Duplexer questions

[ke7vvt]

I would like to run this by the WWRF expertise before I connect everything and break something...

Wacom WP-637 Band-Reject Duplexer
Two pics attached (before I tuned it)
One datasheet
One response curve graph

1. I thought I would measure the resistance across the center and outside of the SO-239 connector just to make sure it was zero before connecting the transmitter to the duplexer. I am surprised to find that the resistance is 0.0 ohms? Is that OK?

2. Small jumper cables in between the higher frequency absorbing cavities? See pg 6 of .pdf. I removed them because after I tuned the individual cavities at the desired 145.25MHz, reconnected short jumpers, and measured again, the frequency reject peak was somewhere else. I still have these and can put them back in, and retune at the end of that cable instead of directly at the cavity connector if they do need to be there.

3. The measured response curves are shaped differently at the top (near 0dB) than the typical response curve (see pg 2). There is an additional 12dB insertion loss (due to the shape of the lower frequency notch) that I don't want, because the transmitter is going to go there.

4. At the moment I think that the 75dB notch for the receiver will be enough to block out the 25W (40dBm) from the transmitter, so I am not going to connect the transmitter through any cavities. Instead, if all goes well and I can fix the response curve, I would like to tune the other three cavities to 144.390 (only .3MHz separation!) and try to get room for a 25W aprs transmitter too. TX @ 144.390MHz, RX @ 144.650MHz, TX @ 145.250MHz, all connected to the same antenna.

Does that all sound OK or am I crazy.

 

[C2]

Why is there 400 kHz spacing?

 

[ke7vvt]

It is 600kHz between RX and TX.

If I try to put an APRS digipeater in there, between APRS and RX it is 260kHz. And between APRS and TX there is 860kHz.

Thanks C2

 

[dudmuck]

First of all, 25 watts is about 44dBm (not 40).

2nd, if you put 25 watts into a 12dB loss, then something is going to go up in smoke.

To tune those correctly you need to view swept insertion loss and return loss at the same time.
The only way to optimize insertion loss is by seeing the return loss.
The only way to optimze notch is to see the swept freq response curve in real time, because there is interaction between the cavities and thats the only way to see it.
Its something best done with a VNA such as HP8753. FInd somebody that has one.
Maybe hobby product such as mrs miniVNA is capable of that, i dunno.

 

[ke7vvt]

First of all, 25 watts is about 44dBm (not 40).

Right, 10*Log[25/0.001]=43.979 . Edit: My mistake, I was looking at the 10W setting.

2nd, if you put 25 watts into a 12dB loss, then something is going to go up in smoke.

THANKS! I will not connect said 25W transmitter behind those cavities. May I ask why?

To tune those correctly you need to view swept insertion loss and return loss at the same time.
The only way to optimize insertion loss is by seeing the return loss.

OK, how do I do that? The response curve graph I posted is measuring the Transmission/Source ratio.

The only way to optimze notch is to see the swept freq response curve in real time, because there is interaction between the cavities and thats the only way to see it. Its something best done with a VNA such as HP8753. FInd somebody that has one.
Maybe hobby product such as mrs miniVNA is capable of that, i dunno.

Got one and am not afraid to use it, just need to know what to do with it. It can measure Transmission/Source, Return/Source, and Source, Transmission, Return.

Thanks much dudmuck

 

[ke7vvt]

bump

 

[ke7vvt]

We are still unsure if it is OK for the resistance across the duplexer cavities SO-239 center conductor and shield. Does anyone know if that is OK or not OK?

Also, I still don't know why I can't connect a 25 W transmitter to a 12 dB insertion loss duplexer. I would assume it would just reduce the power output slightly.

10 * Log(25,000 mW / 1 mW) = 44 dBm
44 dBm - 12 dB = 32 dBm

Is that bad to do this for some other reason?

 

[Mike Mercury]

1. I thought I would measure the resistance across the center and outside of the SO-239 connector just to make sure it was zero before connecting the transmitter to the duplexer. I am surprised to find that the resistance is 0.0 ohms? Is that OK?

I just checked my pass/reject duplexer and it read the same zero ohms (short). I am equally surprised.

2. Small jumper cables between the cavities... I removed them because after I tuned the individual cavities at the desired 145.25MHz, reconnected short jumpers, and measured again, the frequency reject peak was somewhere else.

the only way I've ever known to tune a duplexer is with the jumpers installed.

3. There is an additional 12dB insertion loss

Talking insertion-loss (and not isolation between sides)...

YOWZA !!! an average duplexer will have a 3-4db loss per side; some of the better ones are less lossy (2 to 2.5); and of course that depends on the channel seperation. The closer the two frequencies are - the more you'd need a higher-end duplexer (obviously).

 

[ke7vvt]

MM,

Thanks!
1. OK, thats good that there is at least one other that has the same short as mine.

2. Right. I put the short cables back in there and...

3. The massive insertion loss is gone and the response curve flipped.

I haven't had the time to learn about how adding various lengths of cable to make open/shorts into a system, but I think that is what happened here. The response curve is much nicer. Not sure on the insertion loss yet. I have all of the data but haven't plotted it up yet, but that will be soon.

jcm