hello everyone, i need help on the best resistor values to use in adjusting a bandwidth. for cell phone operating frequencies
pilas

Posts: 7
Joined: Wed Dec 14, 2011 3:32 pm

What hardware are you using? What is it operating at, and what are you trying to change it to? etc.
MrGlass

Posts: 90
Joined: Tue Jun 03, 2008 3:50 am

ok thank you.
am working on my final year project. building a signal neutralizer.
from my research and some reviews including ladyada's work on wave bubble, am suppose to design a high pass filter that will work as the bandwidth adjustment component of the device.

well my problem is the best resistor and capacitor values to use to serve the purpose of frequency bands between 800mhz to 1900mhz. plus, i read from a certain literature about how the bigger the bandwidth the larger the noise generation.

i have tried various combination of RC to determine all sorts of cutoff frequencies but my simulation on (multisim - spectrum analyser) still shows up lower frequencies.

thank you.
pilas

Posts: 7
Joined: Wed Dec 14, 2011 3:32 pm

Still need more info. What filter topology are you using? How many stages? Active/passive? Or is all that up to you to decide? Are you familiar with poles and zeroes on the complex plane?

The subject of filter design can involve some fairly heavy calculus. A starting point might be Wikipedia, http://en.wikipedia.org/wiki/Network_synthesis_filters . There are links there to various types of filters: Butterworth, Chebyshev, etc. But Wikipedia mostly gives an encyclopedia-type treatment of the subject. To really make the choices needed to design one for a particular purpose, you may need something closer to a textbook.
uoip

Posts: 127
Joined: Wed Sep 23, 2009 5:48 pm

yes i am very family with filters, am not designing a digital filter, so i guess i dont have to worry about poles and zeros. its an analog filter. ie a passive analog filter.

upon reviewing ladyada's circuit, her bandwidth adjustment component, was a high pass filter in parallel with a resistor, which i guess will be a variable resistor. i have attached that part of ladyada's circuit below
samp.JPG (31.79 KiB) Viewed 2521 times
.

thank u very much.
pilas

Posts: 7
Joined: Wed Dec 14, 2011 3:32 pm

Poles and zeroes happen in purely analog passive filters. For example, a single stage RC filter has a pole at s=-1/RC. If used as a high pass filter (if the output is taken across the resistor), it has a zero at the origin.

Generally, when looking at RC circuits, the product of R*C is the time constant, and it's what you use if you want to change the corner frequency of a filter.

Your intial post suggested that undesired frequencies were still getting through after you adjusted the corner frequency. That could be a problem that the cutoff isn't sharp enough. If that's the issue, you may need more filter stages (a higher order filter).
uoip

Posts: 127
Joined: Wed Sep 23, 2009 5:48 pm

thanks man for that enlightenment.
so by more stage, do u mean, i cascade several high pass together or something.

also do i use that same zero pole diagram concept i know for digital filters to determine the stability of the filter or some other way?

thanks
pilas

Posts: 7
Joined: Wed Dec 14, 2011 3:32 pm

Yeah, you can cascade multilple stages together. Some background:

http://alignment.hep.brandeis.edu/Lab/F ... ilter.html

A page with a bunch of applets to illustrate various filter types

The pole/zero concept is strictly a mathematical concept that applies regardless of how the filter is implemented. It works for digital, analog electronic, and even analog mechanical filters.
uoip

Posts: 127
Joined: Wed Sep 23, 2009 5:48 pm

If that's a screen shot of the Wavebubble then I believe you have the wrong end of the stick. A Wavebubble works by having 2 Voltage Controlled Oscillators (VCOs). These are typically tuned by varying the tuning voltage between 0V and 28V, this tuning voltage is set by a Phase Locked Loop (PLL). A PLL knows the desired frequency, it measures the output frequency of the VCO and then adjusts the tuning voltage until the output frequency is correct. It also corrects for phase making it rather handy for most radio applications.

The bandwidth of a Wavebubble is constrained by the VCOs. e.g. the Minicircuits ROS-2500+ has a bandwidth of 1600MHz to 2500MHz, the Minicircuits ROS-1300+ has a bandwidth of 400MHz to 1300MHz. Then within these ranges the PWM (a poor mans DAC alternative) and a digital pot control the DC-offset of a triangle wave generator made with an NE555 timer.

If you want to design a filter that stops mobile phone frequencies then this is not a Wavebubble related question. Oh and you definitely still need to worry about poles and zeros with analog filters. The Texas Instruments Filter Designer: "http://www-k.ext.ti.com/srvs/cgi-bin/webcgi.exe?Company={5761bcd8-11f5-4e08-84e0-8167176a4ed9},kb=analog,case=obj(26717),new" (sorry, can't get the link to work) is an online tool that may help. Although you're semi doomed as mobiles can go up to 2GHz which gets into stripline filters.
TheFallen

Posts: 92
Joined: Mon Feb 08, 2010 1:28 pm
Location: UK