Current Driver

[Yossarian]

Hey Guys,

I have a question concerning a project i am doing for University: I am supposed to help build an active magnetic field compensator for an electron microscope.

 For that i need a current driver (I'm not even sure if that's the right word) that can convert a small voltage (tens of mV) from the sensor into a large current (1+ A). It should also be quite fast and accurate between DC and about 1kHz.

I have tried to build a circuit but it's not working so far (i'll try to upload a picture later) and i don't really find a lot of helpful stuff on google neither.

 I hope you guys can help me out or maybe tell me what to search for.

Many thanks in advance.


Cheers

[awallin]

If it doesn't need to be ultra-low-noise then look at OPA569
here's a board I did a few years back: http://www.anderswallin.net/2014/09/tec-drive-v2/

if noise is crucial, look at Libbrecht-Hall or Ericksson circuits for laser constant current drives.
https://arxiv.org/abs/0805.0015
http://www.submm.caltech.edu/kids_html/DesignLog/DesignLog179/MillerMUSICReadoutDocs/HEMT%20Power%20Supply/Libbrecht%20and%20Hall,%20A%20Low%20Noise%20High%20Speed%20Diode%20Laser%20Current%20Controller.pdf

[Cerebus]

For that i need a current driver (I'm not even sure if that's the right word) that can convert a small voltage (tens of mV) from the sensor into a large current (1+ A). It should also be quite fast and accurate between DC and about 1kHz.

Do you have a specific transconductance that you're aiming for, i.e. a specific ratio of ^current/_voltage?

A few minutes thinking about that question might jog your memory as to exactly how you might achieve this. (Yes, I am deliberately adopting a didactic style rather than offering up any neatly packaged answer(s) as you've admitted yourself that it's your homework.  )

[MasterT]

Below 20 kHz any audio power amp, except D-class.

[David Hess]

So essentially you need a Howland current pump with a power output stage although there are other topologies which may be better.

Another way to do it is to use a fixed precision constant current source and then *subtract* another precision current from it with a parallel current source of the opposite polarity.  This has the advantage of only requiring one quadrant operation.  Usually it would be power intensive but in your case with an inductive load, the required voltage compliance is low so the power is low.

Check out figure 142 on page 65 of Linear Technology application note 47 for one implantation of a fast high current voltage controlled current source.