Operational Amplifier used as Comparator
- Thread starter LMAO
- Start date
- Status
Phil Corso
Senior Member
- Location
- Boca Raton, Fl, USA
LAMO...
The major disacvantage is that its functioning properly is entirely dependent on the Op-Amp chosen!
Regards, Phil Corso
The major disacvantage is that its functioning properly is entirely dependent on the Op-Amp chosen!
Regards, Phil Corso
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120626-2126 EDT
LMAO:
It is time that you get some books, datasheets, and application notes and study these. You keep wandering around your problem without adequate background to make your own analysis and judgements.
First step is to adequately define your problem and create some specifications.
.
LMAO:
It is time that you get some books, datasheets, and application notes and study these. You keep wandering around your problem without adequate background to make your own analysis and judgements.
First step is to adequately define your problem and create some specifications.
.
gar,
just because you don't know my "specifications" and "defined problem" it does not mean that they do not exist. There are several reasons I don't post my entire project specifications, one of them trying to keep the problem simple and inside the scope of this discussion board. Also, I believe you are reading my threads and start extrapolating your perception and making assumptions by connecting the dots and judging my work.
Some answers are simply not found in datasheets and you don't really have to bother responding to my threads if you you feel your time is being wasted by my inadequate info.
A comparator is usually just an op amp with a high frequency cut off circuit. This circuit helps prevent oscillation that may occur near the switching point if a simple op amp is used open loop.
But will it damage the op amp if the output happend to keep oscillating? I don't think so. But I guess anything is possible.
And you can always add a capacitor to the op amp to make your own high freq. filter.
But will it damage the op amp if the output happend to keep oscillating? I don't think so. But I guess anything is possible.
And you can always add a capacitor to the op amp to make your own high freq. filter.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120627-1230 EDT
LMAO:
It is hard to give you help when you do not define the parameters of your application.
What information is lacking in datasheets that is presenting a problem for you? The datasheets I have used over the years have been fairly complete.
Power dissipation and/or excessive voltage are the likely causes of failure.
steve066:
Comparators are usually designed to be very fast response devices. For example 45 ns propagation delay. These tend have more of an oscillation problem, than do modern compensated op-amps. And op-amps tend to be slower than comparators, but typically have higher gain.
There is no indication that LMAO's application has need for very much bandwidth. 1 Hz or less might be adequate. A 741 probably can swing near rail-to-rail in less than 1 microsecond. I believe the typical is 0.5 microsecond. So the typical comparator is about 10 times faster.
.
LMAO:
It is hard to give you help when you do not define the parameters of your application.
What information is lacking in datasheets that is presenting a problem for you? The datasheets I have used over the years have been fairly complete.
Power dissipation and/or excessive voltage are the likely causes of failure.
steve066:
Comparators are usually designed to be very fast response devices. For example 45 ns propagation delay. These tend have more of an oscillation problem, than do modern compensated op-amps. And op-amps tend to be slower than comparators, but typically have higher gain.
There is no indication that LMAO's application has need for very much bandwidth. 1 Hz or less might be adequate. A 741 probably can swing near rail-to-rail in less than 1 microsecond. I believe the typical is 0.5 microsecond. So the typical comparator is about 10 times faster.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120627-1709 EDT
LMAO:
I will respond more later.
But quickly op-amps generally have no problem being driven into + or - saturation.
I would put a pair of back to back 1N4148 diodes directly across the + and - inputs. Then maybe a 10,000 ohm 1/4 W resistor in series with each input. Those resistors go to your 0.01 ohm shunt.
You have to worry about output loading, not too much current, and common mode voltage which includes how the power supplies are referenced.
.
LMAO:
I will respond more later.
But quickly op-amps generally have no problem being driven into + or - saturation.
I would put a pair of back to back 1N4148 diodes directly across the + and - inputs. Then maybe a 10,000 ohm 1/4 W resistor in series with each input. Those resistors go to your 0.01 ohm shunt.
You have to worry about output loading, not too much current, and common mode voltage which includes how the power supplies are referenced.
.
ggunn
PE (Electrical), NABCEP certified
- Location
- Austin, TX, USA
- Occupation
- Consulting Electrical Engineer - Photovoltaic Systems
We have a few circuits that use them in open loop mode and they work just fine. These are the 741 series.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120627-2139 EDT
LMAO:
See http://en.wikipedia.org/wiki/Operational_amplifier
and http://www.ti.com/product/lm741 then pick datasheet.
