2.4 GHz link - best performance with one antenna H and one V! WTF?

[Delta]

I've got a cheapo analogue (PAL) CCTV camera in a shed, and I'm sending the signal to the house using a pair of cheapo 2.4GHz (analogue) A/V Tx/Rx boxes (not these actual ones, but you get the idea - http://www.accessories-shops.com/images/pdimg/WVT810BD.jpg

The Tx is against a window, with the standard antenna.
The Rx is in the loft, with a big Ebay Yagi on it.
Distance about 20m. The internal roof is boarded and the roof tiles are quite hefty.

Initially I set it up with horizontal polarisation (ie, the standard aerial pointing sideways, parallel to the house) and the Yagi also horizontal (elements parallel to the floor), my thinking being that most WifI routers use vertical pol.

It worked OK, but I thought I'd try V, so I first moved the Rx Yagi to V, I notice the picture immediately got much better!  Then moving the Tx ant to V (pointing it to the sky) the picture degraded!

I've left it with Rx V and Tx H, and whilst I'm happy with the picture, the OCD side of me says this is not right!  What is going on?
 

[AF6LJ]

Surrounding materials will have an effect on your antennas performance.

While you say you are using a yagi for your RX is it a yagi made up of dipoles or loops??

[Delta]

Surrounding materials will have an effect on your antennas performance.

While you say you are using a yagi for your RX is it a yagi made up of dipoles or loops??

It's this kind of affair...
http://en.arronna.com/imageRepository/850f6cf1-40aa-405f-b97c-83776bf328e3.JPG

[PA0PBZ]

The polarization probably changes due to reflections and such. Stop worrying and enjoy.

[AF6LJ]

Surrounding materials will have an effect on your antennas performance.

While you say you are using a yagi for your RX is it a yagi made up of dipoles or loops??

It's this kind of affair...
http://en.arronna.com/imageRepository/850f6cf1-40aa-405f-b97c-83776bf328e3.JPG

Okay Dipoles...
I think the above post has the actual answer, not knowing your installation my best guess also would be the polarization has flipped due to reflections.

[Delta]

Ah, that would explain it then!

There is a "plane" of about 2.5m of wire mesh directly in front of the Tx antenna! 

PS.  I'm wasn't worried, just confused! 

[AlessandroAU]

Surrounding materials will have an effect on your antennas performance.

While you say you are using a yagi for your RX is it a yagi made up of dipoles or loops??

It's this kind of affair...
http://en.arronna.com/imageRepository/850f6cf1-40aa-405f-b97c-83776bf328e3.JPG

That antenna is less than useless. Some antennas of that style from china have more gain when facing the wrong way. Throw it away and make your own.

[Delta]

I wouldn't say less than useless, as it does at least work.  It seems to have some directivity, as I needed to be quite accurate when aiming it...

[AlessandroAU]

I wouldn't say less than useless, as it does at least work.  It seems to have some directivity, as I needed to be quite accurate when aiming it...

That may have been harsh. I'm a little bitter about these having wasted quite a bit of time trying to get one to work.

Have a read here https://asdfghjkl.me.uk/blog/antennas/

[Delta]

I wouldn't say less than useless, as it does at least work.  It seems to have some directivity, as I needed to be quite accurate when aiming it...

That may have been harsh. I'm a little bitter about these having wasted quite a bit of time trying to get one to work.

Have a read here https://asdfghjkl.me.uk/blog/antennas/

Erm, were you using a 2.4GHz Yagi for 2.1GHz 3G.....?

[PA0PBZ]

Erm, were you using a 2.4GHz Yagi for 2.1GHz 3G.....?

 

Edit: removed picture of antenna showing 2GHz band, site no longer wants to show it...

[AlessandroAU]

Of course not, the point was to demonstrate the horrible construction quality. Regardless, if you read the site I linked you will see that person examines the common 2.4ghz antennae as well.

Personally, I got not better performance on my eBay yagi than the standard 5dbi dipoles I was trying to improve on.

[Delta]

Sorry, my mistake, I zoomed in on one of the photos that said 2.4GHz, it must have only been an example photo...

[Delta]

That antenna is less than useless. Some antennas of that style from china have more gain when facing the wrong way. Throw it away and make your own.

Right, I owe you an apology Alessandro; you were spot on, the eBay Yagi is in enormous pile of shite.

It is indeed worse than useless.

I've been playing about whilst looking at the demodulated video signal with my scope (I don't know what that says about me: I don't have a TV I can take up to the loft, but I do have a scope!) and it was an utter pile of pish - the telly was doing a pretty good job getting a picture out of that!

I opened the grey box, and there's an attempt at a 1/4 stub BalUn, but shouldn't that be shorted at then end? (It wasn't!)  I ripped that out, and have now got a decent signal with the "Yagi" pointing in a completely random direction, up and away from the Tx.

So I will now take your advice, chuck it away, keep the SMA tail, and make my own.

Any tips please Sir?   

[PA0PBZ]

I opened the grey box, and there's an attempt at a 1/4 stub BalUn, but shouldn't that be shorted at then end? (It wasn't!)  I ripped that out, and have now got a decent signal with the "Yagi" pointing in a completely random direction, up and away from the Tx.

