My First Crack At Making A Power Divider

[tcp2525]

I've been making phasing harnesses throughout the years, but never a power divider. I figured I'd give it a shot. Did all my calculations to get to where I needed to be with the materials that are over the counter. Finding the right combination of outer and inner pipes is the biggest challenge.  My target was the high end of the GMRS band so that's where I did my calculations. Of course UHF being extremely critical of measurements a few thousandth of an inch either way can be a PITA. Testing it on the NanoVNA yields just barely a 1.4:1 at 467.650 MHz and under 1.2:1 at 444.175 MHz, my most used frequencies. Now I know what to look out for and compensate should I decide to build another.  Anyway, here's the finished product.

Everything all soldered and watertight waiting on a coat of battleship gray epoxy paint.

 

Everything soldered waiting for the final top cap after testing. Two 50 ohm terminators for testing.

 

Center conductor (11mm Brass Tube 1mm wall thickness) ready for installation. The 12 gauge copper center conductor is left intentionally long and will be much short once installed.  N-connector soldered on other end.

 

Brass tube with 8-32 threaded brass nut with rounded corners pressed in prior to solder.

[JeepCrawler98]

Clean work; nicely done! how's the insertion loss on this guy - close to (I assume) the theoretical 3db per port? Do you have a link for the design formulas? Those things are kind of pricey in the commercial market!

[tcp2525]

Thanks. Here's the link to the page I used.

 

https://www.qsl.net/dk7zb/Stacking/splitter.htm

 

And here's the software. Software seems pretty good.

 

http://www.hp.woodshot.com/

[Lscott]

What is the final application for this?

[tcp2525]

Stacking two yagi antennas that will eventually go on the tower.

[WRKC935]

On 3/6/2022 at 5:44 AM, tcp2525 said:

Stacking two yagi antennas that will eventually go on the tower.

This is NOT the way to be doing that.

You need to find some 75 ohm cable and cut two EQUAL lengths of it and put the required connectors on it. 

Then you use a Tee connector fed with a 50 ohm cable.  But the two 75 ohm cables have to be the same length.  If you split the power in a divider, then only 50% of the total signal goes to each antenna.  It will maintain the correct impedance, but cuts the power in half.  When you run the two pieces of 75 ohm cable you correct the impedance mismatch that would be 25 ohms (2 50 ohm loads in parallel) by adding 25 ohms of impedance to each leg. 

And I would be looking for proper cable for this like RG11 with a copper shield and NOT TV cable.  TV cable typically will not take solder as well and is a pain to terminate.

 

[JLeikhim]

This is NOT the way to be doing that.
You need to find some 75 ohm cable and cut two EQUAL lengths of it and put the required connectors on it. 
Then you use a Tee connector fed with a 50 ohm cable.  But the two 75 ohm cables have to be the same length.  If you split the power in a divider, then only 50% of the total signal goes to each antenna.  It will maintain the correct impedance, but cuts the power in half.  When you run the two pieces of 75 ohm cable you correct the impedance mismatch that would be 25 ohms (2 50 ohm loads in parallel) by adding 25 ohms of impedance to each leg. 
And I would be looking for proper cable for this like RG11 with a copper shield and NOT TV cable.  TV cable typically will not take solder as well and is a pain to terminate.
 

Both schemes apply 50% ( less resistive losses ) power to each antenna.

Sent from my SM-T350 using Tapatalk

[Lscott]

10 hours ago, WRKC935 said:

This is NOT the way to be doing that.

You need to find some 75 ohm cable and cut two EQUAL lengths of it and put the required connectors on it.

What he built is called a "Wilkinson" power spilter or combiner. Also there is nothing wrong with building one the way he did it either. You can buy professional ones built the same way.

https://www.tessco.com/product/380-960-mhz-2-way-splitter-w-n-females-313184

The coax impedance should be 70.7 ohms for the ideal match, not 75 ohms. Using 75 ohm coax the calculated SWR assuming a pure load of 50 ohms is 1.22 : 1 under ideal conditions. Still not bad.

Also if you don't use type "N", BNC or SMA true 50 ohms impedance connectors but use SO-239/PL-259 instead which will really mess things up.

