boylesg
Advanced Member level 4
- Joined
- Jul 15, 2012
- Messages
- 1,023
- Helped
- 5
- Reputation
- 10
- Reaction score
- 6
- Trophy points
- 1,318
- Location
- Epping, Victoria, Australia
- Activity points
- 11,697
I tried disconnecting all those diodes and it didn't seem to have much effect.The signal at the input to the opamp is limited to about +12.6V by D4.
Yes. See post 3 above.Could it be the OP amp itself?
Yes. See post 3 above.
Before I read your post 3, it occured to me to try increasing the 2 DC sources to 18V and then 24V and the clipping disappears.
So it is the opamp by virtue of the fact there is not enougn 'room' between 0V and +12V to accurately reproduce the input signal given that it is 0 - +20V with a +10V offset.
Ah well, it will be another interesting exercise to see if I can find or figure out a way around this.
I have read about amplifiers that convert amplitude differences to pulse width differences. Might be worth finding and re-examining some of those.
You are often linking this simulator in your posts, but clicking on the link does never show the simulation circuit for me you are apparently intending to show, only a standard RLC simulation. I guess there should be way to show the correct circuit for everyone, e.g. posting a file.Here you can see the various sources of clipping: https://tinyurl.com/92lhnlk
You are often linking this simulator in your posts, but clicking on the link does never show the simulation circuit for me you are apparently intending to show, only a standard RLC simulation. I guess there should be way to show the correct circuit for everyone, e.g. posting a file.
$ 1 5.0E-6 10.20027730826997 65 5.0 50
R 48 176 0 176 0 1 200.0 10.0 10.0 0.0 0.5
r 128 176 128 272 0 1000000.0
g 128 272 128 304 0
w 128 176 176 176 0
c 176 176 176 272 0 2.0E-11 1.6247195974630806
g 176 272 176 304 0
w 176 176 224 176 0
w 224 176 224 208 0
c 224 208 288 208 0 1.0E-10 0.3141657128141626
w 224 176 224 144 0
r 224 144 288 144 0 47000.0
w 288 144 288 176 0
w 288 176 288 208 0
w 288 176 336 176 0
d 448 192 448 96 1 0.805904783
a 544 208 624 208 1 12.0 -12.0 1000000.0
w 624 208 624 288 0
r 624 288 544 288 0 27000.0
w 544 288 544 224 0
w 544 192 448 192 2
d 448 288 448 192 1 0.805904783
g 448 288 448 320 0
174 624 288 656 384 0 100000.0 0.32180000000000003 Resistance
g 624 384 624 400 0
R 448 96 448 48 0 0 40.0 12.0 0.0 0.0 0.5
w 656 336 688 336 0
c 48 176 128 176 0 2.0000000000000002E-7 0.0
w 48 176 48 112 0
w 128 176 128 112 0
s 48 112 128 112 0 0 false
r 512 224 512 288 0 3000.0
s 448 288 512 288 0 1 false
w 624 384 688 384 0
s 688 384 752 384 0 1 false
d 752 320 752 384 1 0.805904783
d 752 256 752 320 1 0.805904783
d 752 192 752 256 1 0.805904783
w 752 384 800 384 0
d 800 384 800 320 1 0.805904783
d 800 320 800 256 1 0.805904783
d 800 256 800 192 1 0.805904783
w 800 192 752 192 0
w 752 192 688 192 0
w 688 192 688 336 0
O 800 192 880 192 0
w 512 224 544 224 0
r 336 176 336 272 0 100000.0
r 336 272 336 368 0 100000.0
g 336 368 336 384 0
w 336 272 384 272 0
S 448 192 384 192 0 1 false 0
w 384 272 384 208 0
w 336 176 384 176 0
o 0 16 0 290 37.41444191567111 1.8268770466636287E-4 0 -1
o 19 16 0 290 18.707220957835556 9.765625E-55 1 -1
o 44 16 0 34 9.353610478917778 9.765625E-55 2 -1
Also even without the diode, you are supplying the TL081 with +12 and -12.
