Batboard

Riddle me this...

I removed a Rackmount VHF Quantar duplexer 144-160 working to spec 154 input, 156 output.
>75db isolation, ~1.2db insertion loss.

Now take the same duplexer, and retune it with the notches on the new freq's 159 input, 151 output.
The notches are as they should be. >75db isolation. The problem is insertion loss.
159 input shows -13db loss
151 output shows -19db loss

It's easier to show a picture, but...
On the 159 notch, for example, on the high-pass side of the notch, it peaks up around 2MHz from notch with -1.2db loss. Then it has a gradual slope down as you move away from the notch (to -13db @159)

It's an RFS duplexer, Part/Model # 0185417U02, Rev 00
Using a R2600 with track gen.

Any ideas?

tpg

Since your input is now higher than the output (compared to before re-tuning) did you swap the input & output on the dupes?
Likely the duplexer needs to be thot of as "high pass" and "low pass" not input and output.
Those numbers sound about right (wrong) for the inter cavity cable lengths being wrong.

159 on the high side
151 on the low side.

Reversed from earlier.

tpg

I think that's the RFS 5042 series duplexer... Is it the rack mount one with the black balls on the ends of the tuning rods?

If so, that thing is only designed to work on splits between 0.5 and 5 MHz. They don't make this very clear on their old spec sheets or catalogs, which is stupid. It doesn't even appear on their website now, so maybe they don't make it anymore?

I know someone who got one to work with a 7 MHz split. I think he had called RFS tech support, and they advised him of the 5 MHz max split limitation, but said it still might tune up to 7, but wouldn't guarantee it.

Brass knobs on the end? This thing has no balls.........

It is rack mountable.

God this sucks...

tpg

I googled a bit and found a pic of the 5042. It's shown in this hamsexy installation right behind the MASTR II, on the floor.

Here's an interesting email exchange on the 5042, as well:

http://www.mail-archive.com/repeater-bu ... 42436.html

PETNRDX wrote:Those numbers sound about right (wrong) for the inter cavity cable lengths being wrong.

What SHOULD they be? I'm leaning this way also.
I'm using Belden RG142.

Edit: Found calc based on 1/4wave x vf(.709 for RG142)
151 = 13.8"
159 = 13.11"

Look right?

Thanks,
tpg

Looks right, but you really need to be using a network analyzer rather than just a service monitor when you're getting this deep into it. I've seen duplexers that are tuned perfectly according to a spec-an/TG combo, but are terribly mismatched. Make sure to check SWR from the TXer into the duplexer before deploying the unit.

Calling EMR and ordering what you need is probably not an option, right?

it going back... screw it.

You probably need to swap the ports around. You can't just retune the slopes of the
cavities by retunning them. Think about the way they were before you started to play with hem.
Keep the low frequency on the low side cavity and the high frequency on the high cavity. just
because your input and output are backwards from the way it was used originally makes no
difference. Just change the in and out labels and you should be fine.

There may be enough of a frequency change that the cable lengths may come into play,
but I would try it first with the original cables. The change should not be that much in
the cable lengths.

Jim

The Pager Geek wrote:it going back... screw it.

Yes,

I swapped ports around for giggles. I tried tuning both the labeled "high" as 159 and then the "low" as 159.
Needless to say when tuning the Low as 159 it was EVEN WORSE than the -13db insertion loss seen on the high port.

The ports are where they should be.
High = 159 pass, 151 notched
Low = 151 pass, 159 notched

High port has -13db insertion loss at 159
Low port has -19db insertion loss at 151

It's out the door to the factory, so I'm on to the next project...

tpg