OpAmp Triangular wave oscillator

Thread Starter

Thanos_husk

Joined Nov 6, 2018
12
Hi!
I need to create a triangular wave oscillator for a project. I was wandering if i can get such a waveform using only one OpAmp. I found this schematic witch takes the waveform from the red dot , but i can't figure out how should i choose the resistor and capacitor values. I need it to have a frequency of 38kHz and input voltage of 20V dc (as shown in the schematic).
Any help would be awesome.
Thanks a lot.
 

Attachments

Alec_t

Joined Sep 17, 2013
14,280
That simple circuit would need to have an opamp supply voltage greater than 20V to get a 20V 'triangle' wave. The wave is actually an exponentially rising and falling wave rather than a true triangle, and would become distorted as soon as any significant load is applied to the red dot point.
What will you drive with the 38kHz?
 

Thread Starter

Thanos_husk

Joined Nov 6, 2018
12
That simple circuit would need to have an opamp supply voltage greater than 20V to get a 20V 'triangle' wave. The wave is actually an exponentially rising and falling wave rather than a true triangle, and would become distorted as soon as any significant load is applied to the red dot point.
What will you drive with the 38kHz?
The wave's Vp-p should be around 0.5 to 1V. Not even close to 20.
 

Alec_t

Joined Sep 17, 2013
14,280
The wave's Vp-p should be around 0.5 to 1V.
So why the 20V rail? It would certainly help for getting a nearly linear chunk of the exponential waveform, but otherwise seems excessive.
Methinks you are going to need another opamp to buffer the triangle wave in order to drive anything with it.
 

Thread Starter

Thanos_husk

Joined Nov 6, 2018
12
So why the 20V rail? It would certainly help for getting a nearly linear chunk of the exponential waveform, but otherwise seems excessive.
Methinks you are going to need another opamp to buffer the triangle wave in order to drive anything with it.
I just need to take a triangular wave form , filter it with a first order low pass filter and then turn it into a sin wave with a second order low pass filter.
 

danadak

Joined Mar 10, 2018
4,057
The attachment is the development tool for PSOC UP. Ignore it if
you have no experience with processors. Note it can generate
two waveforms at same time, tri and sine if you want.

If the goal is to make sine then many oscillators can be done with one
OpAmp all generating a sine, phase shift osc, colpitts, wein bridge....

Regards, Dana.
 

grahamed

Joined Jul 23, 2012
100
As has been said what you ask for is not possible.

For a small voltage range, and given that the output voltage swing of the OPA is much higher (and that is why you do need 20V supply despite what may have been said) the unloaded waveform will be tolerably triangular.

However any loading of the point, especially by something like a low-pass filer, will destroy the linearity. You must buffer the signal (generally using another OPA). If this is some kind of homework or test exercise then I would complain about its validity, otherwise I'd use an additional OPA.
 

Veracohr

Joined Jan 3, 2011
772
I just need to take a triangular wave form , filter it with a first order low pass filter and then turn it into a sin wave with a second order low pass filter.
If the ultimate goal is a sine wave, why not make a sine wave oscillator? That's perfectly feasible with one opamp.
 

AnalogKid

Joined Aug 1, 2013
10,986
As has been said what you ask for is not possible.
Sure it is.

As pointed out in posts #9 and #11, the triangle waveform does not have to be linear. In fact, the more of an exponential curve there is to it, the less filtering it will take to turn it into a sine wave.

ak
 
Top