Sawtooth Voltage Generator

[AMSA84]

Hi guys, I am trying to think about a sawtooth votlage generator and I concluded that I will need a comparator, current source (through current mirror) and a capacitor). However, I can't find any circuit which can do this.

Does anyone knows any good reference for this?

Kind regards.

 

[andre_luis]

Have a look here at this valuable suggestion:

• https://www.edaboard.com/threads/324112/#post1385132

 

[D.A.(Tony)Stewart]

Many different solutions exist depending your requirements.

Pls state them first, always.

frequency range KHz, MHz mHz?
Max slew rate of step to origin V/us
Amount of overshoot tolerated%
Output voltage range?
source Impedance ?
Expected load impedance R,C,L?
Linearity error on ramp? %

Hi guys, I am trying to think about a sawtooth votlage generator and I concluded that I will need a comparator, current source (through current mirror) and a capacitor). However, I can't find any circuit which can do this.


Does anyone knows any good reference for this?


Kind regards.

 

[AMSA84]

Hi and thanks for the reply.

The sawtooth generator is to be used inside an IC as part of a system.

It's to be operated >300MHz, output voltage range from 0V up to 3.3V and it's output is to be connected into a comparator (this sawtooth waveform is to be compared to a reference). Hope this info is enough because the other questions I don't have a clue of the requirements. I wonder if for the application is important.

Regards.

 

[D.A.(Tony)Stewart]

Hi and thanks for the reply.

The sawtooth generator is to be used inside an IC as part of a system.

It's to be operated >300MHz, output voltage range from 0V up to 3.3V and it's output is to be connected into a comparator (this sawtooth waveform is to be compared to a reference). Hope this info is enough because the other questions I don't have a clue of the requirements. I wonder if for the application is important.

Regards.

They are very important. A Sawtooth has two ramps. 3.3V at 300MHz or 3.3ns is 1000V/us. Thus the end of the sawtooth ought to be 100x faster ramp,, depending on application. 100,000V/us is not practical.

100kV/us with 1 pF load.
Let's see
Ic=1pF 0.1V/ps= 100mA with a switching time of 3.3ns/100= 33ps




Pls try again.. Why sawtooth and what are the slew rate requirements?

 

[AMSA84]

TO be honest, everything is open and nothing is defined. Yet, I don't know what kind of requirements a sawtooth needs.

It is a sawtooth because it is for a buck/boost converter.

Regarding the slew rate requiremente, we could go other way around, the current consumption on this block should not be greater than 1mA.

 

[D.A.(Tony)Stewart]

Then 300MHz is too high.

 

[AMSA84]

I have seen one or two works where this was done (high frequency PWM generators) where the sawtooth was implemented. Who says sawtooth, says a triangular waveform.

 

[D.A.(Tony)Stewart]

I have seen one or two works where this was done (high frequency PWM generators) where the sawtooth was implemented. Who says sawtooth, says a triangular waveform.

Triangle is easy, quote ref.

 

[AMSA84]

Maybe the authors that I am talking about made a triangular instead of ramp.

Another thing that came up to my mind is that maybe the sawthooth doesn't need to come from 0 up to 3.3V but from a certain voltage level, accounting with the transients, up to some other voltage level. For example, from 1V up to 2.5V.

https://ieeexplore.ieee.org/xpl/art...hip+components+for+integrated+dc-dc+converter

https://ieeexplore.ieee.org/xpl/art...icient+dc-dc+down+converter+based+on+dual+die

https://ieeexplore.ieee.org/xpl/art...mber=4672009&queryText=fully-integrated+dc-dc

https://ieeexplore.ieee.org/xpl/art...mber=6264597&queryText=fully-integrated+dc-dc

https://ieeexplore.ieee.org/xpl/art...mber=5316434&queryText=fully-integrated+dc-dc

And thinking now on the triangular, how can that be done?