A very useful explanation once again. We encountered a situation in motor control where we are faced with exactly this dilemma of using an opamp instead of a comparator. Basically we want to switch the commutation when we detect zero crossings of the backemf. However the mcu has only 1 built in comparator and we need 2 more. So we've been looking at the possibility of configuring the internal (to the mcu) opamps as Schmitt triggers. Probably the lower slew rate won't be an issue but the offset from rail voltage could be.
Remarkable! Thank you professor. This reminds me of the physics wars back in Einstein days, when physicists would call eachother's physics as rubbish. Except your engineering is superb. If I'm permitted to use the colloquial term, you the man.
Aside from the very helpful guidance that Sam always provides in his videos, a lesson to be learned here is not to trust some random tutorial on the internet. One shouldn't assume that some sloppily put together guide that was never peer reviewed or even edited by someone other than the original author, is a good cohesive and correct representation of reality, let alone a good unambiguous and comprehensive explanation of a particular subject matter. The virtual students of such tutorials often blame themselves for not understanding the subject being presented, in whatever manner it is being presented. More often than not, unfortunately, the lack of understanding among said students stems from a deficient or downright awful presentation. The presenter is the one at fault, often lacking the qualifications that should be a pre-requisite for teaching the subject matter, and once again, more often than not, those qualities that make one a good instructor or teacher. Members of the audience being presented to, assuming of course that they meet the qualifications of the target audience for whom the presentation was originally created, should feel the urge to just give up and assume the subject being presented is too complicated for them and it is their fault for not understanding it. Fortunately, there are brave caring souls, like Sam in this video, that when given the opportunity or while feeling a compulsion to act, voice their disdain for these sorts of presentations/tutorials, and make explicit the various inaccuracies, ambiguities, and shortcomings found therein.
Actually, NOT every operational amplifier (OA) can be a comparator, and the reason is NOT the phase-compensation capacitor. The reason is different: to be a comparator an OA should have a significant admissible input differential voltage that is comparable to the power supply voltage. It is necessary to make a comparator usable within an undetermined input differential voltage range and assert the Schmitt trigger state by setting the hysteresis to +/- 1 or more volts. Modern OAs have anti-parallel protection diodes between the inputs, which makes the admissible input differential voltage just +/- 0.5V. This is why many manufacturers do not even publish the admissible or maximum differential input voltage in the spec since these OAs are NOT INTENDED for work as comparators. Nowadays, you can find just a few models of OAs that have no anti-parallel input protection diodes and can be used as comparators. OA manufacturers do not even show protection diodes in the datasheet, although they are used everywhere, even in comparators, not between the inputs like in OAs, but between each input and power supply rails.
Hi Gregory, thanks for contribution. The compensating capacitor slows down the OA. Other restriction when OA is used as a Schmitt trigger apply, like common mode range at the input, output, and maximum differential input voltage. This presentation was referring to the LM741 Schmitt trigger design demo and it does not cover all the issues. I would disagree though with two of your assertions. The facts are: 1. Most, if not all, modern OA do not have diode protection between the inputs. 2. All datasheets specify the maximum permissible voltage between the inputs which is normally the full swing of the power supply. And one more point: if diodes are present between the inputs, the OA can still be used even with large hysteresis band. This is because it is OK to exceed the diode voltage if the current of the diodes is limited to a safe value (which is normally specified in datasheet). So by, for example, adding two resistor in series with the inputs, the diode restriction can be circumvented.
@@gregorymirsky8707 So? This article also refutes your assertions: "the outputs of most modern op amps can swing close to their supply rails". So modern Op Amps do not have diodes between the input terminals. And :"It is necessary to read the data sheet very carefully to determine if the worst-case voltages expected in the comparator application sufficiently affects the op amp performance to prevent proper operation" so datasheets do include the information. All the best. Sam
@@sambenyaakov Sam, I have never written what you are referring to above: I was not referring to the outputs. That was probably a different person. All our correspondence is above. Also, the article I provided the link to, on page 3 (Fig. 9), shows the said protection diodes between the inputs of the OA, therefore not refuting my assertions. The person above interfered with my account.
I think it is incredible that their definition of "high speed" is ~10uS slew rate, for a digital circuit nonetheless! I guess it is high speed in the context of morse code but that's about it. The voltage level issues could be somewhat alleviated by using some diodes to introduce voltage offsets. I've commonly seen a couple diodes stacked in series with the output to an indicated LED or input to an arduino. The uA741 can't swing the output to the rails obviously, but do the inputs even work down near the rails?
