Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Can someone explain this circuit?

Status
Not open for further replies.

technogeek

New Member
The circuit that was posted by Russelk here:

https://www.electro-tech-online.com/threads/adjustable-current-output.27691/

Can someone explain how it works? Maybe I just don't understand how the voltage regulator fits in to the big picture?

I think the fet works like a voltage controlled variable resistor to vary the current through the load... But I don't know where to begin to select part numbers for a specific "variable" constant current 1-100ma source.
 
technogeek said:
The circuit that was posted by Russelk here:

https://www.electro-tech-online.com/threads/adjustable-current-output.27691/

Can someone explain how it works? Maybe I just don't understand how the voltage regulator fits in to the big picture?

I think the fet works like a voltage controlled variable resistor to vary the current through the load... But I don't know where to begin to select part numbers for a specific "variable" constant current 1-100ma source.
The voltage regulator has no role in the current source.
You want a 1-100ma corrent source? What is your control voltage range? Or would you rather use a pot for control?
 
Yes, I basically need a voltage controlled current source.

0-12v input controlling a 1-100ma output. Doesn't have to be linear. The input voltage will be locked at one voltage, and the output current must not change, even if the output voltage (1-6v) or Vcc changes (nominal 12v, can vary 9-18v).
 
technogeek said:
Yes, I basically need a voltage controlled current source.

0-12v input controlling a 1-100ma output. Doesn't have to be linear. The input voltage will be locked at one voltage, and the output current must not change, even if the output voltage (1-6v) or Vcc changes (nominal 12v, can vary 9-18v).
Do you need a source or a sink? It sounds like you want a source, as you said the load voltage is positive.
Before I jump into a design, what is this for? I have designed more than one circuit that the user discovered did what he asked for, but wasn't what he needed.
 
Last edited:
Yes, a source. Basically I need to turn an analog DC output from a PLC into a variable current output. There doesn't need to be any correlation (really) between input voltage and output current (nor does it need to be repeatable), but once the input voltage "locks in", the current must not change for any situation inside it's limits.

It's for a small heater element. Physics and stuff.
 
Can you use surface mount parts, or do you need to stick to through-hole? What amount of drift can you tolerate, over what period of time? No circuit is absolutely stable.
Suppose the current went from 0 (instead of 1ma) to 100ma over 0 to 12V? That's a little simpler, but is not really a problem if you need the 1ma floor.
 
Last edited:
absolutely, 0-100ma is fine. Surface mount is not a problem. SOICs are fine, TSSOPs and fine pitched things get a little iffy.

+/- <1ma drift would be ideal. +/- 2ma would be the outer limits. I would want that kind of regulation indefinitely. If current is set, and bounces back and forth between 35 and 37ma forever - that's perfectly fine. However if it's set, and starts climbing at 0.5ma/hr on it's own accord, that's not acceptable.

So what is wrong with using the circuit in the previous thread?
 
technogeek said:
absolutely, 0-100ma is fine. Surface mount is not a problem. SOICs are fine, TSSOPs and fine pitched things get a little iffy.

+/- <1ma drift would be ideal. +/- 2ma would be the outer limits. I would want that kind of regulation indefinitely. If current is set, and bounces back and forth between 35 and 37ma forever - that's perfectly fine. However if it's set, and starts climbing at 0.5ma/hr on it's own accord, that's not acceptable.

So what is wrong with using the circuit in the previous thread?
It's a current sink, not a source, and it would probably oscillate with that big honkin' MOSFET loading the op amp.
The circuit below should work well for you.
 

Attachments

  • voltage controlled current source 0-100mA.PNG
    voltage controlled current source 0-100mA.PNG
    15.1 KB · Views: 278
Last edited:
Thanks!

How does it work? I don't like building things I don't have a thorough understanding of...

It looks like the U1a/M1 combo set up a 0-0.33ma controlled current source.... Which would set the input of u1b at 0-0.99v below vcc. The 2nd half seems to be doing much the same thing, which would mean the voltage across R1 is 0-0.99v.. Therefore the current through it is 0-100ma. Q1 and Q2 seem to be some kind of current following deal I guess.

