With the advent of a new 70cm repeater in Sydney that tends to attract some good technical (and microwave!) conversation topics, I started using it only to find that as little as 200mW from my handheld will blue-screen and/or kill the audio of all digital TV channels.
I'm sure most will remember, but our TV setup here is one antenna and a nasty Kingray wideband DC to daylight preamp, driven to the max, feeding a distribution amp, that then cascades to another distribution amp, which then feeds 10 individual coax runs to each house. Fortunately, at the moment, I can still get to the coax between the antenna and the pre-amp.
The neighbours have been complaining, and one is campaigning for their own TV antenna and pre-amp. If that happens my problems will start all over again and it could be the end of amateur radio for me from home.
A local amateur kindly put together a coax stub filter for me over the weekend, which helps a lot, but I am still seeing blockiness and audio chop in the picture on some channels, the worst beiong ABC on 570-odd MHz.
I tried building a simple 70cm LC trap filter today from a design on the Southgate ARC website, but the thing simply does not work - there is no null at all.
Just wondering if anyone has a tried and tested 70cm-only filter that I can put in series with my existing 2m and 6m filters, and the 70cm stub filter? My only means of tuning it is off-air signals. Oh for a nice VNA! My 2m and 6m TVI is solved - just need to tackle 70cm now. Hoping someone can advise.
73 - Rob VK2GOM / G0MOH
70cm TVI... Filters?
-
VK3BJM
Re: 70cm TVI... Filters?
I reckon the stub filter is hard to beat for simplicity, and bang for buck. You don't mention if the stub filter you have is a single or double stub, Rob. If it's a single stub, I'd increase it to a double. I might try knocking up a couple of others to try at other points in the system - at the input to the TV, for example.
I made a filter, with 2 stubs for 144 MHz and 2 for 432 MHz, 10 years ago when I had an issue with my neighbour in Box Hill South (and the ex...). The double stubs dealt with 300 w on 2m and 100 w on 70 cm. Admittedly, there were no TV pre-amps involved in either house. Sadly, the filter was left with the neighbour when I relocated to Kyneton in 2003. I don't need it here, but it would make a good template. I can't remember the depth of the notch acheived with pairing stubs, but it may have been between 30-40 dB; it was pretty impressive for a couple of short bits of coax!
Good luck...
Barry
VK3BJM
I made a filter, with 2 stubs for 144 MHz and 2 for 432 MHz, 10 years ago when I had an issue with my neighbour in Box Hill South (and the ex...). The double stubs dealt with 300 w on 2m and 100 w on 70 cm. Admittedly, there were no TV pre-amps involved in either house. Sadly, the filter was left with the neighbour when I relocated to Kyneton in 2003. I don't need it here, but it would make a good template. I can't remember the depth of the notch acheived with pairing stubs, but it may have been between 30-40 dB; it was pretty impressive for a couple of short bits of coax!
Good luck...
Barry
VK3BJM
-
VK2GOM
Re: 70cm TVI... Filters?
HI Barry,
Thanks for your reply. Yes, it's a single. We tuned it by trimming to null a known 70cm beacon. It all but disappeared on the radio S meter.
I have read another good technique is to fire about 200mW of RF into the stub from a handheld etc., with a power meter on the other end. Trim the stub for least RF on the meter at your chose frequency. I might try and make another to go in series with this one. I hear coax like RG400 teflon is the best stuff to use, but I don't have any of that, and a search on eBay showed some suitable donor short cheap patch leads, but with crazy postage costs from the USA.
I have RG6 and RG58 here to try. I will try the feed RF into it / power meter trick and see how it performs.
73 - Rob VK2GOM / G0MOH
Thanks for your reply. Yes, it's a single. We tuned it by trimming to null a known 70cm beacon. It all but disappeared on the radio S meter.
I have read another good technique is to fire about 200mW of RF into the stub from a handheld etc., with a power meter on the other end. Trim the stub for least RF on the meter at your chose frequency. I might try and make another to go in series with this one. I hear coax like RG400 teflon is the best stuff to use, but I don't have any of that, and a search on eBay showed some suitable donor short cheap patch leads, but with crazy postage costs from the USA.
I have RG6 and RG58 here to try. I will try the feed RF into it / power meter trick and see how it performs.
73 - Rob VK2GOM / G0MOH
-
VK2GOM
Re: 70cm TVI... Filters?
