The design of current sensing circuits

[Faizan Jawaid]

I wanna ask how current sensing ckts are designed...Like i have a load of 1A driven by a mosfet and a load has a series .1ohm resistor...According to ohm...the voltage across the sense resistor would be 100mV...so i should use a comparator or something to convert it into TTL signal then make the switch off if it cross 100mV...??

 

[kender]

current sensing switch

so i should use a comparator or something to convert it into TTL signal then make the switch off if it cross 100mV...??

If you're describing a low side current-sense resistor connected to ground, then yes - a comparator with a 100mV threshold will trigger, when the current is above 1A.

 

[Faizan Jawaid]

Re: Current Sensing??

Are they any precision comparator to use or i can use any voltage comparator (LM311) to do a comparison of 100mV...And the reference should be set at 100mV?

 

[A.Anand Srinivasan]

Re: Current Sensing??

a low offset opamp would be very useful... rather than using an opamp u can go for comparators that r available.... they have better performance...

 

[AVIANO]

Re: Current Sensing??

I wanna ask how current sensing ckts are designed...Like i have a load of 1A driven by a mosfet and a load has a series .1ohm resistor...According to ohm...the voltage across the sense resistor would be 100mV...so i should use a comparator or something to convert it into TTL signal then make the switch off if it cross 100mV...??

If you particular about SMPS then most of the PWM controller IC has build in comparator/OpAmp for this purpose.

 

[Faizan Jawaid]

Re: Current Sensing??

Actually i dont intend to use a PWM ic...im generating the PWM signal from a controller...I was planing to add a AND gate at the Gate of the mosfet..And one input of AND will be from controller (whether to on the device or not) and other will be from current sensing cktry...

 

[A.Anand Srinivasan]

Current Sensing??

Dont use logic gates to drive the input of MOSfet.... it is not advisable.... it may lead to gate breakage....

 

[Faizan Jawaid]

Re: Current Sensing??

Dont use logic gates to drive the input of MOSfet.... it is not advisable.... it may lead to gate breakage....

No im not...Actually im using Gates only to use logic 1 or 0...I have a gate driver to drive the mosfet after the AND GATE...

 

[A.Anand Srinivasan]

Current Sensing??

k.... that is a cool way to do it... has ur question been answered...

if not please rebrief it a bit more so that we can answer clearer...

 

[Faizan Jawaid]

Re: Current Sensing??

My question was that do comparator are able to compare between low voltages with low reference...Like 100mV?

 

[Siniša]

Re: Current Sensing??

See TLV3011

 

[A.Anand Srinivasan]

Current Sensing??

ya.... there are comparators which can do so.... 100mV is not that low a reference.... if u r dealing with uV then maybe there is a reason to be worried....

 

[FvM]

Re: Current Sensing??

Dont use logic gates to drive the input of MOSfet.... it is not advisable.... it may lead to gate breakage....

Never heard about that phenomenon...

Actually, the drive current may be too low to achieve an intended switching speed. That's definitely the point, when a driver stage comes into play. But for slow switching applications, driving FET from logic is rather common. There is a whole family of logic-level FETs particularly supporting this option. A gate series resistor is generally recommended to achieve defined behaviour, but that's also the case when using a dedicated gate driver, at a higher current level.

Actually i dont intend to use a PWM ic...im generating the PWM signal from a controller...I was planing to add a AND gate at the Gate of the mosfet..And one input of AND will be from controller (whether to on the device or not) and other will be from current sensing cktry...

That sounds dubious. You're probably generating a high frequency self oscillation, that can quickly destroy the FET due to high switching losses. In a current mode controller, there's typically a flip-flop that allows only for one switch operation per clock cycle.

 

[Faizan Jawaid]

Re: Current Sensing??

Ok i got it...So u mean self oscillation does not work and damages gate?...and thats y PWM controllers have clocked latch in them...Ok thanks first i`ll try working on precision voltage comparators...

 

[A.Anand Srinivasan]

Re: Current Sensing??

Dont use logic gates to drive the input of MOSfet.... it is not advisable.... it may lead to gate breakage....

Never heard about that phenomenon...

Actually, the drive current may be too low to achieve an intended switching speed. That's definitely the point, when a driver stage comes into play. But for slow switching applications, driving FET from logic is rather common. There is a whole family of logic-level FETs particularly supporting this option. A gate series resistor is generally recommended to achieve defined behaviour, but that's also the case when using a dedicated gate driver, at a higher current level.

Actually i dont intend to use a PWM ic...im generating the PWM signal from a controller...I was planing to add a AND gate at the Gate of the mosfet..And one input of AND will be from controller (whether to on the device or not) and other will be from current sensing cktry...

That sounds dubious. You're probably generating a high frequency self oscillation, that can quickly destroy the FET due to high switching losses. In a current mode controller, there's typically a flip-flop that allows only for one switch operation per clock cycle.

i learnt that driving the gate of MOS using logic gates would destroy it from the "Art of electronics"
i'll try searching the book and posting the page number... it is owing to the capacitance of the gate junction which tries to cause a surge current....

 

[rvionics]

Re: Current Sensing??

I have attached an basic arrangment for constant current sink (The mosfet to be mounted on heat sink depending upon dissipiation)
the 0~100mv referance will decide the current through the FET