Magnetic Loop vacuum capacitors from USSR

CT2GXW

CT2GXW

Star Member
Hi all,

I live in a 6 story building, and I plan to install a mag loop on the rooftop as my main antenna. I want to build it myself, and will go the vacuum variable capacitor way. My question is, does anybody has experience with these old soviet capacitors available on Ebay? I've seen lots of online projects of hams using them with success, but I wonder after 40-50 years (?) these capacitors still hold the vacuum inside, even if they are in NOS condition. I don't want to pay 150? + and come to the conclusion it doesn't go as expected...

Thanks
Pedro

 
After following loop designs/builders over many years I have noticed that now and again some of the capacitors arrive either smashed in transit, or else they have lost their vacuum. Without that vacuum a 10,000v rated capacitor suddenly becomes a 1-200v item! :o

All of the cases I read about where the capacitor was damaged or faulty, the sellers have generally refunded all monies, and allowed the buyer to keep the capacitor.

73, Mark...
 
Thanks Mark, that proves my concerns are right. Even if my money was return, I would hate to be in such a situation. To come to the conclusion that the vacuum is gone would have me assemble all the antenna and find it out only when everything was in place.

I will check for new capacitors from current manufacturers, if the price is not super high. Or in the last case will reconsider the air variable ones, maybe build one, but it will have to handle 5kv...

Pedro

 
If you do a search on Google for "diy high voltage variable capacitor" you'll find LOTS of info bd

I don't know if you are aware of another technique for making them, you essentially take a high capacitance air spaced variable (think radio tuning capacitor), and then remove some of the plates from both the fixed section and the moving vanes. You end up with much less maximum capacitance than before the mod, but the voltage rating becomes very much higher.

73, Mark...
 
Thanks Mark, I will consider building the air variable myself. Then I need to sort out the motor drive system for tuning, which I will probably do with a stepper motor.

To make one of these antennas for a permanent installation is quite a task. No wonder the commercial models are so expensive.
 
CT2GXW said:
Thanks Mark, I will consider building the air variable myself. Then I need to sort out the motor drive system for tuning, which I will probably do with a stepper motor.

To make one of these antennas for a permanent installation is quite a task. No wonder the commercial models are so expensive.
Click to expand...

You could use a stepper motor, but an easier way (which I've used a number of times before), is to use a servo (the kind used as actuators in radio control models), they have very high torques for their size, and have pretty robust gearboxes too.

It is possible to modify the servos by disconnecting the internal servo amplifier board and positional feedback potentiometer (just remove then both), then you can use DC to wind the motor/gearbox back and forth. The gear ratio is not normally high enough to give accurate tuning on a loop, so I normally add an extra couple of pulleys or sprockets to give a further reduction to the capacitor shaft. Sometimes you can get slow motion drives too, maybe 8:1 or 16:1 etc.

With a modified servo, pulleys/reduction drive, and a variable capacitor you have everything you need to remotely tune a loop. You could PWM the motor (which allows for a high torque but a slower speed than a continuous DC feed, this is how MFJ control their loop motors), or else you could use a really simple control system like I used sometimes in the past. Just a few batteries and four push button switches. Two switches give either positive or negative at one voltage, the other two give positive or negative at a higher voltage. The low volts allows for slow tuning, the higher volts for fast tuning (changing bands or going from one end of a band to the other). Also. you can tap the switches with your finger to give a kind of manual PWM output.

It may sound crude but it works. I used it to tune an 80m mag-loop I built in my attic here, worked fine. As an extra refinement I added two microswitches that were triggered by the position of the variable capacitor. So if the capacitor was fully one way it would open the switch and stop rotation any further in that direction. Same at the opposite direction of travel.

Once set up correctly you just give the loop some RF drive and press the up and down tune button while keeping an eye open for a dip in the SWR.

There's a design floating around on the web for an auto tuning mag-loop designed by G4WIM. That uses a much more advanced micro to control everything, and a stepper motor plus reduction drive for the 'muscle' of the system.

https://www.youtube.com/watch?v=BZ4HedpjqOY

It all depends how elaborate you want to get with the tuning control gear, but I think if you go too crazy with it too soon you miss the fun and amazement of using magnetic loops. Over the last forty to fifty years many of the loops I've tinkered with have just been made with sections of coaxial cable, and using old capacitors removed from MW/SW radios to tune them. No good for QRO power, but I don't normally run more than 5 watts anyhow, in fact a lot of my QRP rigs are only 1-2 watts at most, so flashover isn't an issue here.

