The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> comparator with tens of mV hysteresis
https://designers-guide.org/forum/YaBB.pl?num=1153098375

Message started by hunk on Jul 16^th, 2006, 6:06pm

Title: comparator with tens of mV hysteresis
Post by hunk on Jul 16^th, 2006, 6:06pm

now i design a comparator which should have tens of mV hysteresis, for example, 20mV, but the spread should be less than +- 10%~20% under all supply, temp, process. many schmitt circuits have larger hysteresis and larger spread. i designed some circuits. but the simulated results seems bad.

can anyone help me. i need any advice, circuits, IEEE papers.
tanks in advance!

Title: Re: comparator with tens of mV hysteresis
Post by sheldon on Jul 17^th, 2006, 12:06am

Hunk,

Try adding positive feedback to a comparator using resistive
feedback. Since the amount of hystersis is controlled by ratioed
resistors, it should be repeatable. There is a nice summary at

http://www.ecircuitcenter.com/Circuits/op_comp/op_comp.htm

Best Regards,

Sheldon

Title: Re: comparator with tens of mV hysteresis
Post by hunk on Jul 18^th, 2006, 12:12am

sheldon wrote on Jul 17^th, 2006, 12:06am:

sorry, i haven't describe my specs in detail. my comparator is used in a chip. and it used to compare with a reference voltage. i need this compatator should "one side" hysteresis. it means when input rises from low to hign the switching point should equals reference voltage just like a normal comparator, while from high to low the hysteresis should take effect.
for getting "one side hys", i use a MOS switch. but this swtich will intruduce Ron which will vary with supply,Temp,and processes. so if using the resistor positive feedback, comparator will suffer from those. and need more area in my chip.

Title: Re: comparator with tens of mV hysteresis
Post by jhuang on Jul 25^th, 2006, 7:33pm

Hunk,

I think you could get your 2nd threshold (Vref2 = V_ref - V_hysteresis) by using a resistor divider. Avoid using a switch in series with the resistor divider (as you mentioned R_on error) instead using 2 tranmission gates to connect either V_ref or Vref2 to the input of the comparator. This way, the Ron of transmission gate has negligible effect due to the high input impedance of the comparator.

The other concern is the offset and its variation over processs/temperature (from the comparartor itself) which could easily mess up your spec (+/-10%, which is +/- 2mV). You need do somthing special there to meet the spec.

Regards, jhuang