ideas for adding current/voltage measurements to my load

[croyleje]

HI everyone I am working on adding current/voltage measurements to my DC load and have been looking around on line for ideas.  I have come across several different setups but most are very convoluted basically I want to be able to see actually current from my supply and voltage without having to use my DMM.  I already have an LCD display so changing code to add the displays isn't an issue just need to work out how to do the measurements.  I will try to draw a simply schematic of circuit but it is basically the one Dave used on his dummy load but I am using a Cypress PSOC PWM DAC to drive the FET.  Any ideas would be greatly appreciated.

Thanks Jason

[theatrus]

Voltage: resistor divider feeding an ADC.

Current: Measure the voltage over your current sense resistor.

[VK5RC]

In a recently built power supply (a linear 'dinosaur', 35V 15A) I used a high side 0.01Ohm sense resistor with a LT6106 amplifier to get the sensed I (voltage) up to a reasonable level to stick into an arduino that ran the display, and an over voltage shutdown. Used a simple voltage divider to get the voltage down to Arduino levels.

[Asim]

Dave was measuring the current by measuring the voltage across the 1 ohm resistor. It is a direct conversion . 1 volt corresponds to 1 amps.

As for the voltage use a voltage divider as the other guy said , just be aware of the voltage going to your ADC not to accede the maximum voltage of the ADC

[Harvs]

When I get the time (thesis due on Friday, so it's going to be after that) I'll be writing up a little project I did with a PSOC 4200.  It's a DC power meter with auto ranging voltage and current measurement using the mux buses to switch the ranges.  It's simple to build and use because all the switching and opamps are done internally.  Only needs resistors externally.

You could still get the PWM out of the device on one of the few spare pins, but both op-amps are used.  Anyway I'll probably write it up this weekend so it may give you some ideas. 

[croyleje]

The voltage isn't much of an issue i will be using simple voltage divided parallel to the input and feed that into the ADC and at 12 volts your only looking at about 100uA so it wont effect the load current much.  But if my logic is right you have to sense the actual current with a separate sense resistor other then the one ohm resistor because you will always have the voltage of the op-amp across that resistor.  So to read actual current i was thinking of adding current sense resistor before the FET.  Below is a copy of the basic dummy load circuit not the one I am using but the design is the same.  also any ideas on how to sense the voltage without effecting the load current at all would be nice but uA is not really an issue.  Any input would be great thanks.

Jason

[croyleje]

also HARVS I would be interested in seeing your design even if it is in the future please keep us in mind if your draw it up

have a wonderful day everyone

Jason

[zapta]

You can use a I2C IC such as the LTC2943 all it needs is a shunt resistor (50millihom for 1A range, 25milliohms for 2A, etc). You can close your voltage regulator feedback after the shunt resistor to compensate for the shunt voltage drop. The IC comes out of the box with pre calibrated voltage and current A/Ds (for current it has two versions, integrating Coulomb counter and a momentary A/D). For better shunt accuracy use a kelvin shunt such as Digikey FC4L64R025FERCT-ND.

Edit: here is a schematic of a 0-2A arduino compatible power supply that uses the LTC2943 https://github.com/zapta/power-monitors/blob/master/pmon_oled/eagle/pmon_oled_schematic.pdf?raw=true . It also uses a 1.3" OLED display to show the current and voltage.

[theatrus]

The voltage isn't much of an issue i will be using simple voltage divided parallel to the input and feed that into the ADC and at 12 volts your only looking at about 100uA so it wont effect the load current much.

With very high value resistors, don't discount the load placed on the sense point by the ADC. For simple circuits you can ignore it, but buffering it with a unity-gain op-amp is an option.

But if my logic is right you have to sense the actual current with a separate sense resistor other then the one ohm resistor because you will always have the voltage of the op-amp across that resistor.

I don't follow this logic - the op-amp is trying to drive a constant current through the sense resistor, which it sees as a voltage difference between that node and ground. Remember the laws of op-amps!

[Harvs]

Hey Jason,

But if my logic is right you have to sense the actual current with a separate sense resistor other then the one ohm resistor because you will always have the voltage of the op-amp across that resistor.

Nope you're logic isn't right.  You can just sense the voltage across that resistor.  It has one end hooked to ground, and all the input current passes through it.  Therefore that's exactly how you want to measure your current.

As theatrus said, if you've got one of the opamps spare, use that to buffer the voltage at the high impedance node of the voltage divider.

[nctnico]

When an opamp is used as a buffer you'd need to use one with a low offset.

[croyleje]

Hello again everyone maybe I am over thinking this to much but if the op-amp non inverting input is 1V then your going to get 1V across the 1 ohm resistor = 1 amp but if the load under test can only supply 500mA you will still show 1 amp across the 1 ohm resistor so to be able to measure actual load amperage wouldn't you have to sense its current before the FET.

Jason

[theatrus]

Inputs don't (*ignoring bias) sink or source any current. The output of the op amp is tied to the gate of a FET.

[croyleje]

Forget I asked I am an idiot set current would be measured at non inverting input of op-amp actual current across 1ohm resistor making things more difficult then they are is something I am really good at but thank you for the help.

PS Hasz I believe it was I would really like to see your design when you write it up.

Thank you for all the help next time just tell me I am  an idiot.

Jason

[theatrus]

No need to call yourself an idiot - it's all part of the learning process

If you can breadboard this with opamps and a pot instead of the arduino (or use the arduino), you'll be able to see how everything interacts.

Best of luck!

[gardner]

I have a couple of projects that need to monitor the voltage of a 12v lead-acid battery.  For lead-acid, the "interesting" voltage range is in the 11v..15v range, and the whole way from 0v..11v is uninteresting.  To concentrate my crappy A2D resolution where I need it, I use a 10V zener on the high side to act as a voltage reference.  Then I can feed the remaining 1v..5v voltage difference to ground into the A2D.

There are some down-sides:
-- you need software calibration constants for your diode
-- you may need software zener dynamic impedance calibration for your diode
-- you may want software tempco calibration for your diode
-- you have to ensure there's enough current through the zener to keep in the stable range, and that could be a couple milliamps of extra drain on the battery

But it's simple and cheap.