RF impedance matching PCB layout / analysis?

[Jesse Lackey]

Hello all,

I've managed to get myself into a pickle with a client project.
I'm doing some work with Zigbee (2.4-2.5Ghz band) and am having RF problems.

I should preface this by saying I know little about RF, but have a fair
amount of experience in a bunch of other areas, I'm not new to
electronics but am definitely new to RF.

The specific problem is transmitted power is 50dBm lower than expected
(-60 vs. -10).  This is very likely because (due to space necessity) the
design isn't an exact copy of the suggested RF output circuit/layout
from the manufacturer of the balun (anaren) I'm using.

Rather than getting into the details of it all, I'd like to educate
myself in this area and do some analysis.  This issue will come up again
for this client so it is well worth the time.

This is where the piclist wisdom comes in!

What I need to be able to do now and in the future is to match a given
antenna to the RF output of the various low-power RF chips out there.
In this case it is a TI CC2430.  So the datasheet for that chip
specifies the differential output as being 115+J180.  TI does some
trickery with long traces and passives to match to a 50-ohm antenna;
this works, but is considerable board space.  Anaren makes a balun and
has an appnote on how to use it with this part, and goes through an
analysis of their design.  This is great, however I don't know:

a) how to go from physical pcb trace shapes to the equivalent m+Jn for
each segment
b) the larger concept of what is going on here, and how to take each pcb
trace segment and various caps/inductors in series to shunted to ground
and plot them on a smith chart to wind up at 50 ohms for the antenna.

Anyone still with me?

I know all this is a big subject, and good software is probably big
bucks, but tips of all sorts would be helpful.  A book (or chapter in a
book, even), software free to kilobucks, etc.  My client has some money
to spend to solve all this, I just don't know where to begin looking.

Thanks everyone-
J

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[PAUL James-3]

Jesse,

The first place I'd look is at an Amateur Radio Handbook.  These books
are full of stuff such as this.
I am a ham (Extra Class), and I have a number of these books.   I'll
look at one tonight and see if I can
Find anything helpful to you.  

I also have a book on PCB's for RF functions.   I'll look there too.

If I find anything, I'll let you know tomorrow.   I hope this will help
you out.



       
Regards,

       
Jim

-----Original Message-----
From: piclist-bounces@... [mailto:piclist-bounces@...] On Behalf
Of Jesse Lackey
Sent: Wednesday, July 23, 2008 4:40 PM
To: Microcontroller discussion list - Public.
Subject: [EE] RF impedance matching PCB layout / analysis?

Hello all,

I've managed to get myself into a pickle with a client project.
I'm doing some work with Zigbee (2.4-2.5Ghz band) and am having RF
problems.

I should preface this by saying I know little about RF, but have a fair
amount of experience in a bunch of other areas, I'm not new to
electronics but am definitely new to RF.

The specific problem is transmitted power is 50dBm lower than expected
(-60 vs. -10).  This is very likely because (due to space necessity) the
design isn't an exact copy of the suggested RF output circuit/layout
from the manufacturer of the balun (anaren) I'm using.

Rather than getting into the details of it all, I'd like to educate
myself in this area and do some analysis.  This issue will come up again
for this client so it is well worth the time.

This is where the piclist wisdom comes in!

What I need to be able to do now and in the future is to match a given
antenna to the RF output of the various low-power RF chips out there.
In this case it is a TI CC2430.  So the datasheet for that chip
specifies the differential output as being 115+J180.  TI does some
trickery with long traces and passives to match to a 50-ohm antenna;
this works, but is considerable board space.  Anaren makes a balun and
has an appnote on how to use it with this part, and goes through an
analysis of their design.  This is great, however I don't know:

a) how to go from physical pcb trace shapes to the equivalent m+Jn for
each segment
b) the larger concept of what is going on here, and how to take each pcb
trace segment and various caps/inductors in series to shunted to ground
and plot them on a smith chart to wind up at 50 ohms for the antenna.

Anyone still with me?

I know all this is a big subject, and good software is probably big
bucks, but tips of all sorts would be helpful.  A book (or chapter in a
book, even), software free to kilobucks, etc.  My client has some money
to spend to solve all this, I just don't know where to begin looking.

