Antenna - Theory .com: The Antenna Site

Antenna Basics Antenna Types Smith Charts Antenna Measurements

antenna-theory.com :: View topic - Frustrated
antenna-theory.com Forum Index antenna-theory.com
Antenna Theory
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Frustrated

 
Post new topic   Reply to topic    antenna-theory.com Forum Index -> Questions Related to Antenna Theory
View previous topic :: View next topic  
Author Message
dbvanhorn
Antenna-Theory.com Newbie


Joined: 31 Jul 2015
Posts: 7

PostPosted: Fri Jul 31, 2015 3:20 pm    Post subject: Frustrated Reply with quote

Ah the fun of real life antenna design. No dipoles in vacuum here, messy real world stuff.

Here's my conundrum:

I have a four layer PCB, in plastic enclosure, with about 12mm free above the PCB and almost no room below. The enclosure and feet put the bottom of the PCB about 4mm above whatever the device is sitting on.

I have a corner of the board for an antenna, 57 x 25mm.
Frequency of operation is 2.4-2.5G, and I need better than 12dB return loss to prevent harmonics from the radio chip.

I have a broadband antenna that's working out pretty well, provided the device is sitting on a wood shelf, but of course I have to allow for it to be sitting on something like a DVD player or other device with steel case.
When that case comes into the picture, the broadband goes away, I get sharp resonances, and of course this is all highly variable.

I was hoping to find something like a patch that I could use where the material under the antenna wouldn't matter so much, but my available area isn't large enough.

I really need to stick to a PCB track antenna.

I think my only real option to minimize the effect of the metal under the PCB is to flood the bottom layer of the PCB with ground.

Is there some antenna type out there that can sit on top of a ground plane on the other side of an 0.062" FR-4 board and radiate reasonably well?
Back to top
View user's profile Send private message
admin
Site Admin


Joined: 03 Jan 2007
Posts: 210

PostPosted: Fri Jul 31, 2015 6:42 pm    Post subject: Reply with quote

Good question, but yours is a difficult problem. What's the overall size of your PCB by the way?

I don't think you really need 12 dB return loss for harmonics though, you could filter that out easily enough. What radio is it?

Are you measuring the antenna with a VNA and then watching it shift when placed on metal (i.e. is this where the "sharp resonances" comes from, or what are you referring to)?

Short story: if you have a horizontal planar antenna, no matter what you do, you will have an "image antenna" below the surface which will hurt your bandwidth/efficiency/impedance
Back to top
View user's profile Send private message Send e-mail
dbvanhorn
Antenna-Theory.com Newbie


Joined: 31 Jul 2015
Posts: 7

PostPosted: Fri Jul 31, 2015 7:01 pm    Post subject: Reply with quote

Board size is 96x96mm with large rounded corners.

Experience with this chip is that if the return loss gets worse than 12dB, the harmonics rise fast. We don't have any real data on how fast, but it's a cliff we don't want to walk off. We've been using 12dB as our cutoff.

I could mask it with a pad, but that costs range..

I'm in sim (xfdtd) with this design, the models track reality pretty well in our other products.

Image antenna: Yup.. They were going to use a magic chip antenna, or a piece of wire cut to "a quarter wave" and just stuffed into the box.
Back to top
View user's profile Send private message
dbvanhorn
Antenna-Theory.com Newbie


Joined: 31 Jul 2015
Posts: 7

PostPosted: Fri Jul 31, 2015 8:33 pm    Post subject: Reply with quote

So am I right in thinking that the most practical answer to reduce sensitivity to the objects below my device, is to flood the entire bottom layer with copper, and then to use some antenna structure suspended up above the PCB, like a PIFA or similar?

I didn't want to abandon trace antennas without looking for some less obvious variant that might work.
Back to top
View user's profile Send private message
admin
Site Admin


Joined: 03 Jan 2007
Posts: 210

PostPosted: Fri Jul 31, 2015 9:50 pm    Post subject: Reply with quote

I'm no fan of simulations...too many engineers use it as an excuse not to think. And then they can't figure out what to do when the built product doesn't work and of all things they go back to the simulation to figure out what's wrong....but that's another rant for another day.

For the harmonics, can't you just use a low pass matching network (shuntL series C) to help with that?

