# generating sine wave

Discussion in 'Electronic Design' started by pinku, Aug 9, 2006.

1. ### pinkuGuest

Is there any simple method to generate a sine wave of frequency 1MHZ by
using LM324
op-amp/by using filters /by using oscillators (please mention the
specifications and schematic diagrams).

2. ### EeyoreGuest

1 MHz is a bit fast for an LM324 btw ! Would you like to consider another device
?

My hourly rate is \$80 btw.

Graham

3. ### Frank BemelmanGuest

And payment term 24 hours

4. ### Guest

The answer is no, whch isn't going to help get your homework done.

If you look at the data sheet for the LM324

http://www.national.com/ds/LM/LM124.pdf

and check out page 7, you will see that amplfier runs out of small
signal gain at about 1MHz (top of page, left hand side) while the large
signal output swing falls below 1V at 100kHz.

If your instructor was silly enough to specify an LM324 as the op amp
to use for this job you've got an easy answer to the question.
Otherwise you are going to have to start out by finding a faster
amplifier.

How much distortion in the sine wave can you accept? What sort of
amplitude do you need? Does the frequency have to be exactly 1MHz? If
so, tell us what sort of range frequencies would count as "exact" -
+/-10Hz?

How stable do amplittude and frequency have to be?

In the meantime, googling on Wien bridge and phase-shift oscillators
might give you some ideas.

http://hyperphysics.phy-astr.gsu.edu/Hbase/electronic/oscphas.html

http://www.ecircuitcenter.com/Circuits/opwien/opwien.htm

The circuit given above for the Wien bridge will produce a rather
distorted sine wave - there are better versions of the Wien bridge, See
Jim Williams Linear Technology application note AN43 on bridge circuits
- the first half of the application notes talks about Wheatstone
bridges, but the second half gets onto Wien bridges.

5. ### Tony WilliamsGuest

If you have around >12V supply a quick and dirty way of
getting a sinewave in the low MHz-region is with a
tuned cross-coupled long tail pair.

Vs-------------------+
|
+--/////---+---/////--+
| --------L------- |
| C |
+----------||---------+
| |
+--------/\/\/--------+
+---------, ,--------+
| \/ |
\| /\ |/
npn|--/\/\-' '-/\/\--|npn
e/| R R |\e
| |
+----------+-----------+
|
\
/Rtail
\
|
0v------------------+

L is the centre-tapped primary of a transformer and can be
directly as shown, or via a secondary winding.

L+C form a tuned circuit, resonant at 1MHz.

The load on L+C is Rload and 2R in parallel (more or less).

Make the impedance of the C about 1/10 of the effective
collector-collector load then calculate the value of L for
resonance.

Assume a voltage across Rtail of about 1/3 to 1/2 of Vs.
Start off with a tail current of around 1-2mA dc.

A bit quick and dirty, but it oscillates quite readily

6. ### budgieGuest

at least that roll-off helps subdue harmonics.

7. ### Phil HobbsGuest

I don't know that that's true. Given that you can make a Colpitts
oscillator out of an emitter follower plus an impedance transforming
network, it might be sort of fun to do the same with a 324. I bet you
could make a 1 MHz oscillator out of one by using a pi network between
the output and input. Not a very _good_ sine wave oscillator, but all
you need is power gain, not voltage gain.

Cheers,

Phil Hobbs

8. ### TuTGuest

The big problem with the 324 is its slew rate, (about 0.5V/us). At 1MHz,
this would limit the output amplitude to about 160mV p-p for a sinusoidal
waveform.

9. ### Michael BlackGuest

One might as well go to a single transistor, which of course works
fine for the purpose.

This is once again an example of someone stuck in low frequency
mentality trying to extend it to radio frequencies. Time after
time, we see people needing an oscillator at 1MHz or so, and want
to know why their op-amp or function generator IC doesn't work very
well up there. Of course, it's because they are hitting the top
of the frequency range for the device, and while the range may
include 1MHz, that doesn't mean the IC designer every intended good
operation up there.

Yet, they are trained in op-amp and IC solutions, so they can't see
that long before op-amps came along, such an oscillator would be
made with a single active element, tubes in the past, a transistor
in recent decades. The 324 is a stretch at 1MHz, but it's close to
DC for building an LC oscillator with a single active element. It's
all perspective.

Michael

10. ### Rich GriseGuest

With an LM324, no. According to the data sheet:
http://www.national.com/ds/LM/LM124.pdf ,
At 1 MHZ, its gain is unity, so it will not oscillate.
(well, maybe just a little, but it'd be a nightmare trying to get it
stable.)

Sorry,
Rich

11. ### Ken SmithGuest

I haven't tried this but here's an idea to get around that.

Use only one output of the LM324 and make a Colpitts common collector
oscillator by placing an impedance in the ground line of the package.
IIRC the compenation of the LM324 controls the slew rate of the output vs
the ground pin.

12. ### Fred BartoliGuest

Hehe...
I thought about boostrapping the supplies, but this one is a nice weird
idea.