Filter hacking for 2395 ISS TV

[G3GJA]

Yes it will change the tuning Gareth.

I used PTFE tubing placed over the tuning pistons to lower the operating frequency of a 10-cavity duplexer originally intended for use at 1496MHz and 1354MHz so that I could tune it to 1308MHz and 1275MHz for use in a repeater.

Although the the 33MHz separation was much less than the original 142MHz, it still gave useful isolation between the TX and RX ports. However, it didn't cope with the power level needed and was prone to flashing over, but it now has use as a look-through filter at home so I can monitor the repeater whilst transmitting.

Clive

[radiogareth]

As there was a piece of PTFE in the workshop I cut off a 25mm wide strip (1.25mm thick) and placed it between the tuning screws of the 10MHz filter.
Pictures show the effect - no attempt made at retuning as it was fine as it was for RX of the ISS.
Gareth

No_tune_with_PTFE.jpg (159.13 KiB) Viewed 454 times

No_PTFE.jpg (161.28 KiB) Viewed 454 times

[G3YKI]

Hi Gareth,
"So I wonder what the peak CW/DATV power they might be capable of? "
For sure, there is not a simple answer to this question to be derived by analysis.

1) you may see the breakdown of air quoted to be 30kV/cm but it is also dependent on pressure (a lot) and the size of the gap , the materials, the nature of the surfaces, humidity etc. (and the safety factor you would want to include)
2)The voltage on the resonators will be proportional to the input voltage, but multiplied up by the 'Q' that it is operating at in the filter. That varies tremendously over the band of operation and is different for each of the resonators when there are more than one.
3)If you had figures for the voltage and the dielectric strength of air you would still need to analyse the exact shape of the gap to derive the field strength. Sharp edges or points dramatically decrease the breakdown voltage because they concentrate the electric field.*

When it comes to filling the gap with PTFE or similar strong dielectric there is a not so obvious catch to be aware of. I will try to explain by an illustration. Imagine the tuning screw air gap represented as a capacitor, maybe 1pF.
We would have exactly the same result if we divide it into two halves; namely 2 x 2pF capacitors in series.
Replace one of the air capacitors with solid dielectric, dielectric coefficient=2, that is, we filled half the air gap with solid.
We now have 2pF in series with 4pF. That is 1.33pF. So yes the tuning has been affected and needs to be compensated somewhere.
What about the voltage? Well now with 2pF and 4pF in series 2/3 the voltage is across the 2pF and 1/3 across the 4pF.
The 2pF capacitor representing the air gap has 2/3 the voltage and is 1/2 the distance it was. So THE FIELD STRENGTH IN THE AIR IS INCREASED BY 1.33 TIMES!
By putting in a filler to increase the voltage rating you may inadvertently increase the risk of failure.
In practice, you can get arcing in the air which will eventually damage and burn through the PTFE, or the filter can fill with nitric oxide which will corrode any metal parts. It has happened.**

*A filter failed as a result of a 'burr', a small spike left when a screw on a tuning plunger had been tightened. The makers repaired it by using a nylon screw. Well that lasted a few minutes. We might have agreed to PTFE had they consulted us.

** avoid breathing the yellow gas, obviously, but also don't touch silver plated parts or you will have silver nitrate on you hands and they will turn black on exposure to light.

[G3YKI]

Seeing that my junk box contained a piece of 1 5/8 OD copper tube, I thought that may be a candidate to make a filter. At 38mm I.D. the waveguide cut off frequency is 4.6GHz; so between that of the 40x40 and 30x40 tubes in the other designs. At only 160mm long it is only enough for a three pole filter.

img_5898.jpg (60.52 KiB) Viewed 59 times

img_5903.jpg (96.8 KiB) Viewed 59 times

Copper Tube 2395a.png (49.51 KiB) Viewed 59 times

I am not suggesting this is a better solution than the other filters in this thread. (I expect it would be harder to find and more expensive that the Al sections)
But shows that you don't need to follow an existing design slavishly, there are many variations possible and if the filter is properly adjusted then the performance of any 3 pole filter will be very similar. It is shorter in overall length because the ends are sealed rather than relying on the evanescent waveguide to terminate the filter. Rough sketch attached.

49_2395tubefilter.png (23.15 KiB) Viewed 59 times

[radiogareth]

Well that's nicely answered one of my 'wonderings' and prompted a quick scoot into the garage/workshop where there is a good stock of sundry copper tube, including a good length of 1&5/8". Not sure I have any copper rod but plenty of suitable brass.
Could you post a picture of the feed cap arrangement as the reduced dimension/feeding arrangement may be useful for developing a really compact solution.
Thanks.
Gareth

[g0mjw]

Gareth, remember there are other ways to make filters in tubing. You might want to save that tube.

See here https://www.g4dbn.uk/?p=517

Mike