Variable current source with AD8610 opamp

[nickishere]

Hi Edaboard,

I was looking over an Analog Devices Application Note #968 where they describe different op-amp and transistor circuits to drive a variable current. The application note can be found here:
https://www.analog.com/static/imported-files/application_notes/AN-968.pdf

The circuit I'm interesting in building is mentioned in Figure 5 at this link:
https://i.imgur.com/TyaJkXP.png

I'm looking to generate a variable current from 0 to 1 Amp. I successfully prototyped the circuit on my breadboard and can sweep the current from 0 to 1A by changing the input voltage from 0 to 2 Volts. I used the opamp in a bipolar configuration with +- 10 Volts with a External supply of +5V.

I tried lowering the opamp supply to +-5.0V and just single supply (GND and +5.0V) but under both cases the circuit ceases to function. Is a bipolar opamp required for this circuit? The current sense amplifier across the 100 m-ohm load should have a positive output voltage. I'm a little confused as to what exactly the upper opamp with 10M and 22 pF exactly does.

I'd like to run this circuit with a single supply with a 5 Volt supply if possible.

Thanks,
Nick

 

[D.A.(Tony)Stewart]

The only problem is the Op Amp used requires a bipolar supply because the input does not operate near the supply rail. This device also has high slew rate dv/dt to switch the MOSFETs fast. The 10 MOhm feedback to (+) creates positive feedback or hysteresis for full speed turn on to drive the capacitive gate during transition. This puts the switch into a PWM relaxation oscillator mode so high bandwidth is needed as well. But this also creates output ripple if there is any inductive trace, wiring effects or load. the small cap to (-) adds some stability & noise reduction with negative feedback.

A better choice is a rail-to-rail op amp with high current drive capability such as this one. https://www.analog.com/en/AD8028 (Dual) which operates on a wide range of single supplies and has some other features for input crossover selection. This also has very high GBW and slew rate with 100mA short circuit drive. (dont try it very long, that will blow the output)

I have no idea how much noise out it will generate (yet) nor what frequency it will regulate at with the modest small ratio (+) feedback

 

[nickishere]

The only problem is the Op Amp used requires a bipolar supply because the input does not operate near the supply rail. This device also has high slew rate dv/dt to switch the MOSFETs fast. The 10 MOhm feedback to (+) creates positive feedback or hysteresis for full speed turn on to drive the capacitive gate during transition. This puts the switch into a PWM relaxation oscillator mode so high bandwidth is needed as well. But this also creates output ripple if there is any inductive trace, wiring effects or load. the small cap to (-) adds some stability & noise reduction with negative feedback.

A better choice is a rail-to-rail op amp with high current drive capability such as this one. https://www.analog.com/en/AD8028 (Dual) which operates on a wide range of single supplies and has some other features for input crossover selection. This also has very high GBW and slew rate with 100mA short circuit drive. (dont try it very long, that will blow the output)

I have no idea how much noise out it will generate (yet) nor what frequency it will regulate at with the modest small ratio (+) feedback

SunnySkyguy: Thanks for the response.

Do you have a reference for the positive feedback circuit with the 10M resistor and 22 pF cap? If I probe on the output of this amplifier should I see a RC decay exponential waveform? Or is a switching comparator style output? I was playing on a breadboard so perhaps this impacted the performance I was observing.

Thanks for lead on AD8028. Solid rail to rail opamp with a nice GBW. I'm not sure what the bandwidth spec is on this circuit the IRF640 can only switch at 20 MHz which is comparable to the GBW on the AD8610 of 25 MHz. I'll order a couple of samples and prototype on a copper protoboard.

 

[nickishere]

I replaced the AD8610 with a single supply rail-to-rail opamp from Linear LT1413 and the performance improved. Unfortunately, when powering the opamp from +3.3V or +5V the drive on the IRF640 is too low and the drive current saturates at around 250 mA with a external FET drain bias of +5 Volts.

In reality the FET drain bias will be from the battery so it could range up to +8.4V. Increasing the drain bias voltage (labeled ext in Figure-5) does increase the maximum current.

I ordered some Power MOSFET's (n-channel) with a lower threshold voltage so I can perhaps achieve larger drain currents. The specific FET's i'm trying out are:
595-CSD16340Q3 -
**broken link removed**

STL75N3LLZH5 -
https://www.st.com/web/en/resource/technical/document/datasheet/CD00277592.pdf

SIS414DN-T1-GE3 -
https://www.mouser.com/ds/2/427/sis414dn-244523.pdf