Thanks everyone-
J

--
http://www.piclist.com PIC/SX FAQ & list archive View/change your
membership options at http://mailman.mit.edu/mailman/listinfo/piclist

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Sean Breheny]

Hi Jesse,

Did I understand properly that you think the TRANSMIT power is too
low? How are you measuring it?

When you say that the manuf specs the differential output at 115+J180,
is this the output impedance of the transmitter which you need to
match to the antenna?

What kind of antenna are you using?

There are several levels of skills which could be involved here. One
is overall RF link calculations. Another is basic antenna theory and
deviations from it in typical practice with microwaves. Another is the
set of skills needed in working with RF ICs (i.e., what specs are
important, which ones are not so important, what numbers can be relied
upon vs. which ones must you allow a good tolerance for). Another, as
you pointed out, is impedance matching and transmission lines.
Depending upon what problem you are having, its cause could lie in any
one of these areas.

Sean


On Wed, Jul 23, 2008 at 5:39 PM, Jesse Lackey <jsl-ml@...> wrote:

> Hello all,
>
> I've managed to get myself into a pickle with a client project.
> I'm doing some work with Zigbee (2.4-2.5Ghz band) and am having RF problems.
>
> I should preface this by saying I know little about RF, but have a fair
> amount of experience in a bunch of other areas, I'm not new to
> electronics but am definitely new to RF.
>
> The specific problem is transmitted power is 50dBm lower than expected
> (-60 vs. -10).  This is very likely because (due to space necessity) the
> design isn't an exact copy of the suggested RF output circuit/layout
> from the manufacturer of the balun (anaren) I'm using.
>
> Rather than getting into the details of it all, I'd like to educate
> myself in this area and do some analysis.  This issue will come up again
> for this client so it is well worth the time.
>
> This is where the piclist wisdom comes in!
>
> What I need to be able to do now and in the future is to match a given
> antenna to the RF output of the various low-power RF chips out there.
> In this case it is a TI CC2430.  So the datasheet for that chip
> specifies the differential output as being 115+J180.  TI does some
> trickery with long traces and passives to match to a 50-ohm antenna;
> this works, but is considerable board space.  Anaren makes a balun and
> has an appnote on how to use it with this part, and goes through an
> analysis of their design.  This is great, however I don't know:
>
> a) how to go from physical pcb trace shapes to the equivalent m+Jn for
> each segment
> b) the larger concept of what is going on here, and how to take each pcb
> trace segment and various caps/inductors in series to shunted to ground
> and plot them on a smith chart to wind up at 50 ohms for the antenna.
>
> Anyone still with me?
>
> I know all this is a big subject, and good software is probably big
> bucks, but tips of all sorts would be helpful.  A book (or chapter in a
> book, even), software free to kilobucks, etc.  My client has some money
> to spend to solve all this, I just don't know where to begin looking.
>
> Thanks everyone-
> J
>
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Dennis Crawley-2]

Dear Jesse,
microstrip antenna design handbook
IEEE microstrip antenna technology
RFID handbook Fundamentals as Applications in contactless samrt cards.

...but it is a science appart.
I know nothing about, ...but I have the phone who knows :)


On Wednesday, July 23, 2008 6:39 PM [GMT-3=CET],
Jesse Lackey  wrote:

> Hello all,
>
> I've managed to get myself into a pickle with a client project.
> I'm doing some work with Zigbee (2.4-2.5Ghz band) and am having RF
problems.

>
> I should preface this by saying I know little about RF, but have a fair
> amount of experience in a bunch of other areas, I'm not new to
> electronics but am definitely new to RF.
>
> The specific problem is transmitted power is 50dBm lower than expected
> (-60 vs. -10).  This is very likely because (due to space necessity) the
> design isn't an exact copy of the suggested RF output circuit/layout
> from the manufacturer of the balun (anaren) I'm using.
>
> Rather than getting into the details of it all, I'd like to educate
> myself in this area and do some analysis.  This issue will come up again
> for this client so it is well worth the time.
>
> This is where the piclist wisdom comes in!
>
> What I need to be able to do now and in the future is to match a given
> antenna to the RF output of the various low-power RF chips out there.
> In this case it is a TI CC2430.  So the datasheet for that chip
> specifies the differential output as being 115+J180.  TI does some
> trickery with long traces and passives to match to a 50-ohm antenna;
> this works, but is considerable board space.  Anaren makes a balun and
> has an appnote on how to use it with this part, and goes through an
> analysis of their design.  This is great, however I don't know:
>
> a) how to go from physical pcb trace shapes to the equivalent m+Jn for
> each segment
> b) the larger concept of what is going on here, and how to take each pcb
> trace segment and various caps/inductors in series to shunted to ground
> and plot them on a smith chart to wind up at 50 ohms for the antenna.
>
> Anyone still with me?
>
> I know all this is a big subject, and good software is probably big
> bucks, but tips of all sorts would be helpful.  A book (or chapter in a
> book, even), software free to kilobucks, etc.  My client has some money
> to spend to solve all this, I just don't know where to begin looking.
>
> Thanks everyone-
> J

