Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Comparator characterization for variable input

Status
Not open for further replies.

iaf

Member level 3
Joined
Apr 11, 2019
Messages
57
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
433
Hi, I'm using a single stage amplifier IC design as a comparator for generating a reset signal when the input voltage goes below the refrence voltage (0.7V), the input voltage is produced by a voltage divider from the voltage supply (3.3V) so i'm getting an input range from 0-0.9V.
Now i want to characterize the comparator calculating the gain, delay time, slew rate etc. The problem that what matters for me is the voltage drop rate which varies with the fall rate of the supply. I tried simulating with small input voltage overdrive 0.65-0.75V but the comparator is too slow for reaching the rails(which are not really reached) depending on the load i'm using (which i don't really know) so for every change in the overdrive voltage or the voltage fall rate i get different values.
My question is how to calculate these values while the values of the overdrive voltage and voltage fall rate are variable. Also which load should i expect when this output is going to be connected to the reset pin in the Microcontroller? and which charectertistics would be important in this case?
Thanks
 

Microcontroller pin should be modeled as small C and may also have a weak-pull up/down. You should find that info in your datasheet.

I don't see any way to answer your question because its a hard question and application specific. But I can suggest you work backwards. The circuit isn't there for fun it's there to solve some problem. So maybe you know the parameters of the problem and can verify the circuit solves it:

How long can your circuit tolerate 0.69V -> Ensure your comparator responds in time
How long can your circuit tolerate 0.68V -> Ensure your comparator responds in time
etc.

And consider the the whole picture such as the microcontroller's reaction time. If the microcontroller response time is 10ms then a 1ms response from the comparator is probably fine in all cases.
 

Hi,

it seems you have the circuit before you have the requirements....

Forget the circuit for now, just decide and tell us the requirement for the function.
reset low active,ir high active
Where is the threshold supply voltage? What tolerance? What hysteresis?
What´s the timing requirement: from getting low voltage --> to asserting RESET?
What´s the slew rate requirement?
When the supply voltage is considered good: do you want the RESET to be deasserted immediately or do you want some delay time?
you say: "input voltage range 0-0.9V" ... "reference voltage 0.7V" ..."small input voltage overdrive 0.65-0.75V"...If the threshold is 0.7V and you add some "small" 0.7V overdrive, then you are beyond input voltage range....
(btw: 0.7V isn´t really small overdrive in any case. 0.7millivolt maybe is. )

***
After that...

so i'm getting an input range from 0-0.9V.
I don´t see wher the 0.9V is coming from.

Now i want to characterize the comparator calculating the gain, delay time, slew rate etc.
&
the comparator is too slow
first you need to know the requirements...

I tried simulating
then, why don´t you show us your simulation?

which i don't really know
If not you ... who else?

Also which load should i expect when this output is going to be connected to the reset pin in the Microcontroller?
and which charectertistics would be important in this case?
The microcontroller datasheet will tell you.

Klaus
 

Microcontroller pin should be modeled as small C and may also have a weak-pull up/down. You should find that info in your datasheet.

I don't see any way to answer your question because its a hard question and application specific. But I can suggest you work backwards. The circuit isn't there for fun it's there to solve some problem. So maybe you know the parameters of the problem and can verify the circuit solves it:

How long can your circuit tolerate 0.69V -> Ensure your comparator responds in time
How long can your circuit tolerate 0.68V -> Ensure your comparator responds in time
etc.

And consider the the whole picture such as the microcontroller's reaction time. If the microcontroller response time is 10ms then a 1ms response from the comparator is probably fine in all cases.

I don't have a specific MC i have looked some datasheets but couldn't find information about the capacitance i just used a 100fF C as a guess for the capcitance of the MOSFET, i also have to consider the parasitic capacitance.
I forgot to mention that the clock is 50MHz, so the fewer the clock cycles the better it is, other than that there are no strict requirments.

- - - Updated - - -

Hi,

reset low active,ir high active

it's low active

Where is the threshold supply voltage? What tolerance? What hysteresis?

i'm not sure what you mean by the threshold supply voltage

What´s the timing requirement: from getting low voltage --> to asserting RESET?
What´s the slew rate requirement?

There are no strict requirements for that

When the supply voltage is considered good: do you want the RESET to be deasserted immediately or do you want some delay time?