I do have an MFJ-1788X loop which I think is rated at 150 watts or so, but again it only ever sees QRP levels of RF bd

73, Mark...







 
Great information Mark, thanks. I've also seen one using a 555 timer circuit to generate a PWM for a 12V DC motor, with a button for CW, one for CCW and a pot for adjusting the speed. That might also be a simple way to do it.

In the meantime, I came across this Etsy "store" in Turkey that sells sets of plates for a butterfly air variable, and even the 3D printed end stands:

https://www.etsy.com/listing/888848143/5-sets-capacitor-plates-1-mm-homebrew?ref=shop_home_active_1

I feel very tempted to order some sets and build one around 200pF.

Regarding your loops performance, how would you rate them against other wire antennas? It's a lot of work compared to much simpler solutions, but since I want to install an antenna on the rooftop - and I'm trying to avoid verticals for noise, lack of a good radial ground and higher risk of lightning strike - I thought it could be a good solution. On the other hand, the need for constant retuning is not very atractive...

Pedro

 
Hi Pedro.

Those capacitor plates are a good find!  bd

The two initial things that effect the performance of a loop are the diameter of the loop conductor, and the spacing of the capacitor plates. Really thin wire just does not perform well if the loop is very much reduced in size, but if the circumference of the loop is around the maximum that will work (approx. 1/4 wavelength long, though technically you can go a little longer than that until the loop becomes self resonant at the top end of its coverage range), then you can get away with thinner conductors, though what you gained by increasing the loop diameter you loose in conductor losses!

I have used the braid of cheap 50 Ohm feeder (about 5mm dia), and also thicker coaxial cable such as RG213, and all worked well (though I've also used 15mm copper water pipe and pretty much anything else conductive!)

Obviously with the capacitor the greater the air space the more power your loop will take before flash-over occurs. I just use whichever is the best one to hand at the time, and also rather than use a really large expensive capacitor to tune a small loop I will very often make the loop a little larger so that all that is needed is a capacitor large enough just to trim the loop to the required frequency (this is only really effective for single band use though).

You can also use fixed value high voltage capacitors in parallel with the tuning capacitor, again this allows you to use a less expensive capacitor to tune across your desired frequency range. They use this idea in the Chameleon Loop, it has a switch to change from the upper range to the lower range.

The loops generally perform well because they are resonant on the desired frequency, which means in turn you will reduce your feeder losses to a minimum. A lot of guys are using coaxial feeder to connect up a random or non-resonant ant antenna and then rely on the ATU back in the shack to allow their radios to 'see' a good impedance/SWR, but this is only OK between the ATU and the rig. After the ATU all sorts of strange impedance/SWR values may be present with the result that maybe as much as 90-95% of the power they think they are supplying to the antenna is simply lost in the feeder.

So a guy running say 5 watts in to a magnetic loop may end up with an ERP as good as another guy running 100 watts in to a totally mis-matched lossy antenna system!

If you have your loop elevated on the roof of a building then in theory it should work well either horizontal or vertical. f the loop is mounted low down, or close to a conductive surface (or reinforcing in concrete say) then it is best to mount it vertically or perpendicular to the conducting item as this will reduce losses greatly.

Although vertical whips etc. do appear noisy in a lot of urban locations, I've found this is not the case with the loops. All of my loops are used vertically and they work fine. I have tried side by side comparisons with vertical antennas and horizontal antennas, and the loops have pretty much always won on both signal strength and/or signal to noise ratio.

The only downside with mounting to loop vertically is that because they are bi-directional when mounted that way, you have to choose the orientation carefully (I use a Great Circle beam heading map to figure out the 'best' position). I am lucky though since a NW/SE orientation gives me Europe in one direction and USA/Canada etc. in the other direction. Also it turned out to be the ideal orientation to null out some local QRM!

73, Mark...
 
Just thinking.... folks don't use normal ATUs for 630m because of the high voltage flashover problem, and the standard approach is a stonking big tapped  inductor with a variometer for fine tuning. Is there any reason apart from cost that a vacuum variable capacitor would work for matching the antenna?
 