The datasheet states the max input of the TL081 is +/- 15V or the supply range, whatever is less.
You have a 20V peak signal going into a TL081 with a +/- 12V supply. The diode is protecting the op amp from damage and clipping the input but if you intend to drive a 20V signal into a TL081 it will not work.
You will either want to find an op amp rated for your application and a larger power supply, or just use a voltage divider to scale your input down to an acceptable range and compensate with op amp gain (note the output of the TL081 is limited to +/- 13.5V, or less if the supply is less, which it is in your case).
Well for starters I did not design the whole circuit. I got it from here: http://xoscope.sourceforge.net/hardware/hardware.htmlIncidentally, what is your intention for R3 + D2?
- - - Updated - - -
switch to a higher rated op amp, such as the LM318.
Well that's the good thing about multisim huh. You can play around with *** abandon without any real consequences.You are exceeding the supply and input limits of the TL081. If you do that, for a more reliable simulation,
switch to a higher rated op amp, such as the LM318.
All I was doing was testing the DC couple switch that I added to the original circuit. I was merely inputing a signalBasically. Also exceeding the maximum input voltage runs the risk of blowing up your amp and permanently
clipping the signal to 0.
If I was to probe a real cicruit with a signal that might exceed the opamp limits or be severely clipped by the protection diodesInput voltage divider, as I suggested, will accomplish your task.
It can be expected to output undistorted 20 Vpp due to the AC coupling. By removing the AC coupling and feeding a 0-20 V input signal, you caused OP clipping (with or without diodes).Well for starters I did not design the whole circuit. I got it from here: http://xoscope.sourceforge.net/hardware/hardware.html
It can be expected to output undistorted 20 Vpp due to the AC coupling. By removing the AC coupling and feeding a 0-20 V input signal, you caused OP clipping (with or without diodes).
I think it's a pretty long discussion about a rather basic OP property: limited input and output voltage range.
WinXP Pro with most recent MS IE and Java versions. Somehow I get the right file by opening "Index of Circuit Examples" and pressing backspace.
Well that's the good thing about multisim huh. You can play around with *** abandon without any real consequences.
All I was doing was testing the DC couple switch that I added to the original circuit. I was merely inputing a signal
with a DC component and seeing what happened. I wasn't particularly concerned about exceeding the opamp limits at the
time and it is probably unlikely that I would probing that sort of signal with the real circuit anyway.
If I was to probe a real cicruit with a signal that might exceed the opamp limits or be severely clipped by the protection diodes
then I suspect I would be better off constructing an external voltage divider and probing said circuit through
that.
From one circuit to another how would you know how far to reduce the opamp input voltage?
Unless I made one of the resistors in your internal voltage divider a pot that I could adjust.
I think it's a pretty long discussion about a rather basic OP property: limited input and output voltage range.
By the way, I added those additional 6 diodes on the opamp output because I have read else where that most sound cards have a 2V limit on their 'line in' socket.
But the original circuit designer states that this circuit has an output of less than 12V. Will probably fry my sound card if I use this circuit without the extra clipping diodes.
What does the diode test on a DMM give you? I assume the voltage drop across the forward biased diode.
I can confirm this behavior in IE. I will report the issue to Falstad.
That is almost true except you are running a model outside the parameters it is designed for, and therefore it's behavior is likely unspecified by the designers of that model. It just so happens that your TL081 model behaves sanely in this situation, but that is not a general truth. I also presume this is for the oscilloscope sound card buffer that you are intending to actually build. Just as you use output clamping diodes in your simulation to protect a soundcard that does not exist in multisim, as a general rule of thumb you will want to use simulation models with behaviors well-defined over your entire operating range, otherwise you tread on uncertain territory.
You should be concerned about exceeding the opamp limits in the real circuit because clipping / distortion / damage are all real effects relevant to your real circuit.