I made another stub filter, using RG58 as the stub.
To tune it, I fired 5W into it from the handheld, with a power meter & dummy load on the other end and started trimming. It's now on-tune, and the stub is doing its thing and the RF out the other end doesn't move the power meter needle at all. I tuned it at 435MHz. Putting RF in it at 430MHz and 440MHz it doesn't move the meter either!
I will put this ahead of the existing notch filter and that should be the 70cm TVI sorted I reckon. I will fit it about 4am tomorrow morning so as not to cause any more aggro with the TV. Every time I have to disconnect something to put a filter in it gives the complaining neighbour more ammunition for their own antenna!
73 - Rob VK2GOM / G0MOH
To tune it, I fired 5W into it from the handheld, with a power meter & dummy load on the other end and started trimming. It's now on-tune, and the stub is doing its thing and the RF out the other end doesn't move the power meter needle at all. I tuned it at 435MHz. Putting RF in it at 430MHz and 440MHz it doesn't move the meter either!
I will put this ahead of the existing notch filter and that should be the 70cm TVI sorted I reckon. I will fit it about 4am tomorrow morning so as not to cause any more aggro with the TV. Every time I have to disconnect something to put a filter in it gives the complaining neighbour more ammunition for their own antenna!
73 - Rob VK2GOM / G0MOH
-
VK2GOM
Re: 70cm TVI... Filters?
...nothing like a 4:20am update - the only time all the neighbours seem to be in bed so I can test TVI filters out.
Test results during 435MHz Tx:
Two stub filters - brilliant. But - major problem: none of the SBS channels received on 571.500MHz stream. Obviously the digital stream is a pretty wide bandwidth from that centre frequency.
One stub filter - most channels knocked out to some extent.
So it looks like I essentially need a vey sharp filter 430 - 440MHz with a sharp roll-off between those frequencies. Sounds like a job for the future with access to a VNA and lots of experimenting.
Until then, I simply can't risk Tx on 70cm. Ah well, I tried
73 - Rob VK2GOM / G0MOH
Test results during 435MHz Tx:
Two stub filters - brilliant. But - major problem: none of the SBS channels received on 571.500MHz stream. Obviously the digital stream is a pretty wide bandwidth from that centre frequency.
One stub filter - most channels knocked out to some extent.
So it looks like I essentially need a vey sharp filter 430 - 440MHz with a sharp roll-off between those frequencies. Sounds like a job for the future with access to a VNA and lots of experimenting.
Until then, I simply can't risk Tx on 70cm. Ah well, I tried
73 - Rob VK2GOM / G0MOH
Re: 70cm TVI... Filters?
Rob
A quick question for you : you have now tested with 2 open-circuit quarter wave coaxial stubs but how have you connected the two stubs ???
Are you using a Pi arrangement : input to a T with O/C stub off one port, cabled to another T with the second stub off it and then to the existing TV feeder wiring ???
If so, the length of the coax cable between the two O/C stubs is critical and should be 1/2 wavelength ( corrected for coax Vf ) at 435 MHz. Random cable lengths will cause other undesirable resonances and will put unwanted peaks and troughs into the overall frequency response of the TV distribution system.
You could also try half wave short circuit stubs if there is no voltage fed up the coax to a masthead amp. Harder to tune though.
Previous experience here with coax stubs is that they are typically far too broad and always cause a high effective insertion loss. Good for something 50% of frequency away at VHF/UHF, too lossy closer in.
You may be better off getting a large-ish new paint can and try to convert it into a higher-Q resonator at 435 with the addition of 1 coupling loop and a flat plate tuning disc soldered to an adjustable plunge rod through the lid, and connected to the TV feeder by a suitable cable length off a T. Maximum circulating current flows through the "bottom" of the cavity so losses must be minimised by using the existing joints (even going so far as to properly soldering inside the bottom instead of relying on the existing squeeze construction) - don't even think about turning the can upside down.
Alternative to that is to build up a small helical resonator at 435 and use it via the T & cable.
A quick question for you : you have now tested with 2 open-circuit quarter wave coaxial stubs but how have you connected the two stubs ???
Are you using a Pi arrangement : input to a T with O/C stub off one port, cabled to another T with the second stub off it and then to the existing TV feeder wiring ???
If so, the length of the coax cable between the two O/C stubs is critical and should be 1/2 wavelength ( corrected for coax Vf ) at 435 MHz. Random cable lengths will cause other undesirable resonances and will put unwanted peaks and troughs into the overall frequency response of the TV distribution system.
