Electronics > Projects, Designs, and Technical Stuff
How to add a current limit circuit to an opamp's output
RawCode:
Hello guys :)
I should limit the output current of an opamp, which is buffered by an LT1010.
There are a lot of current limiting circuits out there, but i don't know how to implement them in my design, so i tried to develop my own current limiting circuit.
This circuit should limit the current between .1uA to several mA. This large current range is given by the current sense method, changing some resistors when needed.
But how can i sense the load's current? Considering that the opamp can drive very resistive loads, even GOhm, i think that i can't use a high-side current measuring technique since i think that the opamp's input can be unbalanced considering that one input sees a low impedance and the other can see a variable impedance. Also a low-side current measuring technique is not feasible because i'm already using a transimpedance amplifier to precisely measure the load's current. I could use this current(voltage) information, but this circuit is very slow (switching relays, ADCs and so on) and it can be blind to some "fast" current variations.
I'm a student, so expect some stupid and hilarious stuff :palm:.
My idea is to use a Wilson current mirror in the opamp's feedback loop. I could use then the mirrored current in a faster and less precise transimpedance amplifier. Its output is then compared to a reference voltage.
To limit the current i could use a Depletion mode NMOS when the source is positive and an Enanched mode PMOS when it's negative, which both are in series with the load.
What do you think about this madness? Can i improve it in some way? There is a better method to limit the circuit in this circuit?
I tried this circuit in a breadboard and it works is the same way it does in the simulations. There is though a small noise in X4's output (in the breadboard there isn't that buffer). I managed to suppress it by a bit with a .1uF capacitor between the opamp's output and its inverting input, but it's still present.
Here some pictures (sorry for the bad quality). I'm using a 10x probe, so the actual values in the screen must be multiplied by 10.
Do you have any suggestions?
David Hess:
Current can be limited *without* any current sense resistor using a diode bridge driven by a pair of current sources. Mismatch in the diodes creases an output offset error but this is removed by the operational amplifier's negative feedback. This idea is shown below being used to protect 50 ohm inputs.
Somewhere I have also seen this used to make a pulse generator with variable slew rate by driving a capacitor with the current limiter and then buffering the output.
Zero999:
Are you intersted in just limiting the output current or making a constant currnet source/sink?
If it's the latter, then how about the Howland current pump?
https://www.ti.com/lit/an/snoa474a/snoa474a.pdf?ts=1596614856141&ref_url=https%253A%252F%252Fwww.google.com%252F
https://circuitdigest.com/electronic-circuits/howland-current-pump-circuit
RawCode:
Thank you both for the answers :)
--- Quote from: David Hess on August 04, 2020, 11:57:04 pm ---Current can be limited *without* any current sense resistor using a diode bridge driven by a pair of current sources.
--- End quote ---
This option seems quite interesting, but I don't understand it. Referring to "Tektronix 7A29 Input Protection", if pin 8 is the input of the circuit, and 18 is its output, how can a DC signal pass through U100? Is this circuit aimed for a small signal application?
--- Quote from: Zero999 on August 05, 2020, 08:08:14 am ---Are you intersted in just limiting the output current or making a constant currnet source/sink?
--- End quote ---
I think that I'm interested only in just limiting the current. At given voltage, if the load requires more current from the opamp, there should be a current limit that I can set before the measurement.
The uA723 has a current limiting feature like i want, but it's based on a transistor high-side sensing.
Also looking at uA723's schematic, I'm not quite sure how i can implement a similar design in my circuit, considering also that the opamp have bidirectional current flow unlike the uA723
Kleinstein:
The diode bridge idea works in a way similar to the depletion mode FETs. It can pass current up to a certain limit and if this limit is exceeded will turn high impedance, limiting the current. The diodes only let as much current pass as the 2 external sources provide. It is a nice protection as it also work for higher frequencies, but the 2 source give quite some waste heat.
Navigation
[0] Message Index
[#] Next page
Go to full version