01/02/07

"W4XE Repeater Tech page ©"
 

I ran a split site 6 meter repeater utilizing a UHF link between them with about a 10 mile separation. Worked great, but there are always twice as many things to go wrong and keep running. The coverage overlap discrepancy wasn't too terrible and it made for a nice system. The range was 50 miles plus and really performed better than 2 meters or 440 with our mountainous terrain in Huntsville, Alabama. With the loss of our south site tower lease, we lost 1/2 of the system. This prompted us to look for a single site solution.

If you are contemplating building a six meter repeater, you'll soon start to look at duplexers for 50 MHz band. These are hard to come by on the used market. Commercial versions can be 6' tall and can be very expensive to purchase new. There are some 70 MHz versions floating around, but these are way too short. The 40-50 MHz low band commercial split look attractive. There are some notch or pass cans only out there but alot of times they don't provide enough isolation and another cavity of two is tacked on to them.

You'll soon find an excellent website:

WB5WPA's http://www.dallas.net/~jvpoll/dup6m/dup6m.html

This provides in great detail his on going decade plus saga of building a duplexer using surplus 1 5/8" heliax.

We were very fortunate that one of our club members is buddies with a guy that offered to build us up 6 heliax cavities. He had the material on hand. We just had to add some brackets and inter-connect cabling.

Six individual cavities arrived and I was soon figuring out how to mount these. I patterned it after the format of a 6 can VHF set by building a box around them using some angle bracket material from Lowes. The sections were precut, so I just added some U bolts and bolted it all together. Some black paint was used for style points. I used some foam around the U bolts so that I would not crush the tubing.

Another local ham W5RM had some 1/4" flex heliax so we made up some jumpers of the right lengths taking into consideration the correct velocity factor.

The cavities were pre-tuned before being sent to us. Initial checks on the spectrum analyzer with tracking generator showed 5 of the 6 cavities had drifted considerably in their notch freq. This duplexer design is a notch only design. Not a BPBR configuration. So the centering of the notch frequency has to be very precise.

Consulting the web, several people had mentioned these duplexers do not hold up well to movement or being mechanically banged around. Just transporting them from the lab to the repeater site would be enough to ruin the precision tuning necessary to achieve the desired attenuation and insertion loss.

It was evident that the main problem with the design was the use of a 'gimic' capacitor. This is what determines the notch frequency. The gimic cap is just a section of center dielectric from a standard piece of 1/2" coax. See photo below. It can provide a very low pF value depending on it's length. For tuning the notch of the duplexer you adjust it's length by moving it in and out of the center conductor of the heliax and solder it where when the notch is tuned. The issue arises when this gimic cap moves around inside the center conductor from mechanical movement, such as transporting it to the repeater site. You could bang on the outside of the cavity and watch the notch move about. Some people use the shield of the 1/2" coax to help retain the dielectric. I've seen others use hobby tubing to attempt to fix it's movement. It seemed like it would be better to just use an actual variable capacitor in the design rather than try to make one.  I chose a piston trimmer cap. Something like a Johanson style seemed appropriate. These are commonly used on commercial duplexers. The gimic cap is far cheaper, but surplus trimmers are available from a number of sources for $10 each. I ran an excel spread sheet calculation and determined a 1-10 pF cap would probably be an ideal tuning range.

In practice I found the 1-10 pF range was more than adequate to tune the duplexer to the necessary notch frequency. It does tune very fast, so the adjustment is a little touchy. I experimented some by padding the cap to slow down the tuning speed but I was afraid the cap I added wouldn't have a good temperature coefficient so I didn't persue that investigation further. The adjustment of notch is VERY reliable now and you can bang all you want on the cavity and it won't change notch tuning. In contrast the larger cap and adjustable inductor were both very broad when tuning the pass band.

The dielectric 'gimic' cap.

Removing the gimic cap. The 'top' construction of the cavity can be seen here. This is done in the style where you locate some BNC's in a hollowed out section of the hardline at the top. The green goop was the glue originally used to seal the top cap onto the cavity to prevent dirt migration. I didn't like the goop approach so I purchased some black flexible plumbing end cables and secured them with hose clamps.

The tiny 1-10 pF piston trimmer. Six of these were donated to the cause.

The finished product: modified transmit and receive cavities with the new piston trimmers.

A spectrum analyzer plot showing both waveforms of the transmit and receive legs.

The finished product up at the repeater site. Just a tad bigger than 440 cans hi hi.

Final comments:

I do have some concerns over these small capacitors being used in the duplexer in regards to power handling capabilities. I have started out at 30W out of the GE Exec II transmitter. These same type caps are commonly seen in commercial duplexers rated at 200W. So hopefully I can increase power later without ill effects.

The other concern I've had since day one is temperature effects. So far the duplexers have been exposed to 17 deg F and up to 100 deg F ranges. I haven't been able to see any real problems so far with temperature.

The addition the the piston trimmers to replace the gimic caps was a success. The cavities are now very mechanically sound and the notch tuning is solid. The tuning of the duplexer didn't change at all in getting banged around during transit from the lab to the mountain repeater site.

We've been running these now for 6 months. We have a very active local 6 meter FM group and the repeater gets a fair amount of use each day.

LINKS:

I did a lot of web surfing on heliax duplexers. There are a lot of 'twist' on the design. It is instructive  to look at how others have conquered the construction of the duplexer. Below are some links that may be helpful to you.

Some other constuction examples of the same design:

http://www.wa7x.com/ki7dx_rpt.html

http://www.qsl.net/kf6yb/newplexer.html

http://www.legitimate.org/iook/sixduplex.html

http://legitimate.org/iook/wi0ok-glenarbor.html

http://chicken.cmdmail.com/Six_Meter_Heliax_and_Helical_Duplexer_info/6m_duplexer_page.htm

I've also seen 440 and 10 mtr versions on the web, but don't have those links handy.

There are a few brave soles out there that totally build up their own cavities at a machine shop:

http://www.bcarn.net/6m-project-duplexer.htm

Here are some examples at a repeater site of real six meter cans:

http://www.localconnect.net/~rikercdr/sh-3hist.htm

http://www.gemoto.com/cnharc/config2.htm

For the real thing and you have the $$$ to buy one:

 http://www.txrx.com/product/product_view.aspx?UID=13C46B7F-82BD-429D-BB08-BDA1F847E017

What you use for your 6 meter repeater antenna is a whole other issue. In my case we used tri-band comet antennas, the 2/440/6 meter ones. The tower is space limited and we couldn't afford wind loading for a single 6 mtr antenna. I know it is not a station master or DB Products and I know we'll probably have to replace it in a few years. But that is what we had to do. At least it offers some 'gain' if you believe the manufacturers figures.

Here are a couple of miscellaneous links on 6 meter antenna designs:

http://members.fortunecity.com/xe1bef/6meters-antenna.htm

http://www.6mt.com/6tech.htm  Everyone on 6 mtrs has probably seen this site a time or two.

73's and I hope this was helpful.

Ralph W4XE