Fast and precision current comparator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
darkfenriz said:
seriously current conveyor is a similar concept to operational amplifier, but it has two low impedance inputs and a high impedance output, nomen omen.
It has a intrisic speed, because there is no voltage gain inside.

Your patterns may work, but every transimpedance stage made around a resistor in negative feedback loop may have problems with stability.
It's because from 'voltage' point of view, input impedance is infinite, and resistance in feedbackloop is given, so the 'voltage' gain is below 1, which is something most opamps hate.

That's why I am going to use video opamps.

Are current conveyors available as ICs?
 
This one looks better, for 20 mA input currents:

comparator-3.gif
 
Actually i am not concious of any current convoyour IC.
edit: maybe you may use a current feedback amplifier with non-inverting input grouded
As for opamp stability, no opamp likes voltage gain below one, especially fast video opamps. A capacitor in feedback loop is often mandatory. I suggest you studdied circuits designed for fast photodiode amplifiers, they serve the same purpose which is converting current to voltage and there is much research there, for example on how to tailor feedback capacitor for macimum gain*bandwidth.
 
Lumba Ogir said:
Wavebourn,
just like in audio, overall negative feedback is really a nuisance here.

Which overall negative feedback?

:xeye:


The first one is a current to voltage converter, with close to zero input resistance to prevent cross-talk between signals I measure.

The second one, is a maximum-possible-gain amp with logarythmic overdrive function; the 3'd one has a positive feedback, a flip-flop with well defined hysteresis to shape output and prevent oscillations.

Do you know what are you talking about, Lumba?
 
AX tech editor
Joined 2002
Paid Member
darkfenriz said:
Actually i am not concious of any current convoyour IC.
edit: maybe you may use a current feedback amplifier with non-inverting input grouded
As for opamp stability, no opamp likes voltage gain below one, especially fast video opamps. A capacitor in feedback loop is often mandatory. I suggest you studdied circuits designed for fast photodiode amplifiers, they serve the same purpose which is converting current to voltage and there is much research there, for example on how to tailor feedback capacitor for macimum gain*bandwidth.


Current conveyor has no voltage gain, *may* have current gain. What you would need is something like a MAX435 or 436, which unfortunately is obsolete, but you can build one from 2*AD844 or 2*OPA866. Edit: should be OPA860

Obviously for comparing current you want current mode circuits for high speed, not influenced by parasitic capacitances on nodes that have voltage swing.

jd
 
Wavebourn said:
This one looks better, for 20 mA input currents:

comparator-3.gif

A minor point: I think there is a (trivial) miswiring". There should not any link between A1a and A3 inverting inputs.

A major point: I would have to experiment with *real* circuit - not spice - but my concern is about precision, noise and speed.

The HA5023 bandwidth is high and comparatively high is the noise. Which is the amplitude of the noise *seen* by A2 input when Current 1 = Current 2?
I think it would be necessary some DC hysteresis around A2 to prevent oscillation and some AC hysteresis as well to improve speed. See Pease's paper.
These improvement would be beneficial also with slow varying currents (Is your signal a 1MHz sine wave or ranges DC up to 1MHz?)

About precision you didn't state any figures: do you need 8 bit or 24 bit precision?

HA5023 is a current feedback amplifier and his inverting input has a bias current *not* comparable with a subpicoampere fet input amplifier. 20 mA is a relatively high current but be aware with bias currents and drifts
 
Wavebourn said:
I need to compare 2 currents when they flow in different directions, if to say strictly. Both may swing between -20 mA to +20 mA, up to 1 MHz frequency.



Ok, the summing amp is the right way to do it. The single amp is fine. But I would use a comparator for the comparator function as they are better suited for that role than op-amps.

Why are you putting a bridge rectifier on the output?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.