I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Thanks in advance!
Dave

dave.harper wrote:
I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Use a 4017 counter, plus resistors.

Or, there are some nice 5-pole analog filters out there. Not cheap, but
they'll do ya.

You could implement an up/down counter in a PAL. Doing it so that it
up/down counts properly shouldn't be too hard.

Or use one of the myriad itty bitty processors out there -- PIC, AVR,
whatever.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com

On Tue, 23 Mar 2010 12:39:59 -0700 (PDT), "dave.harper"
wrote:

I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like.

A lowpass filter is appropriate. You're trying to kill the higher
(odd) harmonics but there's no signal below the fundamental, excepting
DC maybe.

A dual-stage RC lowpass would be pretty good, if you set the -3 dB
point around 2 KHz. A higher-order filter, active or LC, would be even
better. An LC filter isn't unreasonable at 2 KHz.

John

On Mar 23, 3:39*pm, "dave.harper" wrote:
I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). *It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. *One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. *I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). *So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Thanks in advance!
Dave

Start with a triangle wave, there are less harmonics to filter out.
How bad is the EMI? I like the Wein-bridge. Try using an FET opamp
for it. The only problem I've had with the wien bridge was motor-
boating of the frequency. This was finally solved by using good low
dissipation caps. (Ohh I have seen EMI fuzz.. but only with BJT
opamps.)

George H.

"dave.harper" wrote in message
...
I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Thanks in advance!
Dave

One possibility is to use a digital filter. These have very sharp roll-offs
(e.g. 8th order for MAX292), you basically feed the clock of the filter with
a multiple of the fundamental frequency of the filter (say 64x), and the
output is a very pure sine wave. The beauty of this is the amplitude of the
output is independent of frequency and you don't have to change filter
components:
http://electronicdesign.com/content....nd-measurement

Another solution is using a multiplexor to create an approximate sine wave
using an 8x clock followed by a filter (the example below uses a pair of
cascaded 2nd order filters). This may suit you better:
http://www.maxim-ic.com/app-notes/index.mvp/id/21

Mark.

On Mar 23, 3:39*pm, "dave.harper" wrote:
I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). *It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. *One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. *I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). *So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Thanks in advance!
Dave

Function generator. Integrate square wave to get triangle wave. Shape
triangle wave to sine wave using a couple of diodes. See

http://kitsrus.com/projects/k23.pdf

Dan

Function generator. Integrate square wave to get triangle wave. Shape
triangle wave to sine wave using a couple of diodes. See

http://kitsrus.com/projects/k23.pdf


** Contrary to the OP's wishes, that POS uses op-amps.

Have you seen the ugly " sine" wave it produces ??

The diodes will rectify the RF energy.

Not even a starter.



.... Phil

On Mar 23, 8:01*pm, "Phil Allison" wrote:


Function generator. Integrate square wave to get triangle wave. *Shape
triangle wave to sine wave using a couple of diodes. *See

http://kitsrus.com/projects/k23.pdf

** Contrary to the OP's wishes, that POS uses op-amps.

*Have you seen the ugly " sine" *wave it produces ??

*The diodes will rectify the RF energy.

* Not even a starter.

... *Phil

Contrary to what you think, the OP seems like he would be happy with a
sine wave with only 1% distortion. The diodes will rectify RF but at
the output the signal level should be several orders of magnitude
higher than the RF. So would not expect that to be a problem. The
reference to the kitsrus was to show one implementation that had an
explanation of the circuit.


Dan

"dave.harper" wrote in message
...
I'm trying to come up with a filter or converter that takes a square
wave input and outputs something closer to a sign wave (varying
between ~1.9kHz and 2.1kHz). It can have some distortion, but I'm
trying to eliminate the sharp leading and trailing edge. One option
I'm pursuing is a bandpass filter (2 caps and 2 resistors), which
looks to give a reasonable output, but still not quite as smooth as
I'd like. I've also considered using a counter feeding a bank of
resistors, but finding a method for it to start counting up with it
hits 0 and down when it hits the high value might be more difficult.

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Any suggestions?

Thanks in advance!
Dave

I would make a five pole elliptic low pass active filter breaking at about 3
kHz with a deep 80 dB notch at the third harmonic of 6 kHz. This is a dual
op amp (2 section) solution and uses a small number of resistors and caps.
You should be able to attenuate all of the harmonics at least 80 dB leaving
a pretty good sine wave.

dave.harper schrieb:

Originally, I had used a wein-bridge oscillator with op-amps to make
the sine wave, but due to the environment, EMI was a problem and it
damped out the oscillations (even with ferrite beads and modest
shielding). So I'd like to avoid using op-amps (since they are
apparently sensitive to EMI) if possible.

Hello,

a Wien-Robinson bridge oscillator needs a good amplitude stabilisation.
If the amplification is to low, the oscillations will damp out. I guess
there was something wrong with your amplitude stabilisation.

Bye