- Ask a related questionWhat is a related question?A related question is a question created from another question. When the related question is created, it will be automatically linked to the original question.
Hi Ti experts,
We're using INA199A1DCKR as our current sensing application on motor.
Below is our schematic-
While we were using INA199A1DCKR measuring the voltage across shunt resistor on motor output pin from motor driver,
we found the output of INA199A1DCKR acts differently when the load impedance is different.
Ch1 (Red) is the waveform of current clamp Hantek CC-650 (bandwidth 20kHz)
Ch2 (Yellow) is the waveform of INA199A1DCKR output
At 60ohm, 100ohm -
The output voltage waveform rising and falling timing are almost changing concurrently.
But when we changed the load down to 10ohm or 20ohm,
The output voltage waveform did rise simultaneously, but there's a timing delay 15~20us on the falling edge between 2 channels while we did not find this symptom when load is above 60ohm.
Would you kindly help us to investigate why there's such a timing delay between the current amplifier INA199A1DCKR output and measured current (by Current Clamp)?
Motor impedance is roughly between 10 to 20 ohm. So the nonlinearity happens for our application and it causes difficulty for firmware development.
Hi Chieh Wei,
Can you attach the schematic and scope photos again? They did not come through. You can use the "Insert/Edit Media" or "Insert File" button.
Regards, Guang
Hi Chieh Wei,
Thank you for the concise yet to-the-point investigation report, well done.
It is important to point out that the configuration shown is what we call “in-line” current sensing. It is the most challenging from the perspective of the Current Sense Amplifier (CSA), due to the fast switching common mode voltage. As a general rule of thumb, a specially designed CSA such as INA240, has to be used; if an amplifier fits such applications, the datasheet will specifically spell out.
The INA199A1 is not an ideal choice for your application for a number of reasons. Most of the troubles you’re facing are likely related to this limitation. I strongly recommend evaluate and eventually migrate to INA240. However, below is my input regarding the current debug work.
The output takes ~20uS to settle when the common mode (5V) switches with 0V input differential, as shown in figure 15 of the datasheet. With 12V, it is going to be worse. Besides, the disturbances will vary with the magnitude of the differential input. This is likely the primary reason why 10Ohm and 100Ohm loads give different delays as you observed.
I don’t have a good explanation why the “nonlinearity from 2.55V ~ 2.65V” exists for 10Ohm, but not for 60Ohm. And it is puzzling why the output did not seem to move at low duty. although It could be related to the same reason mentioned above. Can you repeat the test for small duty cycles but instead monitor the actual common mode voltage (on top of the load), instead of the control input?
While doing the test, can you possibly test/observe if the INA199A1 is getting noticeably warm or hot? It will be interesting to know.
Regards, Guang