PIFA or monopole would definitely be better. Basically if you can get the antenna to have vertical polarization, you're going to be much better off in regards to the surface effect. 96mm should be sufficient size ground to get a patch/pifa going in there, but the monopole would be preferred.
Back to top
View user's profile Send private message Send e-mail
dbvanhorn
Antenna-Theory.com Newbie


Joined: 31 Jul 2015
Posts: 7

PostPosted: Tue Aug 04, 2015 3:29 pm    Post subject: Reply with quote

I hear ya on sims. I'm not a huge fan, but it does let me get pretty close before I commit to hardware, and the results seem to track pretty well with built hardware.

I can filter out harmonics, but it's better not to make them in the first place.

I don't have the vertical space for a monopole, only about 12mm above the board. A PIFA is a possibility, but the cost of the metal stamping will be an issue.

I'm getting close with a broadband antenna that looks like an apostrophe, 12dB return loss from 1.5-2.7G if the system isn't sitting on a metal box. Realized gain plot looks like an apple and peak gain is about 6dBi.

I still need to reduce the effect of metal below though. When I add a DVD player instead of a wood shelf, there are a bunch of narrow resonances, and the broadband response around 2.4 is just shot to hell.

Unfortunately, the sales folks decide what the product will look like, and how much (if any) room I'll have for an antenna.
Back to top
View user's profile Send private message
helix
Antenna Theory Regular


Joined: 29 Jan 2015
Posts: 64

PostPosted: Sat Aug 22, 2015 5:11 pm    Post subject: Reply with quote

How can you be "not a huge fan" of sims when they get you "pretty close"?!

Simulation is an indispensable tool for antenna engineers. We are lucky to have a relatively simple theory (confirmed by 150 years of experiment) addressed by many outstanding software tools to help us understand ideas and develop designs. Things like fluid flow, plasma, and chemical reactions are a lot more difficult physics to handle computationally.

There is a saying -- "No one believes a simulation except the guy that ran it, everyone believes a measurement except the guy that took it." The point is, everyone has been burned by a bug in a simulation code that gave ridiculous answers, BUT measurements are not perfect, either, and only the guy that took the measurement appreciates what the problems are. In particular, the effect of your VNA cables can have major influence on certain antenna measurements. Meaning the measurement does not reflect what's going on when the VNA cable is disconnected (I assume you're not shipping products with VNAs included in the box). So when measurement and sim don't match, you must suspect BOTH are off, and the truth lies somewhere between.


Last edited by helix on Sat Aug 22, 2015 5:30 pm; edited 2 times in total
Back to top
View user's profile Send private message
helix
Antenna Theory Regular


Joined: 29 Jan 2015
Posts: 64

PostPosted: Sat Aug 22, 2015 5:28 pm    Post subject: Reply with quote

As for your antenna, I would look at a normal mode helix. This is essentially a monopole that has been coiled up. It needs a ground plane, so put that on your PCB. That way when your device is placed against metal (at least on that one side), there's very little difference.

Check this product out, it almost solves your problem, is only 4mm tall, and is in-stock at Digikey:

http://www.molex.com/webdocs/datasheets/pdf/en-us/0479480001_ANTENNAS.pdf

With some matching networks of your own, you may be able to achieve your BW and S11 requirement. You might be able to find something a little taller (you said you have up to 12mm) that works even better. When you're working off a ground plane height is of course key. A few places sell what are essentially special-purpose wire springs to use as antennas.

Heck, you could even use a battery contact (10 cents each!) as a normal mode helix:

http://www.memoryprotectiondevices.com/datasheets/BCAAA-datasheet.pdf

You will need to do some testing or sim to determine the S11 and how to resonate/match at 2.4GHz... If the helix/monopole is a little above its first natural resonance (i.e., it's a little "long" at 2.4GHz -- total wire length is a bit greater than lambda/4), it will look inductive and you can use cheap/small/low-loss NPO capacitors to match it at 2.4GHz.


For a PIFA, which is similar to a monopole and also needs a ground plane on your PCB, you could so something like:

http://www.te.com/usa-en/product-1513150-1.html



OR -- build your own based off these ideas.
Back to top
View user's profile Send private message
dbvanhorn
Antenna-Theory.com Newbie


Joined: 31 Jul 2015
Posts: 7

PostPosted: Fri Sep 11, 2015 2:14 pm    Post subject: Reply with quote

[quote="helix"]How can you be "not a huge fan" of sims when they get you "pretty close"?!

Somehow, the computer lends an air of infallibility. It's been a problem since there have been computers.

Highly precise answers but not always accurate. Smile

I use RF simulation a lot, but I hope that I'm appropriately cautious.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    antenna-theory.com Forum Index -> Questions Related to Antenna Theory All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group