I don't need delay but it will be there due to the propagation delay of the comparator which is also not a problem

you say: "input voltage range 0-0.9V" ... "reference voltage 0.7V" ..."small input voltage overdrive 0.65-0.75V"...If the threshold is 0.7V and you add some "small" 0.7V overdrive, then you are beyond input voltage range....

i still don't get the point of the threshold voltage

(btw: 0.7V isn´t really small overdrive in any case. 0.7millivolt maybe is. )

The overdrive is 5mV

I don´t see wher the 0.9V is coming from.

The voltage divider of the supply

- - - Updated - - -

. I tried simulating with small input voltage overdrive 0.65-0.75V

I've mistaken in that the reference is 0.77V and the input overdrive is 0.765-0.775V
 
Last edited:

Hi,

Thank you for going through my many questions.

i have looked some datasheets but couldn't find information about the capacitance
I would be surprised...
please give a link to the datsheets you used.

as a guess for the capcitance of the MOSFET
Which MOSFET?

i'm not sure what you mean by the threshold supply voltage
If I understand correctly, then you want to monitor the supply voltage ... with the use of a resistive divider ... then to a comparator with 0.7V reference.
If so, then there is a certain level of the supply voltage where the comparator switches....this is called the threshod voltage.

I forgot to mention that the clock is 50MHz, so the fewer the clock cycles the better it is, other than that there are no strict requirments.
Why do you need a clock?

There are no strict requirements for that
If so, then "half an hour" should be OK?
If not --> then you have requirement.

In post#1 you wrote "too slow" ... then obviously you have some timing limit in mind.

****
Please tell me. How do I have to understand this: You try to build a "RESET generator IC" but you have no microcontroller in mind.
Nobody would buy shoes ... and afterwards looking for a grilfriend with the right shoe size.
Usually there is a demand for something .... and because there are hundreds of RESET generators ready to buy .. I assume you look for a solution that the others can´t fulfill.

Or is it some kind of school project?
If so, then I recommend to go some datasheets of existing reset generator ICs ... go through the specifications step by step and ask yoruself where your limits/requirements are.
Do a search for "reset and supervisor ICs"

Klaus
 

please give a link to the datsheets you used.
https://ww1.microchip.com/downloads/en/devicedoc/31003a.pdf

Which MOSFET?
The MOSFETs of the technology i'm using having a reset as a digital input means that the smallest W/L are used
If I understand correctly, then you want to monitor the supply voltage ... with the use of a resistive divider ... then to a comparator with 0.7V reference.
If so, then there is a certain level of the supply voltage where the comparator switches....this is called the threshod voltage.
The threshold voltage is 2V
Why do you need a clock?
That's the clock of the MC so it gives an idea about the timing of the comparator and the reset signal has to be synchronized with the clock
If so, then "half an hour" should be OK?
If not --> then you have requirement.
In post#1 you wrote "too slow" ... then obviously you have some timing limit in mind.
Well as mentioned i would try to reach a reset in fewer clock cycles
****
Please tell me. How do I have to understand this: You try to build a "RESET generator IC" but you have no microcontroller in mind.
Nobody would buy shoes ... and afterwards looking for a grilfriend with the right shoe size.
Usually there is a demand for something .... and because there are hundreds of RESET generators ready to buy .. I assume you look for a solution that the others can´t fulfill.

Or is it some kind of school project?
If so, then I recommend to go some datasheets of existing reset generator ICs ... go through the specifications step by step and ask yoruself where your limits/requirements are.
Do a search for "reset and supervisor ICs"

Klaus
Well yes it's a project and my requirments are not concerned with the timing as much as power consumption and temperature stability. I've looked at Datasheets and papers already and that was my approach. I know this kind of comparator will be slower than other ones but it's actually doing its job for the conditions i set for my circuit i just need to characterize it which what i was asking for. I made the a triangle wave of the supply voltage with fall rate of 8v/µs with an overdrive of around 50 mV (for the comparator) and i'm having a delay of around 40ns so that is acceptable for my circuit but for charectrizing the comparator alone i was confused which values i should use to do so
 

Hi,

Datasheet:
Pin input current and load capacitance....are electrical values, thus you need to read the datasheet section with the electrical specifications.

Klaus
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top