Dear fellow DIYers;
I am looking for a fast and precision current comparator, to compare currents +/-20mA, up to 1 MHz. What I am thinking about, to use a video opamp in inverting connection, then another one with open loop. However it will work, but may be you can suggest some better solutions?
I will share results of the project with the DIY community, when good high-end results will be obtained.
Anatoliy
I am looking for a fast and precision current comparator, to compare currents +/-20mA, up to 1 MHz. What I am thinking about, to use a video opamp in inverting connection, then another one with open loop. However it will work, but may be you can suggest some better solutions?
I will share results of the project with the DIY community, when good high-end results will be obtained.
Anatoliy
Wavebourn said:What are weak points of such a solution?
20 mA input currents that I compare is not necessary, I mean, enough of current to deal with parasitic capacitances.
If I correctly understand your schematic, the main (not the only) weak pont is that the A1a HA5023 under certain condition (ie Current 1=20mA, Current 2=0mA) saturates suffering for delay to recover from saturation.
There is a paper from Robert Pease about the use of an OpAMP as precison voltage comparator
http://europe.elecdesign.com/Articles/ArticleID/9973/9973.html
Two questions:
1) If I understand your goal your circuit doesn't compare if Current 1 > Current 2 but if Current 1 > - (Current 2).
Is it correct?
2) Which is the purpose of the bridge + diode connected transistor on the second OpAmp output?
1. Who is Blomley?
2. As I said, it is just an idea. However, no way I can allow the video opamp to saturate. Couple of fast diodes on input, for example, would prevent this, but I prefer linear input resistance close to zero, so feedback resistor would be smaller, let it be 240 Ohm (for 20 mA+20 mA = 40 MA, +/1 8.6V out).
3. The purpose of a bridge and a Zener is to stabilize threshold of hysteresis. However, it should be more sensitive, say R3 = 1 mOhm, need to be calculated against capacitances.
2. As I said, it is just an idea. However, no way I can allow the video opamp to saturate. Couple of fast diodes on input, for example, would prevent this, but I prefer linear input resistance close to zero, so feedback resistor would be smaller, let it be 240 Ohm (for 20 mA+20 mA = 40 MA, +/1 8.6V out).
3. The purpose of a bridge and a Zener is to stabilize threshold of hysteresis. However, it should be more sensitive, say R3 = 1 mOhm, need to be calculated against capacitances.
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.
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.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.