I've been looking for a simple and "good enough" way to generate sine waves that are independent of frequencies. Obviously generating a triangular wave is much easier, and now I can go from triangular to sine via simple circuits. Thank you again.

The germanium diode whistle woke up my dog XD

Tip -= if you want retention start an ep with a demo of the sound of the waveform & scope display, morphing between the two states.

Hi, very interesting to watch. I was looking to understand how does this triangle to sinus conversion works. As I have bought today a second hand old Function generator Voltcraft FG1617, and did some testing right away and have seen that for the highest frequencies ~ 500 - 2MHz (2MHz is max) I could se the harmonics (multiple of the base frequency) on the spectrum analyzer. I also could notice the distortion (looking on the scope) just itsy bitsy before the top (or at the top) of the sinus shape, and just before the lowest point. Nice thing about these old equipment that a lot of them have complete schematics within the service (or even user) manual. So I have found a adjustment potentiometer SIN.DISTORTION and tried to get it better. I could get it some better for those higher frequencies (like 2MHz) but then the lower frequencies got much much worse, the signal looked more as a triangle then a sinus, so I put the potentiometer back where it was. In their SINE SHAPE circuit they use 4 bridge rectifiers. You can find the user manual with all the schematics online, if you want to look into it. And then after those 4 bridge rectifiers the signal goes to a current mirror I suppose, where some DC.ADJ. and AMP.ADJ. can be made. There are also 3 trim-potentiometers with text SIN.A, SIN.B and SIN.C that are connected to 3 of those bridge rectifiers. By the way, they have specified the THD at

@AccidentalScience

Жыл бұрын

Interesting. That instrument is pretty similar to the one I used when I was at school. Nice. At the time I didn't even know how to measure distortion LOL. It's hard to have a wide frequency response and still maintain low distortion. I rather limit the frequency range and have lower distortion, that could come from a range of reasons, including the physical layout of the circuit. The problem comes with the non linear response of the amplifiers, since the waveshaper works at a specific amplitude. Also the very same waveshaper could be affected by frequency as parasitic reactances become more and more significant. It's always a trade-off.

If the input voltage is variable then we can change the triangular wave from generator to sin wave

can you make video on how one port and two port network synthesis applied practically on designing ckts like filters and amplifiers, I'm confused because all the sources i found talk about the calculations, the parameters, sys function....my questions is how can I convert statement problem in to circuit design using/applying network synthesis. tnx!

A question follows after watching this video few times. Can we generate a sine wave directly at varying frequencies , instead of going through the triangular wave stage? I was not able to find your episode on it. Just about all the videos showing how to generate a sine wave is at a fix frequency. I want to mention that I watched your automatic gain control circuit construction, when you started inventing things like burying an LED and a light sensor into a homemade tube, I have to say that you know the physics of electronics to the point of almost being able to have verbal conversations with these components and ICs.

@AccidentalScience

Жыл бұрын

You can generate a sine wave directly. Try to give a look at the Wien bridge oscillator using an op amp and a small lamp to limit the gain. It is based upon the design invented by Hewlett and Packard. (We'll, they used a tube as an amplifier, not an op amp, but that is the concept.) That oscillator can vary the frequency by changing two resistors that form the bridge, you can conveniently use a double stacked potentiometer. Changing the frequency with a control voltage is less straightforward and a design like the one I used in this video is easier eventually.

@LL-ue3ek

Жыл бұрын

@@AccidentalScience thank you for the insight!

What a weird way to make sine wave. The common XR2206 chip uses this technique internally. Had no idea that HP did this as well.

Hi ! My application requires converting a very low frequency triangle wave (0,1 to 1,4 HZ) input to a sinusoidal one 180ºusing a single power supply so that the output will vary from 0V to maximum what i got by biasing the non inverting input using a potentiometer . I've followed your circuit using only one 5 mm LED and just one Ge diode .The output is useful but the botton of the sine is not "round" . What can I do to correct this? .Any way ,thanks a lot because your circuit was realy a great help for me !!!

@AccidentalScience

Жыл бұрын

If you use Ge diodes you need three of them in series in place of a single SI diode.

What are the resistor values that are used on the PCB?

@AccidentalScience

Жыл бұрын

Sorry at the moment I don't remember. At what point in the video do you see the resistors you're asking for?

@umit3212

Жыл бұрын

@@AccidentalScience at 5:10 when you start explaining the circuit that uses the LED's. I want to use this circuit for my VCO. So it would be nice to have an indication in what range the resistor values should be so I can tweak it easier. Thank you in advance!

@AccidentalScience

Жыл бұрын

Sorry for the late reply, I didn't notice your reply. Here are the resistors: 33K + 100K trimpot (series) for the OPAMP feedback (in parallel with LEDs), 1K + 2K2 (series) at the OPAMP input. At the node a 10K trimpot goes to the diodes. Hope this help. P.s. I am moving to the new website. It will take some time to complete but I will put there the schematic of this circuit as well.

Please make a part 2 follow up for this and improve your designs...because this is not a waveshaper. This is time-dependent network with diode filter and LED feedback significantly filtering the signal, making this unusable for I.e. phase distortion. You are making the assumption that the input signal has to be a triangle wave. That's why this isn't a waveshaper. You didn't see the importance that it maybe a modulated triangle wave with any arbitrary phase input at any moment in time. Try using this with arbitrary phase input and you will see it does not hold up.

@AccidentalScience

10 ай бұрын

So please explain where you see the time-dependent network, neglecting diodes' parasite capacitance of course.

first