Hey guys,
I'm designing a constant current source (on the order of 1A) that needs to be modulated in the hundreds of kilohertz to 1 MHz range. My plan was to have a constant current driver comprised of a mosfet, current sense resistor and an op-amp. There is an inductor in series after the mosfet to keep the current as constant as possible. In order to modulate it, I'm adding another mosfet constant current driver (this one a current sink), in series with the load, to divert current away from the load at high speeds.

The reason I don't want to directly modulate the initial current driver is that this needs to drive loads that are extremely sensitive to surges in current so I'm using the first current driver to set the absolute maximum current, then subtracting from that to modulate the load current.

I've included a diagram. This is my first time posting so I was just wondering if I'm doing anything stupid with this design since I'm relatively new at this (1.5 years out of school, BSEE) and haven't really worked with this kind of stuff.

Thanks a lot

 

The reason I don't want to directly modulate the initial current driver is that this needs to drive loads that are extremely sensitive to surges in current so I'm using the first current driver to set the absolute maximum current, then subtracting from that to modulate the load current.

Hey newothegreat!

I like that approach. You have limited the absolute maximum current your device should see, all the variable part can do is let less current flow. That seems like a good "fail safe" method to protect the device under test.

If the top current source is truly a constant current source then you should not need the inductor. An inductor in a current path like that may give you more trouble then it's worth, as in the large voltage it will generate when you attempt to turn this thing off or disconnect the test load.

And welcome to the boards!

 

I understand what you want to do but I'm curious about the purpose of the end design. Oh, I don't see the benefit of the inductor either

 

If the inductor is omitted, the modulation could only swing less than the current limit; the positive swing would be truncated.

I'm thinking that it may work better to use a broadband transformer for the modulation.

 

The inductor is to prevent a current surge. The load might be a laser diode, which is extremely sensitive to current spikes and is expensive and difficult to replace. I basically put the inductor in there as a fail safe in case there was some problem with the control circuit, like for instance a random voltage spike at the non inverting input causing the "constant current" output to jump suddenly. Even if the event were very short in time, it could still fry the diode.

The modulation isn't going to set at an frequency and will probably include a lot of DC. Would a transformer work for this kind of signal?

Thanks for all the feedback.

 

I guess also I should put a diode in parallel to the inductor instead of connecting it to ground

 

I guess also I should put a diode in parallel to the inductor instead of connecting it to ground

The 'Shunt' concept of this circuit eludes me. It seems counterproductive and energy wasteful. It would seem more logical to modulate the current source. Just don't let the max modulated current exceed 1A.