|
|
Ceramic capacitor cost reduction - RF Cafe Forums
|
Mike_From_Plano
|
Post subject: Ceramic capacitor cost reduction
Posted: Sun Jul 25, 2010 2:30 pm
|
|
Joined: Mon Apr 20, 2009
2:04 pm Posts: 14 |
Hello again, Been awhile since I've been
on the forum, and in that time, I had to take a
new job. Instead of designing cutting edge products
for the semi industry, I'm performing cost reduction
measures.... Easy at first, but now I'm faced
with the mother of all cost reductions - All the
semis are sucked into an asic, and all that surrounds
it are passives. Anyway, the boss want's his
pennies, and that's what I'm paid for, so here's
the question: This device uses a 430MHz RF
link with one end in the ASIC and the other in a
bottom loaded antenna that's comprised of a descrete
inductor and PCB trace. Between, there's a fourth-order,
LC, underdamped low pass / matching network with
the (~4dB) bump centered on the desired frequency.
The filter network is comprised of discretes, and
the PCB lacks space for much in the way of resonant
networks. I'm curious whether it would be a worthwhile
to use PCB inductors (15-90nH) and run-of-the-mill
capacitors - currently, the designer has Johanson
designed in. Also, I'm curious whether the entire
scheme can be scrapped for a tuned loop antenna
- like you find in car alarm remotes. Thanks,
Mike
|
|
|
|
|
RF Head |
Post subject: Re: Ceramic capacitor cost reduction
Posted: Tue Jul 27, 2010 12:00 pm
|
|
Joined: Mon Feb 09, 2009
12:26 pm Posts: 11 |
Looking at this from a power loss point of view,
a lot will depend on the antenna impedance you are
trying to match and required bandwidth. In the nasty
case of a very low antenna impedance and narrow
band the majority of the power loss in your matching
network will be due to the inductors at 400 MHz.
So from a pure component power loss point of view
you can probably get away with much cheaper capacitors
(I have done so many times). However, the higher
quality (not just higher quality factor) capacitors
are often also much tighter tolerance which may
also be required. If it were me I would
proceed as follows:- Measure the input impedance
of the antenna. Run some simulations of your matching
network to determine where the losses are occuring
in the network and what the affect of looser tolerance
is. If you are lucky you may be able to get away
with cheaper caps. You could also get the figures
you need by measurments of your matching network
by by building the network then connecting it to
another matching network which is designed to take
you back to 50 ohms. You can then mess around with
values etc and see what the relative change in the
power loss through the network is. This is probably
what I would do first. You will be able to see exactly
how the cheaper components compare to the expensive
ones. "PCB inductors" might add too much
loss. However, running a transmission line match
might work. I would look at improving the
input impedance of the antenna. Unless you are really
tight for space it would be much nicer to have a
antenna impedance closer to 50 ohms, you may then
be able to do away with the matching network entirely.
If your antenna is a thin PCB track you may be able
to improve the Zin by adding capacity to the open
end of it, like top loading a short monopole with
a top hat. However, if you try to do this with part
of the end of the track working through FR4 it might
proove to be lossy. Loop antenna might not
be great in terms of performance. ENJOY
|
|
|
|
|
biff44 |
Post subject: Re: Ceramic capacitor cost reduction
Posted: Wed Jul 28, 2010 9:34 am
|
|
Joined: Wed Apr 08, 2009
11:07 am Posts: 33 |
Chip inductors cost a lot. So eliminating them would
be my first priority. You need to set up
a spectrum analyzer, calibrated broadband antennas,
and measure the spurious signals. FCC has limits
to what 2nd, 3rd, etc harmonic levels can be. Since
they redesigned the circuit to be an asic, it is
fairly possible that you now meet the spurious requirements
without the lowpass filter, and can throw away those
5 chip parts. If there is no filter, there
is no need for high tolerance chip capacitors, and
common 2 cent chinese chip caps will do the job.
As far as the bottom loading inductor on the
antenna, you need that, AND it needs to be high
Q (or you will sacrifice transmit range--there is
a lot of RF current resonating in it). If you have
the size, you can make the antenna bigger and not
use it! If you still need it--just get a bunch of
chip and aircoil inductors, and use the cheapest
one that gives you the same transmit range as the
current high-priced one.
|
|
|
|
|
Mike_From_Plano |
Post subject: Re: Ceramic capacitor cost reduction
Posted: Thu Jul 29, 2010 1:32 am
|
|
Joined: Mon Apr 20, 2009
2:04 pm Posts: 14 |
Hey guys, Thanks. That's some really good
stuff. I especially appreciate the heads up about
the spiral loading coil. I was enamoured with the
cost aspect of this. Being from another age, I find
it difficult to think of 433MHz as a "low frequency,"
but it's a terribly long wavelength to deal with
when you only have a few inches of trace for antenna.
Fortunately, this thing only has to reach a
couple of meters, so the xmit power can be set very
low. I'm thinking of doing as you say and chunking
the bottom loading on the antenna for more antenna
/ top loading. Also, using a single inductor
at the bias-T to serve as current source / tuning
element. Then following with capacitor divider accross
it to act as the other tuning element / impedance
match. The chip has about 300ohm output z, so this
follows naturally to drive a lower z antenna.
With a fairly low Q, I'm thinking the cap tolerance
won't be a terrible issue and I'll still be able
to knock down the second order that I was left as
a present from the single-ended output. Thanks
again guys, Mike in Plano
|
|
Posted 11/12/2012
|
|
|