Thanks for comment. Yes, the input common mode voltage is even worse but in the "tutorial" they have set the reference around 6V, mid point of the supply.
Hello Professor, I am a big fan of your videos and how simple you make the concepts sound! I was wondering if you could add to your igbt/gate driver series? In particular, how to design working circuits with them and how how ensure all the chosen circuit elements will not blow up or cause the circuit to malfunction. There's just not enough resources out there showing the full circuit with gate drivers and igbts together! Thank you!
Hi , in minute 4:04 vref go to the minus terminal of opamp and signal through resistor to plus, but later on like in minute 4:40 vref go to plus and signal to minus, this is correct? Or not
Not only is the LM741 a poor choice for the given problem, the tutorial also fails to mention that the output signal might exhibit a phase reversal if the input signal happens to get too close to the supply rails.
Hello, this is the author of the article mentioned in the video. I have made an apology/explanation video which I will link here: th-cam.com/video/CYa3C0TAuWo/w-d-xo.html
Thanks for all this... 1) The comparator vs. op-amp info ... 2) The vital service of revealing yet another source of BS on the web. You're like the street-sweeper guy in the Bullwinkle cartoons, identifying for removal just a bit of the tons of detritus that pervades the public streets. Why do these people seek to advertise their ignorance and incompetence? An interesting question. It occurs in all arenas of life.
A very useful explanation once again.
We encountered a situation in motor control where we are faced with exactly this dilemma of using an opamp instead of a comparator.
Basically we want to switch the commutation when we detect zero crossings of the backemf. However the mcu has only 1 built in comparator and we need 2 more. So we've been looking at the possibility of configuring the internal (to the mcu) opamps as Schmitt triggers.
Probably the lower slew rate won't be an issue but the offset from rail voltage could be.
Thanks for sharing your experience.
As always, clear and proffesional. I am always excited to see every video you upload. Thank you for this work. Please keep up the good work.
Thanks for king note.
Remarkable! Thank you professor.
This reminds me of the physics wars back in Einstein days, when physicists would call eachother's physics as rubbish. Except your engineering is superb. If I'm permitted to use the colloquial term, you the man.
Thanks for warm words. Comments like yours keep me going.
Aside from the very helpful guidance that Sam always provides in his videos, a lesson to be learned here is not to trust some random tutorial on the internet. One shouldn't assume that some sloppily put together guide that was never peer reviewed or even edited by someone other than the original author, is a good cohesive and correct representation of reality, let alone a good unambiguous and comprehensive explanation of a particular subject matter. The virtual students of such tutorials often blame themselves for not understanding the subject being presented, in whatever manner it is being presented. More often than not, unfortunately, the lack of understanding among said students stems from a deficient or downright awful presentation. The presenter is the one at fault, often lacking the qualifications that should be a pre-requisite for teaching the subject matter, and once again, more often than not, those qualities that make one a good instructor or teacher. Members of the audience being presented to, assuming of course that they meet the qualifications of the target audience for whom the presentation was originally created, should feel the urge to just give up and assume the subject being presented is too complicated for them and it is their fault for not understanding it. Fortunately, there are brave caring souls, like Sam in this video, that when given the opportunity or while feeling a compulsion to act, voice their disdain for these sorts of presentations/tutorials, and make explicit the various inaccuracies, ambiguities, and shortcomings found therein.
Thanks for note and discussion.
Actually, NOT every operational amplifier (OA) can be a comparator, and the reason is NOT the phase-compensation capacitor. The reason is different: to be a comparator an OA should have a significant admissible input differential voltage that is comparable to the power supply voltage. It is necessary to make a comparator usable within an undetermined input differential voltage range and assert the Schmitt trigger state by setting the hysteresis to +/- 1 or more volts. Modern OAs have anti-parallel protection diodes between the inputs, which makes the admissible input differential voltage just +/- 0.5V. This is why many manufacturers do not even publish the admissible or maximum differential input voltage in the spec since these OAs are NOT INTENDED for work as comparators. Nowadays, you can find just a few models of OAs that have no anti-parallel input protection diodes and can be used as comparators. OA manufacturers do not even show protection diodes in the datasheet, although they are used everywhere, even in comparators, not between the inputs like in OAs, but between each input and power supply rails.