You could also try half wave short circuit stubs if there is no voltage fed up the coax to a masthead amp. Harder to tune though.
Previous experience here with coax stubs is that they are typically far too broad and always cause a high effective insertion loss. Good for something 50% of frequency away at VHF/UHF, too lossy closer in.
You may be better off getting a large-ish new paint can and try to convert it into a higher-Q resonator at 435 with the addition of 1 coupling loop and a flat plate tuning disc soldered to an adjustable plunge rod through the lid, and connected to the TV feeder by a suitable cable length off a T. Maximum circulating current flows through the "bottom" of the cavity so losses must be minimised by using the existing joints (even going so far as to properly soldering inside the bottom instead of relying on the existing squeeze construction) - don't even think about turning the can upside down.
Alternative to that is to build up a small helical resonator at 435 and use it via the T & cable.
Doug VK4ADC, QG62lg51
http://www.vk4adc.com
http://www.vk4adc.com
-
VK2GOM
Re: 70cm TVI... Filters?
Hi Doug,
Thanks for your informative reply. The two stubs were made with short (but random) lengths of coax off each side of the T, each terminated in an F plug, using F joiners as necessary to tap into the system. I have heard about the v/f corrected half-wave between two stubs, but must admit there was by virtue of their construction just a random length between which on measurement is just about a corrected full wave - could be a cause of problems. I could shorten one of the stub T's down to arrive at a half-wave and try that.
But, I suspect - as you say - that they are too low Q and broad and are having detrimental effects on the ability to receive SBS at 571.500MHz.
The paint can resonator sounds interesting... is this something you have built before? Do you have any further details on contruction regarding the coupling loop etc.?
73 - Rob VK2GOM / G0MOH
Thanks for your informative reply. The two stubs were made with short (but random) lengths of coax off each side of the T, each terminated in an F plug, using F joiners as necessary to tap into the system. I have heard about the v/f corrected half-wave between two stubs, but must admit there was by virtue of their construction just a random length between which on measurement is just about a corrected full wave - could be a cause of problems. I could shorten one of the stub T's down to arrive at a half-wave and try that.
But, I suspect - as you say - that they are too low Q and broad and are having detrimental effects on the ability to receive SBS at 571.500MHz.
The paint can resonator sounds interesting... is this something you have built before? Do you have any further details on contruction regarding the coupling loop etc.?
73 - Rob VK2GOM / G0MOH
Re: 70cm TVI... Filters?
Rob
I did make one from a paint can - but that was probably 20-25 years ago. I do not remember the dimensions I used and no longer have the resonator.
A quick web search found the following article that talks about ways of coupling to a cavity resonator : http://www.repeater-builder.com/antenna ... ers/4.html but doesn't contain any diagrams or graphics.
...and a reference : "At 450 MHz the diameter limit is roughly 8 in."
The other article which explains a little more - though more complicated construction than you need for shunting out 435 MHz from a TV distribution system is : http://www.ingenia.org.uk/ingenia/issue ... /david.pdf.
The diameter of the inner conductor (from memory) approximates a 50 ohm transmission line so varies with outside diameter - with an air dielectric, and is used as far as possible into the assembly but with a plunge/screw adjustment on the end.
Try also http://nz5v.net/6m-project-duplexer.htm although that relates to building 6m cavities - formula included.....
Follow the same general test setup for the tuning process as you have with the coaxial stubs but with the cavity on the end of a 1/4 or 1/2 wave Vf'd coax (don't remember which) from a T connection and tune for minimum throughput at 435. One gives a slope on the high side and the other gives the slope on the low side of resonance but very low insertion loss on the other side.
One of the old ARRL Handbooks or RSGB publications may show other methods of building resonators for UHF repeater diplexer/duplexer applications - use ideas from there.
Of course you might be lucky enough to find an existing coaxial resonator for 400-470MHz for sale on eBay.....
I did make one from a paint can - but that was probably 20-25 years ago. I do not remember the dimensions I used and no longer have the resonator.
A quick web search found the following article that talks about ways of coupling to a cavity resonator : http://www.repeater-builder.com/antenna ... ers/4.html but doesn't contain any diagrams or graphics.
...and a reference : "At 450 MHz the diameter limit is roughly 8 in."
