TLE4972 CUR SENSOR PROGRAMMER cannot drive TLE4972 AOUT into data mode

[nwernicke]

When using the software for the TLE4972 STD EVAL kit, the CUR SENSOR PROGRAMMER data pin (the one that goes to TLE4972 AOUT) drives what might well be correct voltages when not connected to the TLE4972. The voltages look like logic levels from Vdd to Vss with the timings described in the TLE4972 manual.

When connected to the TLE4972, the oscilloscope shows a sequence of spikes around Vdd/2 or 1.65V, as if a low impedance digital output is being driven into the low impedance TLE4972 analog output. Nothing is read or programmed.

TLE4972 needs to be driven into data mode by sending 0xABCD on the AOUT pin of the TLE4972. The TLE4972 AOUT pin is typically a low impedance output. It seems the CUR SENSOR PROGRAMMER is not able to drive the AOUT pin into data mode.

1) Does this description of the problem make sense at all? Can I provide better context?

2) Am I doing something wrong? What could that be? 

3) How do I know the CUR SENSOR PROGRAMMER works? Maybe I need new firmware or...? Please help

[Yashraj_P]

Hi @nwernicke ,

How are you programming the sensor, have you referred to Current Sensor User Guide. 
Also, share the oscilloscope captures you are referring to.

Best regards,
Yashraj

[nwernicke]

Here are some images. First you see our Infineon programmer. Its purchased from RuTronik (https://www.rutronik24.com/product/infineon/cursensorprogrammertobo1/13614755.html).

The black circuit board looks scary, but the sensor is disconnected from everything on it (including power). We want to evaluate the TLE4972 with the toolkit-only before doing a new revision that connects our circuitry to the TLE4972. The TLE4972 on this board only connects to the Infineon Programmer for power and signals.

The next pictures are the waveforms of the AOUT signal from the programmer/TLE4972 when the TLE4972 is connected. The last waveform is the programmer AOUT pin with the TLE4972 disconnected. As you can see, it looks good when not connected, bad when connected. Possibilities are: the connection is incorrect, or that the programmer is damaged in some way.

I tested the connection with a multimeter from the programmer to the pin pads on the TLE4972 package, and they are correct. 3.3V, AOUT, and OCD2. The TLE4972 reads current accurately in the Infineon software when test current is applied through the busbar, and linear analog voltage appears on both AOUT and VREF pins with a multimeter.

I believe the connection must be wrong somehow, such as AOUT connected to Vref or something, but I am unable to find where it is incorrect.

 

 

 

 

 

[Yashraj_P · Accepted solution]

Hi @nwernicke ,

Please check if you are supplying the Vin to TLE4972 sensor using the V33 pin of the programmer instead of the VSENS pin. The programmer needs to control the sensor supply to put the sensor in test mode; it does this by sending the ”enter interface” command on SICI (AOUT) within a short time from power up and controlling the supply through the VSENS pin. If the sensors are connected to V33 and not VSENS, the test mode is not entered, the AOUT pin of the sensor never goes in high-z mode and when the master tries to write on AOUT driving conflicts happen on the bus, therefore the strange behavior of the analog pin.

Best regards,
Yashraj

[nwernicke]

I incorrectly connected the programmer power to 3.3V from the programmer instead of Vsens. Vsens needs to power the sensor so the programmer can set AOUT low as it enables Vsens so the TLE4972 comes up in programming mode.