Part Number: INA4230
Other Parts Discussed in Thread: CS-AMPLIFIER-ERROR-TOOL
hi sir,
Currently, in our validation, the INA4230 power error is higher than 1.5% of the datasheet.
It seems to be related incorrect settings.
Please provide adjustment suggestions to us to improve accuracy, thanks.
Hi Zeer,
Apologize for the confusion, when I said maximize the conversion times and averaging settings, I meant set conversion time to the longest (8244µs) and averaging to largest (1024).
Your current register settings seem correct, however, with a current_LSB of 0.0004, you are only utilizing half of the ADC range (80mV), this could add to the error we are seeing. Along with this, the INA4230 measures bus voltage at the IN- pin, check if your DAQ is measuring power after the shunt resistor or before it. Note, as mentioned in the FAQ linked above, the device and system must share the same ground and layout is important to minimize errors. Please share your schematic, layout and maximum current measured in your system.
Best,
Mallika Senthil
hi Mallika,
Could you let us know which register we provided is correct?
The measured Rsense between DAQ and INA4230 is the same one.
Some condition:
It is common GND already.
Estimated max current is 10A.
Rsense is 3mOhm.
The estimated max is 30mV. That’s why we configure ADC range to 80mV.
The variation is still not improved if DAQ bus voltage is at the IN- pin (load side) and IN+ pin (source side).
Below are sch and layout for INA4230.
Left side is power IC.
Here is INA4230
Here is routing.
The center U403 is INA4230.
Please help us with the problem we are currently encountering.
Another question:
Regarding power monitoring accuracy:
– 1.5% full scale (maximum).
Does this mean the error is within 1.5% only at full scale (maximum 20mV or 80mV)? Is there a curve description for what happens if the voltage is less than the maximum? could you share to us, please? thanks.
thanks.
regards,
zeer
Hi Zeer,
The power total measurement error (PTME) is defined as sum of Bus voltage gain error (GBERR) and Shunt voltage gain error (GSERR), i.e. PTME = 0.75% + 0.75% = 1.5%, at full scale voltage (48V) and current.
So the PTME is composed of two different errors, one from bus measurement and the other from shunt measurement. My earlier comment "you appear to be using 0.78V bus voltage, which is a very small fraction of the full bus voltage, this is probably where the bulk of your error is coming from" hypothesizes that most of the PTME error is coming from bus voltage measurement.
You can check this by reading from the shunt_voltage and bus_voltage registers and calculating their individual errors to see which one dominates.
Unfortunately we don't have a curve to error over shunt voltage, however, we do have the error calculator tool [www.ti.com/.../CS-AMPLIFIER-ERROR-TOOL], which we can use to find total error% over current.
Best,
Mallika Senthil
hi millika,
we will have more experiment to validate, and will consult to you later.
please you share again, link is not complete, thanks
hi mallika,
we validated different power rail with VSYS (it is 20V) by DMM and INA. It has quite difference, and should be something wrong.
DMM: 6.5W.
INA: 27W.
Below is design and setting. could you review again and give us suggestions, please, thanks.
Hi Zeer,
1. What is the correct expected power measurement?
2. Where did you probe using the DMM? Please probe the points circled in red.
3. What is the max expected current you used for calculations? Was it ~16.3833A?
Your schematic looks good, the filtering capacitor of 10uF is a little too large, perhaps a 0.1uF may be better, but this is not the likely cause for the error we're seeing.
I'd like to check the layout as well, please share. One more thing to review in terms of layout, since you have 3 shunts in parallel, layout is critical to ensure that the current is split equally. Please see the following app note: https://www.ti.com/lit/an/sdaa115/sdaa115.pdf
Best,
Mallika Senthil
hi mallika,
the measurement point is here,
below is layout.
Please give your suggestions, thanks
if possible, please share your email address, i could send picture, please, thanks.
Regards,
zeer
Hi Zeer,
Thank you for your patience. Layout and schematic all look fine, where you're probing with the DMM is also fine, but I would check that the DMM has enough resolution to measure 113uV.
Please provide a complete register readout, I want to check if the current and voltage measured by the device are correct. Power is calculated off these measurements, so next step is to check whether the issue is in the current or voltage measurement.
Also, please note, you have set max current to 16.5A, but you are measuring 0.37A. This is a very small value in the range you've set. Also the device offset is ~75uV. So if your DMM is measuring 113uV, then the device's offset will add 75uV (max) to this and INA4230 will measure 188uV across shunt.
I will send you a private message with the team's email id on e2e.
Best,
Mallika Senthil
