Digitally controllable current source

[sudiipta]

Hi,

I want to build a current source to test my device. I want to vary current from -1 mAmp to +1 mAmp, in steps of 1 microAmp per second (i.e.2000 steps) and source these current to my device. As I want to vary the current every sec and record voltage at the same time, I want to do it using a computer. I am very new to analog circuit design and micro-controller based systems . Any suggestion will be appreciated.

With regards

Sudipta

 

[Diver300]

Digital to Analog converters usually have a voltage output, and they are usually positive output only.

You need a 12 bit D2A (well you could use 11 bits but they don't exist) and that will output 0 - 5 V or 0 - 3 V. You then need an Op-Amp to convert that voltage to a controlled current.

Do you need the current to be to ground?

What is the maximum voltage of your test device?

 

[MikeMl]

Can you do it with a PC-controlled Data Aquisition Card such as the ones made by NI or Measurement Computing? If so, modifiying one of the DAC channels to deliver constant current is easy.

 

[KeepItSimpleStupid]

We do need to know the voltage in question.

Typically one measures the voltage across the device and the current through the device. One might use Kelvin connections for the DUT (device under test). This basically means that I+, V+ are connected to the same pad and I- and V- are connected to the other. This compensated for lead resistance.

1 mA is generally the top end for 2 terminal measurements. At 10 mA, one usually uses the Kelvin technique.
With a voltage to current converter, one could have multiple ranges for better resolution.

 

[sudiipta]

@ KMikeMl

I have DAQ card Labjack U3 HV (U3 | LabJack). Do you thibk I can use this card to generate current from -1 mAmp to 1 mAmp in steps of 1 microAmp per second ?

@ Diver300
Device (metallic nanosystem) under test will have resistance around 50 ohm. Max Voltage of my test device will be in the range of micro volts to 1 milliVolt.

 

[sudiipta]

Hi,

I have a Labjack U3 HV DAQ card which has build in DAC card. But that will help to vary current by 1 micro Amp every sec to feed in my device? Voltage of my device will be in microvolt to 2 milliVolt range.

Thanks

p.s. My previous reply is still not visible on the site. Worried when this one will come.

 

[blueroomelectronics]

What is your "device"?

 

[sudiipta]

I have a Labjack U3 HV DAQ card which has build in D2A and A2D converter. But how that will help to vary current by 1 micro Amp every sec (I know that using Opamp I can convert this voltage to current) to feed in my device,? Do I need some microcontroller based solution? So that I can program using Labview or MATLAB, etc. Also How to get negetive source? Voltage of my device will be in microvolt to 2 milliVolt range. Resistance of my metallic nano structures are around 50 ohm.

I am attaching one ckt design for generating DC power supply 3-12 V as well as measuring current using DAQ card. I am using this for some other experiment. But here I have to manually set voltage using 10 turn pot . There is no PC control to give input. I do not know how to do that. ANy help will be appreciarted.

 

[MikeMl]

For the experiment, can one electrode be connected to the LabJack's ground?
You are sure that the effective impedance between the two electrode is such that it only takes +-2mV for +-1mA between electrodes?

According to the LabJack data sheet, the DAC is 0 to 5V, with 10bits of resolution (1024 steps). You need 2000 steps, so I recommend using a digital output to switch a small DPDT relay to reverse the electrodes at the center of the range, so you will sweep the DAC twice, once with +- polarity at the electrodes, and a second time using the relay to reverse the two electrodes.

If the impedance between the electrodes is as low as you say, you can convert the voltage from the DAC to a current just by using a single resistor. The value would be ((5*1000)/1024)/0.001 = 4883Ω.

You will have to write a VB program (using the LabJack supplied library) to step the voltage at the DAC output from 4.883V to 0V (1000 steps), flip the relay, and then step the voltage from 0V to 4.883V (1000 more steps). At each point, you will use one of the ADC channels to record the voltage across the electrodes. Since the input range of the 12bit ADC is 0 to 2.4V, you will need to use an instrumentation amplifier with a gain of about 1000 between the electrodes and the input of the ADC.

 

[4pyros]

I want to vary current from -1 mAmp to +1 mAmp

Wait a minute how do you get a negative current?

 

[MikeMl]

Wait a minute how do you get a negative current?

Already answered, above...

 

[sudiipta]

Hi,

I just found that we have a better DAQ card NI PCI 6221 which has 16 bit resolution, and max voltage range is -10 V and +10 V. Does it mean that I dont have to flip relay for negetive polarity? What you mean by flip replay? Can you help me with VB program? Also some example circuit diagram will be very helpful.

I am highly inexperienced in DAQ etc. I have another question, whether same DAQ card can be used for genetaing current (which will be sourced to device) and measuring voltage ?

Thanks so much

Sudipta

 

[sudiipta]

Hi,

I just found that we have a better DAQ card NI PCI 6221 which has 16 bit resolution, and max voltage range is -10 V and +10 V. Does it mean that I dont have to flip relay for negetive polarity? What you mean by flip replay? Can you help me with VB program? Also some example circuit diagram will be very helpful.I also have one PCI DI024, whether this will be also helpful?

I am highly inexperienced in DAQ etc. I have another question, whether same DAQ card can be used for genetaing current (which will be sourced to device) and measuring voltage ?

Thanks so much

Sudipta

 

[MikeMl]

Reread my post above. I specifically told you how to create a good-enough current source starting from the voltage output of the DAC.

 

[sudiipta]

Hi,

I need one more clarification,
Labjack analog output can be set to a voltage between 0 and 5 volts with 10-bits of resolution. Ni PCI 6221 analog out put Maximum Voltage Range -10 V , 10 V. Does that mean I can go from -10 V to 0 V in 1000 steps and then 0V to +10 V in another 1000 steps? I can use single 10 KOhm resistor? (10 V/10000 Ohm = 1 milliAmp or 0.01 V/10000Ohm = 1 micron amp). So If I use NI card, in that case do I need to use a digital output to switch a small DPDT relay to reverse the electrodes at the center of the range?

 

[MikeMl]

You said: "I just found that we have a better DAQ card NI PCI 6221 which has 16 bit resolution,"

This dac covers the range from -10V to +10V in 2^16 = 65536 steps. No reversing relay is needed. At full scale, Vout = 10V, I = E/R = 10/10,000 = 1mA, so the range would be -1mA to +1mA. The smallest step of voltage is 20V/65536 = 0.305mV, so the smallest current step would be 0.000305/10000 = 0.0305uA

 

[KeepItSimpleStupid]

On the spec sheet, I didn't see the encoding mentioned. Don't know if it's 15 bits + sign or 2's complement which can throw off the resolution per bit slightly.

I also need to ask whether you device can generate current. If it can then a 4-quadrant power source is needed.

You might want to consider a true bi-polar current source.

With your requirements of uV to mV and 2 mA instrumentation error will figure predomenantly.

It almost seems as if you would need a nanovoltmeter and a precision current source. Take a look at what Keithley Instruments Inc. - A Greater Measure of Confidence has to offer.