The other article which explains a little more - though more complicated construction than you need for shunting out 435 MHz from a TV distribution system is : http://www.ingenia.org.uk/ingenia/issue ... /david.pdf.
The diameter of the inner conductor (from memory) approximates a 50 ohm transmission line so varies with outside diameter - with an air dielectric, and is used as far as possible into the assembly but with a plunge/screw adjustment on the end.
Try also http://nz5v.net/6m-project-duplexer.htm although that relates to building 6m cavities - formula included.....
Follow the same general test setup for the tuning process as you have with the coaxial stubs but with the cavity on the end of a 1/4 or 1/2 wave Vf'd coax (don't remember which) from a T connection and tune for minimum throughput at 435. One gives a slope on the high side and the other gives the slope on the low side of resonance but very low insertion loss on the other side.
One of the old ARRL Handbooks or RSGB publications may show other methods of building resonators for UHF repeater diplexer/duplexer applications - use ideas from there.
Of course you might be lucky enough to find an existing coaxial resonator for 400-470MHz for sale on eBay.....
Doug VK4ADC, QG62lg51
http://www.vk4adc.com
http://www.vk4adc.com
-
VK2GOM
Re: 70cm TVI... Filters?
Brilliant Doug, many thanks. It's what I like about ham radio; that likeminded individuals will go out of their way to help other likeminded individuals.
Looks like some useful reading for me.
My only experience with cavity resonators so far is at 10GHz and they're tiny , just PCB rivets and an M4 screw!
73 - Rob VK2GOM / G0MOH
Looks like some useful reading for me.
My only experience with cavity resonators so far is at 10GHz and they're tiny , just PCB rivets and an M4 screw!
73 - Rob VK2GOM / G0MOH
-
VK3BJM
Re: 70cm TVI... Filters?
Rob, do you have access to a spectrum analyser and an oscillator?
If the stubs have been cut to the correct length for 70 cm (439 MHz, say) the notch shouldn't be taking out a TV channel 130 MHz away. I'd be expecting the notch to be about 30 MHz wide. Here, again, I wish I had the filter I left in Melbourne - I could quote good, hard specs...
I made mine based on information given by Ron, VK3AFW, in his GippsTech 2000 presentation, "A Simple TVI Fix". Two 1/4 wave stubs, cut a bit longer than the length based on the velocity factor of the cable, then trimmed one at a time before combining the two. I have a spacing of 1/4 wavelength between the "Tees". I did have access to a spec an, and a variable oscillator (sadly, not a proper sweep, but it was adequate for the job), so I could see what was happening each time I went "SNIP"!
I made mine from some thin teflon 75 ohm cable, which allowed the whole thing to be fitted into a small die-cast box, whereas RG-59 or RG-6 would not have fitted - this detail is neither here nor there, other than highlighting the usefulness of the spec analyser, as I didn't know the VF of the cable...
Having two stubs, of similar but not the same length, will broaden the notch. Obviously, this will be quite visible when a spec an is being used during the trimming process. If one of your stubs has been cut a little short, the notch may have shifted up enough to take out SBS.
I was tempted, during my previous posting, to suggest that during your 4 am "Outage Window" you insert a Filter Bypass Switch; an Input, an Output, and two "F" sockets for inserting various filters. That way changes could be made more quickly and easily - and perhaps at a more user-friendly hour! Just switch to "Bypass", connect your filter, then switch it inline. If it isn't behaving, switch back to "Bypass", and take the filter away for a tweak.
73,
Barry
VK3BJM
If the stubs have been cut to the correct length for 70 cm (439 MHz, say) the notch shouldn't be taking out a TV channel 130 MHz away. I'd be expecting the notch to be about 30 MHz wide. Here, again, I wish I had the filter I left in Melbourne - I could quote good, hard specs...
I made mine based on information given by Ron, VK3AFW, in his GippsTech 2000 presentation, "A Simple TVI Fix". Two 1/4 wave stubs, cut a bit longer than the length based on the velocity factor of the cable, then trimmed one at a time before combining the two. I have a spacing of 1/4 wavelength between the "Tees". I did have access to a spec an, and a variable oscillator (sadly, not a proper sweep, but it was adequate for the job), so I could see what was happening each time I went "SNIP"!
I made mine from some thin teflon 75 ohm cable, which allowed the whole thing to be fitted into a small die-cast box, whereas RG-59 or RG-6 would not have fitted - this detail is neither here nor there, other than highlighting the usefulness of the spec analyser, as I didn't know the VF of the cable...