Hi Gregory, thanks for contribution. The compensating capacitor slows down the OA. Other restriction when OA is used as a Schmitt trigger apply, like common mode range at the input, output, and maximum differential input voltage. This presentation was referring to the LM741 Schmitt trigger design demo and it does not cover all the issues.
I would disagree though with two of your assertions. The facts are: 1. Most, if not all, modern OA do not have diode protection between the inputs. 2. All datasheets specify the maximum permissible voltage between the inputs which is normally the full swing of the power supply.
And one more point: if diodes are present between the inputs, the OA can still be used even with large hysteresis band. This is because it is OK to exceed the diode voltage if the current of the diodes is limited to a safe value (which is normally specified in datasheet). So by, for example, adding two resistor in series with the inputs, the diode restriction can be circumvented.
@@sambenyaakov Hi Sam, please read the article at www.analog.com/media/en/technical-documentation/application-notes/AN-849.pdf
@@gregorymirsky8707 So? This article also refutes your assertions: "the outputs of most modern
op amps can swing close to their supply rails". So modern Op Amps do not have diodes between the input terminals. And :"It is necessary to read the data sheet very carefully to determine
if the worst-case voltages expected in the comparator application
sufficiently affects the op amp performance to prevent proper
operation" so datasheets do include the information. All the best. Sam
@@sambenyaakov Sam, I have never written what you are referring to above: I was not referring to the outputs. That was probably a different person. All our correspondence is above. Also, the article I provided the link to, on page 3 (Fig. 9), shows the said protection diodes between the inputs of the OA, therefore not refuting my assertions.
The person above interfered with my account.
First op amp on the market was uA709, circa 1969 or 1970ish, later superceded by the uA741
Hi Ian, thanks for clarification.
very educative and explanatory, thank you
Thanks
Great job!!!, Greetings from an EE from México
😊🙏
I think it is incredible that their definition of "high speed" is ~10uS slew rate, for a digital circuit nonetheless! I guess it is high speed in the context of morse code but that's about it.
The voltage level issues could be somewhat alleviated by using some diodes to introduce voltage offsets. I've commonly seen a couple diodes stacked in series with the output to an indicated LED or input to an arduino. The uA741 can't swing the output to the rails obviously, but do the inputs even work down near the rails?
Thanks for comment. Yes, the input common mode voltage is even worse but in the "tutorial" they have set the reference around 6V, mid point of the supply.
High speed for those ancient 74XX series TTL chips, maybe.
Good tutorial The comparator have an open collector output
Thanks. Not all😊 see the datasheet of the analog unit mentioned in video.
Hello Professor, I am a big fan of your videos and how simple you make the concepts sound!
I was wondering if you could add to your igbt/gate driver series? In particular, how to design working circuits with them and how how ensure all the chosen circuit elements will not blow up or cause the circuit to malfunction. There's just not enough resources out there showing the full circuit with gate drivers and igbts together!
Thank you!
Thanks for comments. Will try.
Hi , in minute 4:04 vref go to the minus terminal of opamp and signal through resistor to plus, but later on like in minute 4:40 vref go to plus and signal to minus, this is correct? Or not
You seem to have skipped or not reached 😊th-cam.com/video/yDickD2tJS0/w-d-xo.html
Great work sir.....
Thanks
Not only is the LM741 a poor choice for the given problem, the tutorial also fails to mention that the output signal might exhibit a phase reversal if the input signal happens to get too close to the supply rails.
Thanks Wei, excellent point! More to the down side of the 741 in this application.
Great !
thank you professor
Thanks
Hello, this is the author of the article mentioned in the video. I have made an apology/explanation video which I will link here:
th-cam.com/video/CYa3C0TAuWo/w-d-xo.html
👍👍👍🙏👍👍👍
😊👍
Thanks for all this... 1) The comparator vs. op-amp info ... 2) The vital service of revealing yet another source of BS on the web.
You're like the street-sweeper guy in the Bullwinkle cartoons, identifying for removal just a bit of the tons of detritus that pervades the public streets.
Why do these people seek to advertise their ignorance and incompetence? An interesting question. It occurs in all arenas of life.
Thanks for sharing you thoughts. Good points. In the making are educational websites that will post reviewed (by expert) videos.