Power loss through Duplexer

[wqpn591]

I'm building a GMRS repeater with two Wouxun KG-1000G units.

Doing some bench testing and noticed that when I put my Nissei RS-50 SWR meter inline with  an XLT 50W dummy load, I get SWR 1.02 and power 42W.

But when I run the signal through the XLT DP-GMRS-50 duplex and measure to the same load, I get SWR 1.0 and power 9W.

Is it customary to lose that much power through a Duplexer?

 

[JeepCrawler98]

What's the make, model and type of duplexer? Duplexers always lose some power, I've measured losses like that first hand on the Jesai, Fumei and other Chinese flatpacks around the 6-7dB range which is exactly the power drop you're measuring. They're really only acceptable down to about a 7mhz split or bigger, not the 5mhz split we're used to.

If it's something with a reputable brand name, it's not normal. a Celwave/Phelps Dodge or Telewave duplexer will have insertion losses around 1.5dB or so, if you take your time tuning those you can get that down to 0.8dB or so; if you get something like that the 42W you measured would be up around 35W after passing through...

Edited August 7, 2021 by JeepCrawler98

[wqpn591]

Its an XLT DP-GMRS-50 duplexer. Bought it with the radios from buytwowayradios.com . made for GMRS repeaters.

I guess I need to find a VNA and see if its tuned correctly.

 

[JohnE]

advertised loss is 2dB so you should be seeing 27-29 W out of it based on the 42 you are seeing out of the radio.

[gortex2]

What are you using for cables to and from duplexer ? multiple adapters and poor cable will also eat up alot of power.

[OffRoaderX]

Having built a KG-1000G repeater myself, and also having a "real" repeater, I can attest that when using one of those small mobile-type duplexers, this is not only normal, but expected.

To get save that power you will have to step up to one of the large band-pass/reject type duplexers.. For most people the $$ is not worth it, unless you can find a good used one.

[wqpn591]

Thanks for the info all.   2dB target is not horrible but can try to make up with antenna gain.

Right now I'm just testing with RG58 short jumpers into a dummy load.

Bought a nanoVNA and will try to tune this thing for low drop on my frequencies of interest.

I'm trying to make this repeater mobile so I can take to our volunteer group events and easily set it up for better coverage than we could get on Simplex.
so I'm a little limited on room for the duplexer. A better one that is a little bigger would be fine. But one of those big can units wont fit. I'll check out some of the units that @JeepCrawler98 suggested.

 

[gortex2]

A decent tuned flatpack will work. It needs tuned properly for the frequencies in use. Anything less than 3db of loss is a plus. IF your dealing with portables at an event dont focus on TX power. A 4 watt portable is not going to talk any further to the repeater. Also running lower power will help the decense and loss factors. IF the repeater can do 25ish watts on TX then after duplexer you should be 10-15. Everyone gets focused on TX power. 

[JohnE]

I will post this again as it is universal to all RF

dB loss            power loss in %
0.5                        10                        100W in 90W out
1.0                        20
1.5                        30
2.0                        37
2.5                        44
3.0                        50
3.5                        56
4.0                        60

                                                               

flat packs tune for Notch/Notch ----\___/----\___/----and should never be tuned W/power

typical loss for the ones I use are 1.5 or less most ~1.

[wqpn591]

Well, after trying to tune that XLT DP-GMRS-50 duplex with a nanoVNA, I've concluded that indeed, the filter design is just not tight enough and there will always be loss through this thing.  If the freq spread where 8Mhz or more, it would work nicely.  But at the 5Mhz spread we have in GMRS, its very hard to get low loss on both the Rx and Tx sides at the same time.

I'm going to try a better mobile duplexer from Sinclair MR3332B-3.  They tune it to order, with something better than a nanoVNA I presume. Still wont be as good at the notch-pass type of duplexer but dont want to fork over $3K+ and loose my mobile friendly system.  Will let you guys know how that one does when it comes in. 

[JeepCrawler98]

So for reference; I just got done tuning up our portable repeater's duplexer for a campout next week with the local GMRS crew - this is a Celwave 633-6A-2N mobile duplexer, measurement device is an Anritsu MT8212B. This is measured through a couple extra fittings that are part of the normal install; so this adds a few tenths of a dB for insertion loss.

For 462.700 Mhz (TX side) - insertion loss is -1.52dB, with a high notch (nothing RX) at -81.72dB

For 467.700 Mhz (RX side) - insertion loss is -1.25dB, with a low notch (notching TX) at -87.97dB

Below are comparative measurements I took a while ago of a similar Celwave unit (left) vs. the chinesium Jesai/Fumei unit (right); you can see the difference - the cheap stuff is unfortunately garbage at the 5mhz split:

Edited August 16, 2021 by JeepCrawler98

[gortex2]

AS Ive said for multiple years you get what you pay for. This is just another example of cheap china stuff that everyone wants to use that does not work well

[wqpn591]

Very nice side-by-side comparison. Exactly what I was seeing on the XLT. 

Your insertion loss is right on target and the nice deep notches........

I ended up ordering the Sinclair MR356N-2    from theantennafarm.com. will see how that does.