From the spec sheet the maximum input voltage can not exceed the power supply rails at +/-15 V.
Personally I do not operate with supply voltages greater than +/-12.
I first put the 741 into a product in 1971. The first IC op amp one I played with was a Fairchild uC702 about 1965. This cost in the range 0f $35 to $45 in 1965 dollars. I have an unused one in the drawer. This obvious is an antique now.
Back to your problem. How did you supply power to the op amp. For your voltage level of the shunt (I believe +24 V relative to ground) you need to reference the op amp to one or the other ends of the shunt. Thus you could generate -12 V relative to the shunt from the 24 V with a 7912. Then, I believe, with a 741 you are going to need some + voltage relative to the shunt, maybe +5 V. You may search for some other op amp that could work with the differential input at the positive rail. It may be that present 741s will.
The present 741 datasheet is lacking on performance of the amplifier when the inputs are near the rail voltages. A possible way to move the inputs somewhat from the positive rail and use just the 24 V supply to power the 741 is to provide two voltage dividers from the shunt to the minus of your 24 V supply. Maybe 1 K from the shunt to tap point and then 10 K to the - rail via the trimpot. Use low temperature coefficient metalfilm resistors with a 1% rating. At the bottom of the 10 K resistors put a 500 ohm trimpot with its center slider connected to the - rail and each end of the trimpot goes to one 10 K resistor. This should be sufficient to balance for differences in the resistors and to possibly set your threshold point. Eliminate the 10 K resistors I referenced earlier that connected to the back to back diodes across the amplifier input. Now the 1 K resistors serve the same purpose.
On the supply voltage to the op amp you may want to place a transient limiter. Transorb is one trade name for a Zener type limiter. This will require a series resistor for current limiting under transient conditions.
A much simpler approach may be the isolated current sensor I referenced in a previous post. I have not worked with it, but it is designed to measure currents offset from common (ground).
.
LMAO:
See http://en.wikipedia.org/wiki/Operational_amplifier
and http://www.ti.com/product/lm741 then pick datasheet.
From the spec sheet the maximum input voltage can not exceed the power supply rails at +/-15 V.
Personally I do not operate with supply voltages greater than +/-12.
I first put the 741 into a product in 1971. The first IC op amp one I played with was a Fairchild uC702 about 1965. This cost in the range 0f $35 to $45 in 1965 dollars. I have an unused one in the drawer. This obvious is an antique now.
Back to your problem. How did you supply power to the op amp. For your voltage level of the shunt (I believe +24 V relative to ground) you need to reference the op amp to one or the other ends of the shunt. Thus you could generate -12 V relative to the shunt from the 24 V with a 7912. Then, I believe, with a 741 you are going to need some + voltage relative to the shunt, maybe +5 V. You may search for some other op amp that could work with the differential input at the positive rail. It may be that present 741s will.
The present 741 datasheet is lacking on performance of the amplifier when the inputs are near the rail voltages. A possible way to move the inputs somewhat from the positive rail and use just the 24 V supply to power the 741 is to provide two voltage dividers from the shunt to the minus of your 24 V supply. Maybe 1 K from the shunt to tap point and then 10 K to the - rail via the trimpot. Use low temperature coefficient metalfilm resistors with a 1% rating. At the bottom of the 10 K resistors put a 500 ohm trimpot with its center slider connected to the - rail and each end of the trimpot goes to one 10 K resistor. This should be sufficient to balance for differences in the resistors and to possibly set your threshold point. Eliminate the 10 K resistors I referenced earlier that connected to the back to back diodes across the amplifier input. Now the 1 K resistors serve the same purpose.
On the supply voltage to the op amp you may want to place a transient limiter. Transorb is one trade name for a Zener type limiter. This will require a series resistor for current limiting under transient conditions.
A much simpler approach may be the isolated current sensor I referenced in a previous post. I have not worked with it, but it is designed to measure currents offset from common (ground).
.
Is the input voltage derived from the op amp's source voltage?? If so, that keeps the input voltage within the right range. If not, there is a chance the input voltage got too high or low.
And what is the output driving? Op amps (and most IC's) have very little output power capability. Since the output will swing to almost the supply voltage, you have to keep the output current fairly low.
And what is the output driving? Op amps (and most IC's) have very little output power capability. Since the output will swing to almost the supply voltage, you have to keep the output current fairly low.
- Status