AC power measurement design not working right

Discussion in 'The Projects Forum' started by electrophile, May 11, 2016.

  1. electrophile

    Thread Starter Member

    Aug 30, 2013
    100
    1
    I built this design to measure AC power (attached). The voltage is measured pretty much the normal way only its stepped down to show about 1.4Vpp for 240V. The current is measured by measuring the voltage across the 4 milli-ohm shunt. The output that goes to the MCU has a DC offset of 0.7V which is then removed in the software. I simulated this in LTSpice and it showed the expected results.

    I built this design today and while I was testing it I observed the following:

    The VSENSE-N and ISENSE-N is where the DC offset of 0.7V is added however I measured -2.9V DC here. Also I see only the positive half of the AC voltage at the V-MCU pin and I'm not sure why.

    The AC waveform at the I-SENSE-P pin is also offset by about -3V DC but I do not see any waveform at the I-MCU pin.

    I used a 40W bulb as the AC load. The LMV324M is powered by a +/- 3.3V supply. The -3.3V comes from a charge pump dc to dc converter TC7660S.

    What am I missing here?
     
  2. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,531
    1,248
    What's missing is a defined ground relationship among the two opamp circuits, the split power supplies, and the signals being measured. Both the positive and negative signals for each signal must be within the common mode input voltage range of the opamps.

    Where does the 8 Vdc come from and what is the relationship between its return and the AC power line signals?

    Also, the two differential amplifier circuits have multiple problems. The LM324 datasheet has a schematic for this on page 22.

    Also, the LMV324 is not rated to operate on 6.6 V.

    Also, this thread violates the terms of service for this forum. The three 470K resistors in series show an above-average awareness about how to do this task, but it remains dangerous.

    ak
     
  3. electrophile

    Thread Starter Member

    Aug 30, 2013
    100
    1
    Thanks @AnalogKid. Here are more specifics. Please bear with me since I've formally studied this a long time ago but I'm very rusty now.

    I researched about this and I am still a little unclear. Could you please explain this a bit? I realize that there needs to be a reference against which the voltages are measured however what I don't understand is where is this ground reference located (image attached)?
    [​IMG]
    Yes I think these are within that range. The datasheet lists 1.7V as the max value for that however, the min value is 0V and not a negative voltage.
    For the purpose of prototyping this 8V is coming in from an external DC adapter, however, the end product will have an onboard flyback power supply.
    Could you post the link to the document? The one from TI doesn't seem to have one.
    The TI datasheet says the maximum differential supply voltage is +/- 5.5V.
    Oh I did not know this. I know I could use an ac-to-ac and a current transformer to convert the voltages and current and provide isolation however what other options would be available for space saving and minimizing footprint area?
     
  4. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,531
    1,248
    Better if you post your datasheet so we can see what you have to work with. All LM324 and 358 datasheets I've seen over the last 45 years have some form of the original National Semiconductor application circuits.
    No, it doesn't.

    ak
     
    Last edited: May 12, 2016
  5. electrophile

    Thread Starter Member

    Aug 30, 2013
    100
    1
    Here is the link to the datasheet. How do I interpret the supply voltages on page 4 under Absolute Max Ratings?

    Also am I doing the ground referencing correctly?
     
  6. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,531
    1,248
    A problem with the original 324/358 datasheets is that although the parts function as standard opamps, National decided to call the negative power input GND, rather than Vee, to emphasize the single-rail capability that was a huge deal at the time. The version 1999 you posted shows Vcc as +5.5 V absolute max above the GND pin, as stated in Note 2. My 2014 version of the same datasheet calls the power pins V+ and V-, and shows the max operating voltage as
    V+ - V- = 5.5 V
    So +/- 3.3V won't work because
    +3.3 - -3.3 = 6.6
    The two datasheets state the same thing in different ways, but either way it is not rated for 6.6 V operation.

    ak
     
  7. electrophile

    Thread Starter Member

    Aug 30, 2013
    100
    1
    Ah I see. Would the regular LM324 work instead of the LMV version?
     
Loading...