Could use some advice from any RF Engineers

[MacGyverS2000]

I've been working on a project on and off for a number of months... as I've picked it back up full force recently, I've also decided to nail down as many requirements as possible. I have some questions, and in other areas I would simply like a "That should work reasonably well" or "I suggest this other method" type of answer.

For the purposes of this post, I would like to add an RF remote control. I expect the receiver to use an external antenna for greatest sensitivity (a shielded wire, rather than a whip, though, for size/cost). The transmitter will be a handheld remote, comparable in size to your average multi-appliance TV remote (so I have some decent board area for a PCB-printed loop). Fairly reliable reception should happen at distances of (let's say) 100 feet (although it's not critical if reception drops out). Of course, FCC Part 15 compliance is a must (automotive application).

I have a decent number of papers/books on hand that discuss the various antenna types, equations to determine component values and trace length/size, etc., but they don't always agree with each other.

1) What frequency range should I be designing for? Something in the 800-900 MHz region?

2) Some papers suggest keep traces well away from the entire transmitting antenna assembly, whereas others say it's not a problem to enclose circuitry in the middle of the loop. Real-world circuits tend to prove out the latter, particularly a number of remote keyless entry units I have... they're small units, so you HAVE to put the circuitry in the middle. Any comments?

3) Since the remote is fairly large (compared to RKE handheld units), should I consider a monopole down one side of the unit rather than a loop along the top? I would get the greater field strength with a monopole, as well as an easier antenna design, but then I may have to worry about close traces for the rest of the circuit being affected by RF.

4) I would like to do as much testing as possible myself to avoid the hevay fees involved with someone else stepping in. What equipment would be wise to purchase/rent? In particular, I'm not looking to make the transmitter the best radiator it can be, only to be relatively strong without breaking the FCC rules. Would a frequency analyzer be enough equipment for this? At the very least, I could see relative levels versus frequency... if my environment is quiet enough, I might actually be able to measure absolute levels. Or should I forget all of that and just hire a consultant from the beginning to do the entire design?

Well, that enough ?s for now...

 

[MacGyverS2000]

OK, came up with a couple of other questions...

To keep resistive losses low, I intend to use fairly wide traces (120-150 mil wide), if possible. When calculating loop area, I assume is it usually done using the middle of the trace as the boundary, correct?

Would I be better off going with 2 oz. copper (versus 1 oz.) and thinner traces, or will the difference in resistive losses be so minimal as to be moot with such a small antenna?

 

[VEBill]

The higher the frequency, the easier it will be to make a compact antenna (obviously).

Simulation (or something...) might be critical to ensure that you provide an antenna with good coverage (isotropic, or low gain) - unless you plan to aim it.

You will need to look up the unlicensed bands as applicable to your market(s). I had previously posted a link to a good website - I'll try to find it again.

You only need to worry about antenna loss if you're trying to use a very small antenna on HF (for example) where the feed point impedance drops to milliohms and any resistive losses start to dominate. Assuming that you're able to choose a UHF band (like 900MHz or higher), then this shouldn't be an issue (with respect to copper thickness).

100 foot range sounds easy. There are RKE keyfobs that get close to that range.

I assume that your shielded wire antenna won't be shielded at the antenna end...

If I were you, I would find a suitable chipset and take it from there.

 

[VEBill]

The link for unlicensed bands up to 1 GHz:

http://www.chipcon.com/files/AN_001_SRD_Regulations_below_1GHz_rev_2_01.pdf

There are more unlicensed bands above 1 GHz of course...

 

[nbucska]

If your project needs a small number of units, you should
just buy one "RF/OEM module" e.g. from <

http://www.digikey.com>

No FCC, easy to use, etc...

<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[MacGyverS2000]

For future reference, this is a US-based product, but designing it to also fit within the limits for eventual UK consumption is ideal.

I definitely intend to use a canned solution for the transmitter, so the questions above are more about design of the actual antenna. Due to the nature of the application (handheld item that would be moving around the vehicle), an omnidirectional setup is what I need.

If I'm reading the FCC rules correctly, the only option I have is the 260-470 MHz band (per section 15.231). It seems the 902-928 MHz range is limited to frequency hopping/spread spectrum type of systems. It certainly explains the large number of canned solution in the lower frequency range.

If I'm reading YOUR response correctly, I will need to worry at least some about resistive losses if I'm going to be operating in the lower frequency range.

I suppose my questions come down to will I need to worry heavily about high tolerances on the PCB material? If I'm careful about following the equations (and pay close attention to the real-world adjustments needed for those equations), can I get away with the variations expected in PCB differences from shop to shop (varying copper thickness, etc.) and be able to compensate with standard value components (or at least variable caps and such)? I would like to design the actual antenna once, try to find standard value components to tune it, and use a minimum of variable value components... is this a realistic request?

 

[nbucska]

If you use an off the shelf TX/RX pair, for 100 ft. I don't
think you would have any difficulty.


<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[MacGyverS2000]

Just to clarify... even if I use a canned transmitter chip (for example, a CC1070 by Chipcon), I still need verification of the final product by the FCC, correct?

It would be HUGE for me if I could use one of those units and remove the need for verification, but I believe the only way to do that is to use a unit that includes an entire transmitting setup (i.e., chip AND antenna) where the antenna cannot be swapped out, and those would be cost-prohibitve.

And no, this will not be a small number of units, so every dollar matters...

 

[VEBill]

My understanding is that certification is not done "by the FCC", but rather by an accredited lab.

It's a simple trade-off to compare the cost for build versus the cost to buy. Obviously, it all comes down to quantity.

If you assume small quantities (buy) to start, then if the product becomes a success with very large quantities... ...well, that would be a very nice problem to have.