating echo effect, with the dB of suppression wildly varying.  Your duplexer, for example, floats between -85dB and -110dB per cavity pair.  With a varying isolation of -170dB to -220dB.  Prior to tuning, the duplexer isolation floated between about -45dB and -70dB.  This can impact how much desense the receiver experiences, giving you the impression that receive sensitivity improved. 

 

To summarize, while the receive sensitivity doesn't improve with improved duplexer tuning, the improved filtering removing more of the transmitter signal and reducing desense, will allow the appearance of improved sensitivity.  

[LeoG]

That's what I meant

[LeoG]

I guess the receiver sensitivity is what it is, but if you are able to receive better with the new duplexer that's real world sensitivity not theoretical receiver sensitivity. 

I get exactly what you are saying.  And I'm pretty sure you know what I mean too.  If I could hit the repeater at 5 miles and not 6.  Then install the new duplexer and can hit the repeater at 7 miles but not 8.  I would consider that an increase of sensitivity.  But in reality it's the duplexer allowing the receive signal to still be heard while the repeater is re-transmitting and not causing desense.

[marcspaz]

To answer question about tools used and how to adjust a duplexer, here is a quick overview.  

 

Tools I used:

• Windows PC
• SAA-2N VNA
• VNA View software
• Box Wrench
• Screw Driver
• Brass Brush

To tune the duplexer, you have to follow a couple of steps.  

1. Calibrate the VNA for the frequency you are testing (via the PC or on the VNA, depending on if you are using the PC or not)
   • I typically set the center frequency to the frequency I am tuning for, with a 10MHz spread.
2. From there, we are going to tune either the low side or high side, separately.  For this example, we can start on the low side.
   • Connect the S11 cable to the "antenna" port.
   • Connect the S21 cable to the "LOW" port.
   • Put a dummy load on the "High" port.
     • Refer to Image 1.
3. Set the Trace 1 format to S21 Thru. 
   • With a mobile UHF cavity duplexer, you can disable the other traces, but if you want to track your SWR at the same time, set Trace 2 format to S11 SWR, and then disable the remaining traces.
   • Adjust your scaling so you can see at least -110dB
4. Adjust each adjustment screw to create the deepest notch possible, to the transmit frequency. In this example, 467.600MHz.
   • only adjust one tuning screw at a time, 
   • loosen the jam nut with the box wrench, just enough to allow the tuning screw to move, and while making adjustments, hold the jam nut with the wrench.
   • Typically, once you set the screw to the proper depth, it can be a bit tricky to set the jam nut without turning the screw a bit at the same time.  A trick I learned is that I put the screw just a few degrees off, counter clockwise, so as I tighten the jam nut, it snugs the tuning screw into the correct location.
   • See Image 2 and Image 3
5. From there, switch the S11 cable to High, and the dummy load to the Low side.  Repeat the tuning steps for the high side, but tuning the notch to the receive frequency.  In this example, 462.600MHz.

 

This guy has a great video on using a VNA to adjust a duplexer.  I have no idea who he is, but it it's good content.  The only thing I would add is, he stated he is using 200 sweep points.  I would use the highest sweep setting you can that provides a good enough resolution of the frequency spread you are tuning.  On the VNA itself, I use 401 points, but on the PC, I use 1024 points.

https://www.youtube.com/watch?v=TbEK4v_3Xuo

 

With this duplexer, specifically, the tuning screws are brass.  Either the manufacturer or B-Tech spray painted the tuning slugs.  I had to spend a bit of time cleaning the paint off the screw threads and heads with a brass cleaning brush to make it so the duplexer could be tuned.  If you have a situation like this and want to repaint the screws for cosmetic reasons, be sure to use non-metallic paint, such as Krylon Fusion for Plastic.  It must be non-metallic, polyurethane free, and lead free.

 

I am including a few extra pictures displaying the results.

 

Image 1

 

 

Image 2

 

Image 3

 

 

 

 

 

 

 

 

[LeoG]

You're as bad as me for staying up late.  LOL

Ya, I saw how they painted the screws.  Shot them from one side to help lock them in I suspect.  Didn't think about how that might impact re-tuning though.

Thanks for the tour.  Do you know how to read the circular graph?  The curly cue lines are intriguing, really curious how they are read.

[Lscott]

I've held off spending money on one of the tiny VNA's since I'm not really convinced they have the dynamic range claimed. I would like to see some verification of the spec's testing against a know high quality lab bench VNA.

For tuning these cavity filters you really need the high dynamic range to be sure you have all of the cavities aliened. 

[marcspaz]

46 minutes ago, LeoG said:

You're as bad as me for staying up late.  LOL

Ya, I saw how they painted the screws.  Shot them from one side to help lock them in I suspect.  Didn't think about how that might impact re-tuning though.

Thanks for the tour.  Do you know how to read the circular graph?  The curly cue lines are intriguing, really curious how they are read.

It's a Smith Chart SWR Circle.  It's a circle that is drawn centered at the origin frequency, passing through the plotted impedance point. This circle represents a constant SWR through the range.  The value where the SWR circle intersects the positive real axis (right side) of the chart indicates the SWR.  The closer the red dot is to one of the lines radiating from the right side, the closer to 1:1 the SWR is on that frequency.

 

37 minutes ago, Lscott said:

I've held off spending money on one of the tiny VNA's since I'm not really convinced they have the dynamic range claimed. I would like to see some verification of the spec's testing against a know high quality lab bench VNA.

For tuning these cavity filters you really need the high dynamic range to be sure you have all of the cavities aliened. 

 

I tried the TinyVNA and it did not work.  It was way off... very inaccurate.  A friend of mine (who is a retired RF engineer) uses a SAA-2N VNA when he is in the field.  He has great results with it, which is why I went with that specific VNA.  It's "good enough" that I would not encourage people who are doing this as a mild hobby to buy a commercial Signal Analyzer.  If professionals are using it, the SAA-2N VNA is good enough for me.

 

I have not compared mine to a commercial grade SA, but I have tuned 5 duplexers (UHF and Commercial VHF) with the SAA-2N VNA since I bought it and have had amazing results.  With a 2m repeater only having a 600KHz offset, accuracy is extremely critical.  My 2m repeater is covering over 8 miles with 5w handheld radios and well over 25 miles with 50w mobiles, and the antenna is only 100' above the ground, 200 feet above average terrain.  Those numbers are both further than the theoretical/calculated RF LOS.  So, I personally have confidence in the SAA-2N VNA.

[WRYZ926]

@marcspaz thank you for the detailed information on tuning the duplexer. 

We are lucky in that one of our members retired from Sound Solutions and he still has access to all of their equipment. He has tuned all of our duplexers.

[WRUE951]

My first Duplexer from BuyTwoWay covered the entire GMRS Band and  my repeater then was on Ch 22. I had a terrible experience with it, ended up buying a tuned dublexer tuned for Ch 18 and moved there.. Been working great since..  I sent the original duplexer back to BuyTwoWay and had them tune for CH 18 to use for my portable repeater which i mounted to the outside of a FAT 50 Ammo Can.   Since this portable set up gets moved around a lot, i'm always concerned that the duplexer will get knocked off tune, but going to almost 3 years now its holding tight on Ch 18.. I did try to tune it myself as Marc did, but didn't have such luck or patience.