-
I prototyped a 0-6VACrms AC Sine Wave Generator Module (AC-coupled, bipolar output) shown in the below schematic.
Specs:
POWER SUPPLY (not on schematic): Single Supply 12VDC Regulated with on-PC board -12VDC Charge Pump Inverter, ICL7662. Maximum ripple on +12V rail is 600uV. Maximum ripple on -12V inverter rail is 900uV.
Principle of Operation: (per accompanying schematic & sequence of photos)
CD4047 provides a precise 50% duty-cycle, 100Hz square wave on +12V power so, it is a unipolar output 0-12V.
That output is AC-coupled and level-shifted by OA#1 to swing +/-12V to produce a bipolar 24V Square Wave.
OA#2 Integrates that square wave into a Triangle Wave which is then fed to OA#3.
OA#3 Integrates the triangle wave into a Sine Wave which is fed to OA#4.
OA#4 Amplifies the Sine Wave to the desired 6VACrms (minimum) bipolar output required.
Objectives:
1) Most important, operation from a single, well-regulated +12VDC power supply. (May also be operated from a single 9V battery with slightly reduced Sine Wave amplitude output.)
2) Precise 50% duty-cycle square wave so, CD4047 was chosen. Op Amp square wave oscillators tested are too sloppy and imprecise.
3) Dual integrators provide the easiest, most stabile method for producing a low-frequency, low-distortion sine wave while allowing the use of non-critical component values. Phase shift oscillators and shaping filter schemes are too finicky and too complicated. I like simple.
4) Amplitude stability to 0.5% is very important. It is entirely dependent on the +/- rail voltages so the need for input voltage regulation is critical. An on-PCB MC7812 seems adequate. No temperature compensation or AGC is provided because the parts count goes way up along with the module's complexity and resultant BOM cost.
HELP NEEDED: Open to any suggestions/criticisms/advice/revisions/corrections/re-design, etc. that could make this a better, low-cost product aimed at the student, beginner, hobbyist, DIY Homeowner markets.
What good is this thing and what is it used for? Well, for those millions of people who have bought cheap DMM's and want to know whether their AC functions are working at least on the low voltage AC range(s), an economical AC Sine Wave module may be something they'd be interested in along with an economical DC reference and a precision resistor module.
Hell, they could verify that their cheap little DMM is worthy of trusting around the house, garage, hobby bench or, even coursework at school. Or, maybe if it
s not, buy a real DMM.
By the way, I'm an admirer of the DMM Check product but, it's too good and too expensive for the market I'm speaking of here.
Thanks in advance for looking things over. SWING AWAY....
-
The two integrators are just acting as a second order low pass filter, you can do that easily with just a single opamp:
https://webench.ti.com/filter-design-tool/filter-response
-
This is a duplicate of your earlier thread :
https://www.eevblog.com/forum/projects/make-it-better-sine-wave-generator/msg5394134/#msg5394134 -
4) Amplitude stability to 0.5% is very important. It is entirely dependent on the +/- rail voltages so the need for input voltage regulation is critical.
That's true only of the square wave.
You might like to ponder what parts affect the sine wave amplitude precision.
-
This is a duplicate of your earlier thread :
NO, it is NOT a duplicate of the earlier thread. If you had actually read what I have said in this thread you would realize that it is a new circuit and a new set of questions, etc.
https://www.eevblog.com/forum/projects/make-it-better-sine-wave-generator/msg5394134/#msg5394134
Thank you. -
Thanks for your comment but it doesn't address my concerns at all. I appreciate your hypothesizing but I don't solve my problems by "pondering". I chose my parts by design and for the reasons I stated in my OP.4) Amplitude stability to 0.5% is very important. It is entirely dependent on the +/- rail voltages so the need for input voltage regulation is critical.
That's true only of the square wave.
You might like to ponder what parts affect the sine wave amplitude precision.
Maybe we have a "language barrier" situation here. -
The two integrators are just acting as a second order low pass filter, you can do that easily with just a single opamp:
I tried a few Sallen-Key type filters but they provided distorted sine waves at the amplitude I need so I went with the dual integrators instead which provides the nicely shaped sine waves you see in the attached photos, not the sort of integrator/sharply-peaked sine waves of the dual low pass filters with single op-amp.
https://webench.ti.com/filter-design-tool/filter-response
Thanks for the design link. Maybe I missed something. I'll give the 2nd order filters another try.....
-
How about the Twin-T oscillator? It's suppose to be low distortion.
Simulation here. Schematic below (12V single supply).
Recently, W2AEW #380 and IMSAI Guy #1807 both did videos on them.
EDIT: I've updated simulation and schematic to correct output part.
You'll need to update simulation speed to allow for the voltage after the output cap to settle about 0V after pot change.
-
See Analog Devices, Linear Technology and National Semiconductor app notes esp from the late and great Jim Williams (ADI App note 43) and Bob Pease on low THD audio oscillators. Also the TEK SG505 app notes,, manual and TEKNOTES article by Rich Cabot, in 1970s.
Jim W was measured and optimized several improvements for solid state, FET control and hi output V anf reduced THD below 0.0005%
https://www.analog.com/media/en/technical-documentation/application-notes/an43f.pdf
The Hewlett and Packard 1936 Wein bridge with incandescent lamp (Stanford U Dr Thesis) led to the HP 200 (valve/tube) oscillator, the classic.
The twin tee, phase shift etc are not as low THD and NOT easy to tune over narrow or especially wide range eg 20 Hz..20 kHz.
For a fixed osc eg 1 kHz they are OK
In audio instruments and measurement, the HP/Jim W/Wien bridge is the gold standard.
Modern implementations replace the 1936 incandescent bulb FB regulator with an FET or lamp-CdS photocell.
Just my experience.
Jon
PS: Was old friend of Jim and Bob, since 1980s
We would see bob at the Silicon Valley AARL Ham fleas selling his analog design books. memories...
-
This circuit design is pretty simple. It has only IC 3 cases: 1 digital and 2 linear double opamps.
With this 3 microchips it gives three different wave shapes: square, triangle and sine. That is quite efficient and convenient for beginners I guess. Yes, there are other ways to make a pure sine, but if I needed a functional generator I would look at this circuit. -
As for me, these days, unless this was purely as an exercise/challenge, or the frequency was much too high to make the following practical (or at least reasonably cheap), I would probably just implement it on some MCU followed by an appropriate DAC, and would call it a day.
But, of course, always depends on the context.