Post Go back to editing

2KV High-side, low-current measurement

Is it possible to use an AD620 connected to an AD210 to measure high-side current through a sampling resistor?  With the REF line on the AD620 and the Icom line on the AD210 connected to the bottom of the resistor (and the top of the resistor is connected to 2KV).  The REF/Icom voltage measured with respect to ground would be very close to 2KV. 

If I were to do that what ground referenced voltage would I measure at +VISS on the AD210?  Would it be 2KV - Rsense*Iload + 15V?  That is, assuming I didn't let the smoke out.  ;-)

Are there any common mode limits I need to be careful to watch out for?

  • Hello Andy,

    ADI has a family of isolated S-D ADCs tailored for current monitoring. The newest device is the AD7403. If the high-side voltage is a DC level then the isolation provided by the AD7403 should be adequate. This would be simpler, smaller and less costly solution.


  • Thanks DaveC, I pulled the datasheet and the AD7403 is a very nice part.  Since I need to get to digital values eventually I was excited to see that too.  Unfortunately, I don't know how to power the input side.

    The designs I have seen typically use a zener to produce voltages of the right separation.  That is, VDD1 would connect to the 2KV power supply output and GND1 would connect to the anode side of a 5.1V zener.  This floats the whole input side up into the 2KV range.  Similarly, VIN+ would connect to the 2KV power supply output and VIN- would connect to the bottom of the sense resistor.  Nothing wrong with that as far as I can tell... at least with respect to the voltage levels.  Is that how it would typically be connected?

    BUT, here is my problem.  The high-voltage supply I would like to use is only 2.4 mW.  I don't think I have enough output power to drive the Zener into breakdown and power the AD7403.

    The AD210 is attractive because I don't think I need to rely on the high-voltage supply to provide the power for the device itself.  The part derives the isolated power for the HV side with it's 3-port configuration.  I'm not entirely confident about the connections on the HV side.  Of course when confronted with an $80 part, it might be easier, cheaper, to use a HV supply with a little more oomph.

    I haven't used any of these parts before so I don't know their capabilities and limitations... particularly related to the high-voltage side. 

  • The ADuM6000 can provide an isolated 5V supply with 100 mA output current. It's ~ 30% efficient, so you'd need to provide enough power on the input side to account for this.


  • Awesome, thanks DaveC! I'll take a look at that.  Good advice.

  • Maximum working voltage for ADuM6000 is 600V DC [Analog Devices, ADuM6000 Datasheet Rev. D, Table 11, pg. 8,, accessed 14 August 2015].

    Is there a way to derate the 50 year lifetime if I intend to use 2KVDC?

    If I plan to build something and get UL listing will I be compromising that plan?

    There is also a very nice application note on this part that includes what appears to be a very robust design for high-side current measurement.  [Analog Devices, Circuit Note CN-0280,, accessed 14 August 2015]  This Circuit Note has a complete schematic and board layout advice.

  • The 600V DC working voltage is very conservative, but there's another factor to consider (which I did not in my initial response): creepage. Using the IEC 60664-1 insulation coordination standard, a 2 kV DC working voltage requires 14 mm creepage. This is assuming basic insulation with the equipment operating in a pollution degree 2 environment.

    You can run two isolators in series to effectively cut the working voltage in half for each component. This reduces the creepage requirement to 7.1 mm, which any of our 16-SOIC packages will meet. The ADuM6000 could support 1kV DC working voltage. ADI has published an app note with an example of this approach. The inefficiency of the ADuM6000 is more of an issue. You would need to place a 2-channel isolator in series with the AD7401A, such as the ADuM2281.

    Sorry I didn't think of this sooner.


  • DaveC, thanks again.  I am beginning to reconsider trying to monitor current.

    creepage... there's a term I'm going to have to ask Google about... same with pollution degree 2 environment... I sure hope I don't kill somebody... like me... although 2.4 mW isn't much more than a carpet spark.