A voltage divider for probing would be useful if you suspected the voltage range was outside the range of your probe. The voltages in this circuit however are well within the range of consumer DMMs.
You would have to decide what signal input range you wanted your circuit to work for, then reduce the voltage down to whatever op amp you choose to build the circuit with.
For example, let's say you are designing your circuit to handle a 0-20V input signal (this is a design decision that you make) and you are using a TL081 with a 12V power supply and 12V max input (another set of design decisions that you make).
Based on your decisions you would then determine that you want to reduce the voltage to 12/20 (60%) of its input value, such that 0 maps to 0 and 20 maps to 12, linearly. For this you can use a voltage divider to ground such as: signal -> 80kΩ -> op amp input -> 120kΩ -> ground.
After passing through the divider you would then clamp the voltage to safe limits for your circuitry using e.g. diodes. These diodes are a final line of defense against input voltages that exceed your design limits.
You "know" how far to reduce it by deciding the input range you want to operate on and looking at the input range of the op amp you decide to use.
That is a perfectly reasonable approach if you decide you want to let the user choose from a range of inputs or make adjustments based on the current situation. Another option is a e.g. a rotary switch with resistors calculated for various preset ranges (e.g. 2V, 5V, 20V, 200V).
I agree with this as well.
The majority of this discussion can be summarized as: Expect distortion and clipping when you exceed an op amps spec limits. All op amps have these types of limits. Refer to the datasheet for details. Design your circuit to operate over the desired input range without exceeding these limits.
- - - Updated - - -
You already posted this question, and it led to the exhaustive discussion here: https://www.edaboard.com/threads/259543/
See the posts at the end of that thread for answers and discussion on this question.
That is correct.
Yes I like this idea. Some preset ranges like a multimeter would add some certainty rather than blindly rotating a pot knob. I could do this as a front end retrofit.That is a perfectly reasonable approach if you decide you want to let the user choose from a range of inputs or make adjustments based on the current situation. Another option is a e.g. a rotary switch with resistors calculated for various preset ranges (e.g. 2V, 5V, 20V, 200V).
Just for interests sake, what opamps do have power inputs in excess of +20V and -20V? All the ones I have come across have power inputs between 12V and 20V.
Wouldn't be the first time that I have got a simulation error that cannot be fixed by the internal analyser. I would assume that is generally the result of exceeding limits of a simulator component.
Clearly the author of the original circuit has already provided protection for the opamp as he states that it is protected by inadvertanly probing 250V mains. Not that I would be remotely likely to have such an 'accident'.
If I was to use the internal voltage divider idea then I would estimate that being able to halve, third and quarter the opamp input voltage would be more than adequate. I can't see me having to probe much more than 24VDC at this point.
Is some one able and willing to explain to me how those diodes between the opamp and input stage do the over voltage protecting on the opamp?
$ 1 5.0E-6 10.20027730826997 50 5.0 50
R 592 240 496 240 0 0 40.0 5.0 0.0 0.0 0.5
R 592 400 496 400 0 0 40.0 -5.0 0.0 0.0 0.5
172 592 320 496 320 0 6 0.0 10.0 -10.0 0.0 0.5 Voltage
d 672 320 672 240 1 0.805904783
O 672 320 784 320 1
w 592 400 672 400 0
w 592 240 672 240 0
r 592 320 672 320 0 100.0
d 672 400 672 320 1 0.805904783
They lift D3's anode voltage one diode drop above the negative supply rail. D3 then prevents the opamp's non-inverting input going below the negative supply rail voltage.Incidentally, what is your intention for R3 + D2?
It may be simpler and better to just put a resistive divider at the opamp's output. e.g. 10K and 2K will reduce 12V to 2V.By the way, I added those additional 6 diodes on the opamp output because I have read else where that most sound cards have a 2V limit on their 'line in' socket.
- [
Given the apparent frequency at which you blow DMM fuses, you may want to reconsider that assessment.
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?