Finished building a signal generator, but one annoying problem exists.

Other Hardware General Electronics
1

Hi!

I have just finished building a signal generator, but one annoying problem exists.

I get a smal negative voltage. See thread Gets strange voltage levels when converting 12V square wave to 5V - General Electronics - Arduino Forum
The answer I got was that the negative voltage was a "glitch", nothing to care about.
Sure I can live with a tiny negative voltage as long the signal is 0-5V.

The problem is that my signal generator have two outputs.
One fixed 0-5V and a second adjustable 0-12V.

That small "glitch" get amplified to a large negative "spike" by my op-amp.

I don't know how to get rid of the negative "spike".
The output of my signal generator IC (XR2206) is open collector.

Is it a bad design to pass a open collector output to a mosfet?

output.jpg
output.jpg1024×768 42.8 KB

scope.bmp (219 KB)

2

Is your scope probe properly compensated? I only see a -160mV minimum
which could easily be a poorly setup up scope probe. nFETs cannot pull below ground!

3

I followed the manual how to test the probe and the reference signal was perfect.

I can not understand where the negative voltage comes from, I also get >5V and that is also strange.

So it should be ok to "level shift" the signal the way I do?

I can not get a screen capture of my oscilloscope right now but found a picture when googling that looks similar after the signal get amplified.

Also the sine/tri, wave (different output pin) look good.

scope2.png

4

If you have a capacitive load it will damp the peak of your square wave.

5

Hi, if your amplifier that is going to get you to 0-12V is only using gnd and a positive supply then the negative bit should be clamped by the amplifier.
It will not get amplified.

Tom... :slight_smile:

6

Looks to me that your scope probe needs tuning
Connect it to a known square wave and adjust the input capacitance (small screw I'm probe) until the picture is a proper square wave
You are getting overshoot due to the probe

7

I tried to adjust the probe. The result was a little better.

I also changed the pullup resistor that went to 5V to 47ohm. That helped a lot.
The signal was much better.

Some small spike remains, maybe my probe is not perfect.

Tom: I have a negative power supply. I am using a ICL7660 to get -5. My op-amp did not go down to 0V without the negative supply.

By the way: I have posted the finished function generator in the gallary.
http://forum.arduino.cc/index.php?topic=220331.msg1604096#new
It has been a fun project for a software guy like me, I learned a lot.