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Peter Loron]

Hi, Jesse. I'd much appreciate getting the information you get on your  
questions, as I'm starting in on some ZigBee stuff, and will likely be  
using the CC2430 in my project.

Thanks!

-Pete

On Jul 23, 2008, at 2:39 PM, Jesse Lackey wrote:

> Hello all,
>
> I've managed to get myself into a pickle with a client project.
> I'm doing some work with Zigbee (2.4-2.5Ghz band) and am having RF  
> problems.
>
> I should preface this by saying I know little about RF, but have a  
> fair
> amount of experience in a bunch of other areas, I'm not new to
> electronics but am definitely new to RF.
>
> The specific problem is transmitted power is 50dBm lower than expected
> (-60 vs. -10).  This is very likely because (due to space necessity)  
> the
> design isn't an exact copy of the suggested RF output circuit/layout
> from the manufacturer of the balun (anaren) I'm using.
>
> Rather than getting into the details of it all, I'd like to educate
> myself in this area and do some analysis.  This issue will come up  
> again
> for this client so it is well worth the time.
>
> This is where the piclist wisdom comes in!
>
> What I need to be able to do now and in the future is to match a given
> antenna to the RF output of the various low-power RF chips out there.
> In this case it is a TI CC2430.  So the datasheet for that chip
> specifies the differential output as being 115+J180.  TI does some
> trickery with long traces and passives to match to a 50-ohm antenna;
> this works, but is considerable board space.  Anaren makes a balun and
> has an appnote on how to use it with this part, and goes through an
> analysis of their design.  This is great, however I don't know:
>
> a) how to go from physical pcb trace shapes to the equivalent m+Jn for
> each segment
> b) the larger concept of what is going on here, and how to take each  
> pcb
> trace segment and various caps/inductors in series to shunted to  
> ground
> and plot them on a smith chart to wind up at 50 ohms for the antenna.
>
> Anyone still with me?
>
> I know all this is a big subject, and good software is probably big
> bucks, but tips of all sorts would be helpful.  A book (or chapter  
> in a
> book, even), software free to kilobucks, etc.  My client has some  
> money
> to spend to solve all this, I just don't know where to begin looking.
>
> Thanks everyone-
> J
>
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce]

> The specific problem is transmitted power is 50dBm lower than expected
> (-60 vs. -10).  This is very likely because (due to space necessity) the
> design isn't an exact copy of the suggested RF output circuit/layout
> from the manufacturer of the balun (anaren) I'm using.

What design have you sort of copied?

Have you looked at TI AN043 on 2.4GHz Antenna design?
http://focus.ti.com/lit/an/swra117d/swra117d.pdf

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce]

>Hi, Jesse. I'd much appreciate getting the information you get on your
>questions, as I'm starting in on some ZigBee stuff, and will likely be
>using the CC2430 in my project.

A good starting point would appear to be the application notes on the TI
page at
http://focus.ti.com/docs/prod/folders/print/cc2430.html

Specially interesting looking items are AN043 (I already referenced), DN003,
DN007, AN040, AN048 (which appears to be what Jesse needs), AN058 (gives
guidance on selecting suitable antenna type), AN055 (seems to reference what
Jesse currently has), DN001 (details using network analyzer to characterize
antenna), AN003 (gives a heap of fundamentals on antennae and doing them in
PCBs), and AN032 (2-4GHz band regulations).

There are a heap of others there too relating to the CC2xxx family of chips,
but these all look pertinent to the RF side. There are a handful of other RF
ones as well which should probably be investigated to make sure I didn't
skip something important.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist