What are the benefits of shunt summation with the INA282

BP101 said:

...please explain the added benefits if any of monitoring shunt current in summation...

Greetings - if I may - there are a number of INA data-sheets currently, "in play."   Thus - minus data-sheet identification - here's my opinion:

I believe the most critical portion of your question revolves around the use of "two" - rather than a single - INA device.   And - I believe the inherent cleverness of this approach is the ability to "simultaneously" perform a "combined current measurement" - across (even more) than two INA devices.   Key to this capability is the routing of signal output (serially) between devices - such that the final device outputs a "summation" of total currents.   (and my belief is that this may even extend to a subtraction from the total - should one/several INA devices "in the chain" report/pass on negative currents.)   Note: this subtraction may only hold true if the individual INA device is properly configured to accept, process & pass-thru, "negative current flow."

I've limited "first hand" experience w/this device - this report may qualify as a (somewhat) educated "swag."   Note too that I've avoided the specifics of your issues (1 & 2) preferring to establish some, "Multi INA baseline" understanding/agreement - prior to expert, experienced, vendor analysis/commentary.

The synchronous nature of the dual (or extended) INA "chain of devices" is what "justifies" (to my mind) the use of two (or more) INA devices.  In time variant systems - such may prove crucial...

The alternative - routing sensed signals into two (or more) of an MCU's ADC channels may void the timing synchronization.    A "Difference or Summing" op-amp configuration should also accommodate the desire for summation - yet may breach the convenience, accuracy & speed of the INA solution.

Perhaps higher up's might need to spend some lab time to answers questions 1 & 2 as might the INA282 pertain to multi phasic current monitors.

Our opinion now swayed from data sheet 1 (11/09 Revised 09/12) - summation or even differential mode as depicted Figure 43 is not well suited for multi phasic 1/2 bridges. That seems indicative of failures to answer question 1 or 2 in being fully qualified reference to real time events of each phase being monitored. 

It would seem same of data sheet 2 (May 2015) illustrations directly connected INA circuits missing totem pole illustration. The earlier (9/12 Fig 43) seemingly pertain only to a power supply single phase 1/2 bridge negate the transformer primary. Our survey instead believes illustrations (May2015) Figures 40-42 do not consider multiple phases and will lead to incorrect current readings in the connections between INA devices of multi phasic inverter bridges.

Hence there seems missing any (warning) or explanation INA282 summation or differential circuit connections cross fire drive pattern indicative of 3 phase inverter 1/2 bridge circuits. User beware (9/12) data sheet including totem pole Figure 43 seemingly does not accurately explain proper use figure 42!   

Hi Mariam,

Green does not infer all the questions were answered yet some reward is given, well deserved!

This scope captured envelope 2x seems to indicate a Single low side INA282 is capable of measuring half bridge Bipolar current in both directions. So it appears the high side FET and low side FET current artifacts exist in a bipolar division developed in the INA switching capacitors. There is no need for summation unless it provides some kind of benefit not yet answered in the posted question.

This blue square represents the PWM wave form event of the Bipolar FET half bridge and the yellow/red circles indicate bipolar 80us/12.5kHz PWM periods. The main voltage rise sets the current threshold and the pulses seem to indicate bipolar FET avalanche current is detected without summation. What is missing in this capture is a sample hold of any linear ratio metric voltage slope in the INA282 output, that which summation may provide?

 Captured signal 2015 data sheet Kelvin connection (-IN) via to ground plane seemingly picks up both sides of the Bipolar switching artifacts. 

Same signal with very thin gage wire 0.250" long (twisted pair) lead into shunt Kelvin connection:

80us is the pulse width of any single pulse. Tektronix storage display playback 1987 did not change the horizontal time base to reflect save history. 1 yellow circle very top is 80us pulse. Other 80us pulses exist outside the blue box exist outside the ADC sample window.  Those stray pulses are seemingly artifacts occurring from 2 other active phases sharing a Delta phase coil array.

Hi Mayrim,

The point of posting does not specifically relate to motor control - more so to the INA282 current monitor behavior as outlined May 2015 data sheet.

The question remains; Why are two INA282 better than simply one INA282 in any current monitor scenario, remains unanswered.

The (May 2015) datasheet fails to address how the INA282 might properly or more accurately monitor current in push pull versus pure bipolar switching scenario past suggested in (Sept. 2009) data sheet. Seemingly the concept behind wheeling capacitors address several areas in that question yet the data sheet takes a benign approach of AC sine wave analysis. AC sine wave has zero volt crossing events where pure DC bipolar pulse wave, zero crossing events occur roughly 1/2 DC supply far above ground. It is our express opinion the INA282 has not been lab tested to reveal the attributes characterized in the (above capture) simply described by (May 2015) datasheet Figure 41 or 43. The best way in discovery is to experiment with the INA282 as we have this past year by trying several configurations, some better than others.

The point in this further posting in part, feedback for TI experts perhaps for performing inclusive analysis on the product for the next revised data sheet.