G0KZZ said:
Hi Pedro.

Those capacitor plates are a good find!  bd

The two initial things that effect the performance of a loop are the diameter of the loop conductor, and the spacing of the capacitor plates. Really thin wire just does not perform well if the loop is very much reduced in size, but if the circumference of the loop is around the maximum that will work (approx. 1/4 wavelength long, though technically you can go a little longer than that until the loop becomes self resonant at the top end of its coverage range), then you can get away with thinner conductors, though what you gained by increasing the loop diameter you loose in conductor losses!

I have used the braid of cheap 50 Ohm feeder (about 5mm dia), and also thicker coaxial cable such as RG213, and all worked well (though I've also used 15mm copper water pipe and pretty much anything else conductive!)

Obviously with the capacitor the greater the air space the more power your loop will take before flash-over occurs. I just use whichever is the best one to hand at the time, and also rather than use a really large expensive capacitor to tune a small loop I will very often make the loop a little larger so that all that is needed is a capacitor large enough just to trim the loop to the required frequency (this is only really effective for single band use though).

You can also use fixed value high voltage capacitors in parallel with the tuning capacitor, again this allows you to use a less expensive capacitor to tune across your desired frequency range. They use this idea in the Chameleon Loop, it has a switch to change from the upper range to the lower range.

The loops generally perform well because they are resonant on the desired frequency, which means in turn you will reduce your feeder losses to a minimum. A lot of guys are using coaxial feeder to connect up a random or non-resonant ant antenna and then rely on the ATU back in the shack to allow their radios to 'see' a good impedance/SWR, but this is only OK between the ATU and the rig. After the ATU all sorts of strange impedance/SWR values may be present with the result that maybe as much as 90-95% of the power they think they are supplying to the antenna is simply lost in the feeder.

So a guy running say 5 watts in to a magnetic loop may end up with an ERP as good as another guy running 100 watts in to a totally mis-matched lossy antenna system!

If you have your loop elevated on the roof of a building then in theory it should work well either horizontal or vertical. f the loop is mounted low down, or close to a conductive surface (or reinforcing in concrete say) then it is best to mount it vertically or perpendicular to the conducting item as this will reduce losses greatly.

Although vertical whips etc. do appear noisy in a lot of urban locations, I've found this is not the case with the loops. All of my loops are used vertically and they work fine. I have tried side by side comparisons with vertical antennas and horizontal antennas, and the loops have pretty much always won on both signal strength and/or signal to noise ratio.

The only downside with mounting to loop vertically is that because they are bi-directional when mounted that way, you have to choose the orientation carefully (I use a Great Circle beam heading map to figure out the 'best' position). I am lucky though since a NW/SE orientation gives me Europe in one direction and USA/Canada etc. in the other direction. Also it turned out to be the ideal orientation to null out some local QRM!

73, Mark...
Click to expand...

Thanks for sharing your experience Mark, it helps me a lot.

Just 2 additional questions:

I?ve read one ham saying he had to retune the loop every 3KHz for transmitting. Is it that bad in terms of narrow bandwidth? That would mean constantly retuning even just for the CW portion of the bands. I know the bandwidth can be greater, but for lower efficiency.

Could you work the US with your loop regularly? With my modest dipole I sometimes can, but the signals are marginal, and usually the stations are on the east coast, close to the ocean.

Pedro
 
CT2GXW said:
Thanks for sharing your experience Mark, it helps me a lot.

Just 2 additional questions:

I?ve read one ham saying he had to retune the loop every 3KHz for transmitting. Is it that bad in terms of narrow bandwidth? That would mean constantly retuning even just for the CW portion of the bands. I know the bandwidth can be greater, but for lower efficiency.

Could you work the US with your loop regularly? With my modest dipole I sometimes can, but the signals are marginal, and usually the stations are on the east coast, close to the ocean.

Pedro
Click to expand...

Two things 'might' cause the op' to have to retune every 3kHz, one is that they are using a small loop (relative to the wavelength they are operating on), the other is that possibly they were using a smallish loop but operating on a low frequency. For example my 90cm dia loop has maybe 200-300kHz on 10m, but down on 40m the same size loop gives a bandwidth of around 20kHz or so. But if I were to use a larger loop on 40m the bandwidth might go up to say 60-80kHz.