Also for a true power spilter/combiner you should have a resistor RF load between the two load ports. That's in case one of the loads fails, opens up, the source still sees a good match. This is very often left out of many designs. It still works, just isn't screw up proof.

https://en.wikipedia.org/wiki/Wilkinson_power_divider

http://www.ittc.ku.edu/~jstiles/723/handouts/Wilkinson Divider Even and Odd Mode Analysis.pdf

[tcp2525]

On 3/11/2022 at 9:58 AM, Lscott said:

What he built is called a "Wilkinson" power spilter or combiner. Also there is nothing wrong with building one the way he did it either. You can buy professional ones built the same way.

https://www.tessco.com/product/380-960-mhz-2-way-splitter-w-n-females-313184

The coax impedance should be 70.7 ohms for the ideal match, not 75 ohms. Using 75 ohm coax the calculated SWR assuming a pure load of 50 ohms is 1.22 : 1 under ideal conditions. Still not bad.

Also if you don't use type "N", BNC or SMA true 50 ohms impedance connectors but use SO-239/PL-259 instead which will really mess things up.

Also for a true power spilter/combiner you should have a resistor RF load between the two load ports. That's in case one of the loads fails, opens up, the source still sees a good match. This is very often left out of many designs. It still works, just isn't screw up proof.

https://en.wikipedia.org/wiki/Wilkinson_power_divider

http://www.ittc.ku.edu/~jstiles/723/handouts/Wilkinson Divider Even and Odd Mode Analysis.pdf

Thanks for the excellent explanation. This project is something I never done before and I love experimenting. Sure, I could have used RG11 as I have a bunch of Belden cable, but I've done that so many times it's boring. My goal is to use the POTS philosophy and build a highly reliable and indestructible divider that will resist weather for decades. Now, if I had a machine shop or access to one I could custom build everything, but this  would go against POTS and wouldn't help other readers duplicate my work should they want to make one. At the end of the day I gained a lot of knowledge, which is very important to me. Now I got to build two yagi antennas.  

[Lscott]

On 3/13/2022 at 9:12 AM, tcp2525 said:

Thanks for the excellent explanation. This project is something I never done before and I love experimenting. Sure, I could have used RG11 as I have a bunch of Belden cable, but I've done that so many times it's boring. My goal is to use the POTS philosophy and build a highly reliable and indestructible divider that will resist weather for decades. Now, if I had a machine shop or access to one I could custom build everything, but this  would go against POTS and wouldn't help other readers duplicate my work should they want to make one. At the end of the day I gained a lot of knowledge, which is very important to me. Now I got to build two yagi antennas.  

Good luck with that project.

I think in another post I had a link for a very excellent antenna modeling software. The guy who wrote it has retired and doesn't want to be bothered with support issues anymore so he just made the software FREE. I have several versions of it and it does a very good job. It does take some expertise and experience to use it but it's a life saver for those that want to do antenna design. 

Perhaps you could model the yagis you want to build before doing the actual construction.

https://www.eznec.com/

 

One other gain antenna that works well on UHF is a corner reflector. You can get some good gain out of one of these.

https://en.wikipedia.org/wiki/Corner_reflector_antenna

This is a simulation of one I did at 435 MHZ for Ham radio use. Instead of using wire elements many use a wire screen mesh to do the same thing.

 

 

[tcp2525]

10 hours ago, Lscott said:

Good luck with that project.

I think in another post I had a link for a very excellent antenna modeling software. The guy who wrote it has retired and doesn't want to be bothered with support issues anymore so he just made the software FREE. I have several versions of it and it does a very good job. It does take some expertise and experience to use it but it's a life saver for those that want to do antenna design. 

Perhaps you could model the yagis you want to build before doing the actual construction.

https://www.eznec.com/

 

One other gain antenna that works well on UHF is a corner reflector. You can get some good gain out of one of these.

https://en.wikipedia.org/wiki/Corner_reflector_antenna

This is a simulation of one I did at 435 MHZ for Ham radio use. Instead of using wire elements many use a wire screen mesh to do the same thing.

 

 

Thanks. I did download EZNEC when he announced it was going to be free. I still have to learn the program and get comfortable before cutting any metal. It's going to be a great learning experience for sure.