Am I close with that?

Now had I said I wanted a current sink, the 2nd half of the circuit would not be necessary, right? Just adjust the values to the proper 100ma vs 0.33ma output. Maybe change M1 (haven't looked at it's specs)

(see, I'm learning again!)

Now, what would happen if the output was short circuited? A full 17V would be across the output transistors, which would mean 1.7W coming out. No big deal.

Is there any common failure part that would cause loss of regulation? Are opamps sensitive to static? If I stick these parts in an oven/freezer will the regulated current change?

Thanks in advance for your help, this is fun!

Oh, one more question. Regarding the previous circuit example you said the circuit might oscillate because of the "big honkin mosfet loading the op amp".... what's the difference between that the the M1 of your circuit? Is there a larger gate capacitance or something?
 
Last edited:
technogeek said:
Thanks!

How does it work? I don't like building things I don't have a thorough understanding of...

It looks like the U1a/M1 combo set up a 0-0.33ma controlled current source.... Which would set the input of u1b at 0-0.99v below vcc. The 2nd half seems to be doing much the same thing, which would mean the voltage across R1 is 0-0.99v.. Therefore the current through it is 0-100ma. Q1 and Q2 seem to be some kind of current following deal I guess.

Am I close with that?
Yep, except the input of U1b is 0-1.00V below vcc. Details.:p
Q1 and Q2 are a Darlington pair. The base current of Q1 will be very low, meaning that nearly all the current you set up in R1 gets passed to the output. You could use a p-channel MOSFET instead, but I had trouble finding one I liked - most were either to small (low power-handling capability) or too large (high gate capacitance - see below).

Now had I said I wanted a current sink, the 2nd half of the circuit would not be necessary, right? Just adjust the values to the proper 100ma vs 0.33ma output. Maybe change M1 (haven't looked at it's specs)

(see, I'm learning again!)
You got it again!

Now, what would happen if the output was short circuited? A full 17V would be across the output transistors, which would mean 1.7W coming out. No big deal.
That's why I picked a medium-power transistor for Q2. If you expect a long-term output short, you should use a small heatsink on Q2.

Is there any common failure part that would cause loss of regulation? Are opamps sensitive to static? If I stick these parts in an oven/freezer will the regulated current change?
I would not shuffle my feet on the carpet and then touch any of the semiconductors, but this is the case with any circuit.
Temperature extremes will cause small current changes. The most sensitive parts, believe it or not, are the resistors. If you are worried about thermal stability, use low temperature coefficient resistors. R1 dissipates 100mw max, so you should use a rating of at least 1/4 watt to keep its from changing excessively. A half watt rating might be better.

Thanks in advance for your help, this is fun!

Oh, one more question. Regarding the previous circuit example you said the circuit might oscillate because of the "big honkin mosfet loading the op amp".... what's the difference between that the the M1 of your circuit? Is there a larger gate capacitance or something?
Gate capacitance is the potential problem. M1 is a small-geometry part, so it has low gate capacitance. Gate capacitance is not an impossible problem to overcome. It basically just requires an extra resistor and capacitor.
 
Thanks for the reply!

Ron H said:
Gate capacitance is the potential problem. M1 is a small-geometry part, so it has low gate capacitance. Gate capacitance is not an impossible problem to overcome. It basically just requires an extra resistor and capacitor.

So what causes the oscillation? Is it because the op amp is trying to regulate the current through the mosfet, but the mosfet is too slow to react?
 
technogeek said:
Thanks for the reply!



So what causes the oscillation? Is it because the op amp is trying to regulate the current through the mosfet, but the mosfet is too slow to react?
Many (most?) op amps don't work well into capacitive loads. The cap adds an unwanted pole to the feedback loop, causing excessive phase shift and making the loop unstable. The problem can sometimes be solved by adding a zero in the loop to (partially) cancel the pole. If you haven't studied control theory, this probably makes no sense.
 

Attachments

  • cap compensation.PNG
    cap compensation.PNG
    5.8 KB · Views: 168
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top