Having two stubs, of similar but not the same length, will broaden the notch. Obviously, this will be quite visible when a spec an is being used during the trimming process. If one of your stubs has been cut a little short, the notch may have shifted up enough to take out SBS.
I was tempted, during my previous posting, to suggest that during your 4 am "Outage Window" you insert a Filter Bypass Switch; an Input, an Output, and two "F" sockets for inserting various filters. That way changes could be made more quickly and easily - and perhaps at a more user-friendly hour! Just switch to "Bypass", connect your filter, then switch it inline. If it isn't behaving, switch back to "Bypass", and take the filter away for a tweak.
73,
Barry
VK3BJM
-
VK2GOM
Re: 70cm TVI... Filters?
Hi Barry,
Thanks for the further information. It looks like trimming the stub with Tx power seemed fairly accurate. Easy to see the power out less and less and more was trimmed off. I don't have a SpecAn here sadly, nor a sweep Osc. I can only use radios to generate the RF.
However, there has been an interesting development...
The TV antenna fitter (I am loathe to use the term 'engineer') was here today doing his survey, and what would you know, whilst he was fiddling, the bain of my life nasty Kingray preamp somehow died. So nobody had ANY picture on their TV for a while. He disappeared in his white van and brought back another pre-amp, this time a Hills one in a flat metal box. It looks a bit primitive, but here is the interesting part: When he wired it in, he took out my 2m/6m notch filter. It was only later when I poked my head through the attic hatch I saw it there lying on the government-mandated insulation batts. I wasn't impressed at the time...
However, knowing I couldn't disconnect again to re-install it right then, I grabbed the Yaesu handheld and did a couple of tests. 5W FM on 2m, put out a call - not a flicker. Tried 70cm, same again, no problem at all! On the old Kingray pre-amp (no filter) 2m at 200mW was enough to 'blue screen' the system, and the same on 70cm.
I tried 50W from the Icom in the shack on 2m, which did blue-screen the TV, which I half-expected. However, the new amp is much less affected by 2m and even less on 70cm.
I will insert the 2m/6m notch filter in the next early hours maintenance window tomorrow. 70cm will then be a lot simpler to solve, I hope if I don't need as much attenuation. I do like your idea of the filter bypass switching arrangement though! If the experiments look set to continue, I will build something to do that.
73 - Rob VK2GOM / G0MOH
**Update. Notch filter fitted. 100W on 2m not a problem. Can't try 6m as the beam is down due to lack of band activity! But I am confident it will be fine.
Thanks for the further information. It looks like trimming the stub with Tx power seemed fairly accurate. Easy to see the power out less and less and more was trimmed off. I don't have a SpecAn here sadly, nor a sweep Osc. I can only use radios to generate the RF.
However, there has been an interesting development...
The TV antenna fitter (I am loathe to use the term 'engineer') was here today doing his survey, and what would you know, whilst he was fiddling, the bain of my life nasty Kingray preamp somehow died. So nobody had ANY picture on their TV for a while. He disappeared in his white van and brought back another pre-amp, this time a Hills one in a flat metal box. It looks a bit primitive, but here is the interesting part: When he wired it in, he took out my 2m/6m notch filter. It was only later when I poked my head through the attic hatch I saw it there lying on the government-mandated insulation batts. I wasn't impressed at the time...
However, knowing I couldn't disconnect again to re-install it right then, I grabbed the Yaesu handheld and did a couple of tests. 5W FM on 2m, put out a call - not a flicker. Tried 70cm, same again, no problem at all! On the old Kingray pre-amp (no filter) 2m at 200mW was enough to 'blue screen' the system, and the same on 70cm.
I tried 50W from the Icom in the shack on 2m, which did blue-screen the TV, which I half-expected. However, the new amp is much less affected by 2m and even less on 70cm.
I will insert the 2m/6m notch filter in the next early hours maintenance window tomorrow. 70cm will then be a lot simpler to solve, I hope if I don't need as much attenuation. I do like your idea of the filter bypass switching arrangement though! If the experiments look set to continue, I will build something to do that.
73 - Rob VK2GOM / G0MOH
**Update. Notch filter fitted. 100W on 2m not a problem. Can't try 6m as the beam is down due to lack of band activity! But I am confident it will be fine.