A balun is like it says to convert from balanced (the dipole) to unbalanced (the coax). If you take that out you will completely destroy the radiation pattern because the coax now is part of the dipole. So your observation of random direction is correct.

[Delta]

It was already random directioning before I touched it!

But should the stub balun be shorted for it to work properly?

[PA0PBZ]

If it is a folded dipole the impedance will be about 300 Ohm, so you will need a 4:1 balun to connect it to 75 Ohm coax, like this:



If you want to calculate the length go here:

http://www.n-lemma.com/calcs/dipole/balun.htm

Find out the velocity factor of the coax used, or assume it's 0.66 if you don't know the type.

[AlessandroAU]

Of course not, the point was to demonstrate the horrible construction quality. Regardless, if you read the site I linked you will see that person examines the common 2.4ghz antennae as well.

Personally, I got no better performance on my eBay yagi than the standard 5dbi dipoles I was trying to improve on.

Edit: Wording

So I will now take your advice, chuck it away, keep the SMA tail, and make my own.

Any tips please Sir?   

I'm a really a big fan of circularly polarised antennas for analogue video because it helps with multipath issues! I would give a helical antenna a try: http://www.w1ghz.org/antbook/conf/Helical_feed_antennas.pdf you can find alot of designs on the internet. I am also a fan of the cloverleaf antenna if you're after a omni circularly polarised antenna. http://www.rcgroups.com/forums/showthread.php?t=1388264

[AlessandroAU]

Of course not, the point was to demonstrate the horrible construction quality. Regardless, if you read the site I linked you will see that person examines the common 2.4ghz antennae as well.

Personally, I got no better performance on my eBay yagi than the standard 5dbi dipoles I was trying to improve on.

Edit: Wording

So I will now take your advice, chuck it away, keep the SMA tail, and make my own.

Any tips please Sir?   

I'm a really a big fan of circularly polarised antennas for analogue video because it helps with multipath issues! I would give a helical antenna a try: http://www.w1ghz.org/antbook/conf/Helical_feed_antennas.pdf you can find alot of designs on the internet. I am also a fan of the cloverleaf antenna if you're after a omni circularly polarised antenna. http://www.rcgroups.com/forums/showthread.php?t=1388264

Biquad antennas also work pretty well and have a small form factor http://martybugs.net/wireless/biquad/ I have used that design on a router and it had excellent performance.

By the way, have you tired something as simple as a parabolic reflector behind the standard dipole antenna? Something like this: http://www.freeantennas.com/projects/template2/

[Delta]

By the way, have you tired something as simple as a parabolic reflector behind the standard dipole antenna? Something like this: http://www.freeantennas.com/projects/template2/

Well much to the Missus's confusion, I got out the scissors and glue, and made a windsurfer...

Fecking feck me with a fecking barge pole, the picture is now PERFECT! 

The engineer in me thought "well, it's a metallic parabolic antenna, this should work", but the bloke in me said "don't be daft, how can something you've made out of card and foil possibly do anything?"

Wow, very impressed, thank you Alessandro! 

[Delta]

I think I've got the radio bug now, so I've going to have a go at making a Biquad tomorrow.  Are the dimensions of the reflector critical?  I've got a few bits of plate kicking around but they are bigger than the 123x123mm (one wavelength) given in the link, and I've seen pictures of them made with all manner of reflectors...

[AlessandroAU]

By the way, have you tired something as simple as a parabolic reflector behind the standard dipole antenna? Something like this: http://www.freeantennas.com/projects/template2/

Well much to the Missus's confusion, I got out the scissors and glue, and made a windsurfer...

Fecking feck me with a fecking barge pole, the picture is now PERFECT! 

The engineer in me thought "well, it's a metallic parabolic antenna, this should work", but the bloke in me said "don't be daft, how can something you've made out of card and foil possibly do anything?"

Wow, very impressed, thank you Alessandro! 

Awesome! haha, isn't it great when some silly simple method works so well! I made a fancier one out of wood and a soda can for the router at my parent's house. Its a pretty easy way to add some directionality. 

I think I've got the radio bug now, so I've going to have a go at making a Biquad tomorrow.  Are the dimensions of the reflector critical?  I've got a few bits of plate kicking around but they are bigger than the 123x123mm (one wavelength) given in the link, and I've seen pictures of them made with all manner of reflectors...

I don't think so, I built one with a 120mm x 120mm PCB and it worked quite well. The most important dimension is the distance between the bi-quad element and the reflector which is 15-16mm depending on the thickness of the element.

My favourite design that the moment is this 3D printed yagi. The elements are floral pins from a craft shop that have been cut to length and its driven by the existing dipole. The yagi adds about 10dBi of gain to the dipole giving about 11-12 total! Takes minutes to build (if you exclude printing time).

[Seekonk]

I have actually never used these.  Providing these two pictures so you can identify it on ebay.  Note the black coax stub and the brass stubs are the pickup.  Wasn't able to open the loop antenna.