By bandwidth I generally look for a 2:1 SWR upper and lower limit for wherever the loop is tuned to. If your rig could tolerate a 3:1 SWR you could use a wider range before having to retune the loop. However, you do find if you are tuned too far away from the desired operating frequency the efficiency drops off rapidly. Also, narrow modes like CW are more tolerant of narrow bandwidths than say SSB or AM.

So if you used a small diameter loop on a low frequency band you may well find that if you used SSB for example, the bandwidth of the transmitted signal is such that some of the energy would be in the range that was 1:1, but other 'parts' of the energy spectrum would fall in areas that presented a 2:1 or even 3:1 SWR. Higher SWR on the loop can mean much less radiated energy, and in our SSB example here you may well sound like you are talking down a tube or have restricted audio bandwidth to some distant station listening to your signal.

In effect the loop is a bandpass filter, and worse case scenario is that it will filter out energy you want to radiate!

This is a worse case example though, but worth keeping in mind. With my 90cm dia loop I find if I tune it around 7020kHz I can happily receive the entire CW section without having to retune, though the signals (and noise) do fall off a bit at each end of the range. For TX though I will retune the loop for the best SWR figure I can get, the hope being that the losses in the feeder will be reduced to the lowest amount possible, and the loop should be radiating at its best efficiency for the frequency used.

As for the other question, within reason if you can hear 'em you can work 'em!  With QRO power levels (say 100W) even a piece of wet string would work! I've used 1.5W on 20m with a simple loaded dipole in the attic, and that managed stations around Europe and the States fine, and I'm pretty sure the loop would be a heck of a lot more efficient.

I have picked up my antenna analyser on SDR sites around Europe and USA when connected to the magnetic loop while performing tests. So if an MFJ-259 can be picked up around the world from the loop imaging what 100W could do! ;D

73, Mark...

p.s. I think I still have some screen grabs of my MFJ-259 picked up on 30m on a web SDR site, I'll see if I can find them later.
 
GM0WEZ said:
Just thinking.... folks don't use normal ATUs for 630m because of the high voltage flashover problem, and the standard approach is a stonking big tapped  inductor with a variometer for fine tuning. Is there any reason apart from cost that a vacuum variable capacitor would work for matching the antenna?
Click to expand...
Short answer is nope! (providing it can withstand the voltages)

The only reason the voltages are so high is because of the high impedance of the antenna requiring lots of 'force' to get current flowing in it. With lots of capacitive loading at the 'top' of the antenna the feed-point impedance would drop considerably and so would need a lot less voltage, also you would have a greatly increased current in the antenna.

Another way is to go 'old school'. Instead of a single top wire you could use a spreader system, running the wire back and forth several times to create an electrically longer antenna.

Also another method is to create a 'meander' antenna, where instead of having the wire running up and down repeatedly as in the spreader antenna, you 'meander' or zig-zag the wire from side to side. This again creates an electrically longer antenna than would be possible with just a single straight run.

Thirdly you could combine the spreader or meander technique with a capacitive top load for even greater effect! ;D

I've seen some of the coil loaded antennas for the LF bands and often thought that the wire used would be much more useful up in the air rather than wrapped around a dustbin or washing basket. bd

73, Mark...
 
GM0WEZ said:
Just thinking.... folks don't use normal ATUs for 630m because of the high voltage flashover problem, and the standard approach is a stonking big tapped  inductor with a variometer for fine tuning.
Click to expand...

Here's an MF loop design idea by W1TAG... http://www.w1tag.com/Camp_Loop/Camp_Loop.htm

Just after finding the above link I found this site by WH2XND where he uses the vacuum variables in an arrangement more along the lines you were possibly suggesting... https://wh2xnd.com/

73, Mark...

 
Found these videos by M0MSN about making your own variable capacitors for transmitting loops, thought it might be of interest...

https://www.youtube.com/watch?v=VGLZgT0psQ8

https://www.youtube.com/watch?v=flJkCNMdpuc

There's another video by a different film maker where they are using double sided PCB material to make a 'linear' variable capacitor (their wording not mine). It's a bit cumbersome as it stands, but it does have possibilities.

https://www.youtube.com/watch?v=5Ji7RJxi4vc

73, Mark...
 
You must log in or register to reply here.
Back
Top