LMP91300: data read problem

Hi Igor,

There are a couple of problems.

The LED pin needs to be connected directly to ground (pin 6) if it is not being used.

The voltage drop of VD2 in your circuit might also be causing the problem.  In order for the SWIF communication to work, the voltage levels need to be 8V and 12V at the SENSE2+ pin (pin 13) with respect to the ground pin (pin 6).  

You will need to use external power supplies to communicate with the LMP91300 when the diode is in the circuit.

Remove U5 and U6 from the SWIF board.

Connect the power supplies as shown in the picture below.

The power supply on the right will provide the 8V.  The supply on the left is stacked on top of the 8V supply and is set to 4V.  This provides the 12V needed.

Turn on the supplies and measure between pins 13 and 6.  Adjust the 8V supply so that the voltage between pins 13 and 6 is at 8V. 

For the first write to the LMP91300 turn off the LED (address 0x72 bit 3=1).  Since you are not using the LED, you can leave it off. 

You should be able to read and write now.

Mike

Setup2.tif

Hi Michael,
I've grounded the LED pin (pin 7) and shorted the protection diode VD2. Now voltage levels and waveform of SWIF signal on pin 13 look good enough.
But the data reading is much worse: the program hangs for 2-4 seconds to get access and the data reads are wrong.
In the LM91300.pdf datasheet in p.7.5.2 (page 15) is said: "The LED terminal must be connected to ground,
either through an LED or directly connected to ground for the LMP91300 to talk to the master." But on page 32 in p.8.2.2.15 there is
an other statement: "If an LED is not needed the LED terminal can be connected directly to the V+/EXT E terminal." The V+/EXT E terminal is at 5V and the ground - at 0V, so there is a too big inconsistency...
So we simply inserted a red light emitting diode between pin 7 (LED terminal) and pin 8 (V+/EXT E terminal), just like in LMP91300NPN user guide.
Now the data reading works, although not every time.

Best wishes,
Igor Gorbounov

SWIF Read Register 0x00_.xlsxHi Igor,

The LED needs to be connected to pin 8 (V+/EXT E).  We will need to fix the statement on page 15.  

What do the signals look like?  I have attached a screenshot of reading register 0x00.  The black trace is the write to the LMP91300 and the red trace is the signal back from the LMP91300 at TP14 on the SWIF board.

Mike

Hello Michael,

this is the signal of writing to register 0x6F:

Time scale is 500 uS per scale division, the voltage scal is 5 V per division, the blue line is at 8 V, the red one - at 12V.

Here is the reading of register 0x79 (the value is 0x59):

The same time scale. The voltage scale is 2 V per division. The red line is here at 0 V, the blue one - at 3.25 V. The values are now reliably read.

Best wishes,

     Igor Gorbounov

Hi Igor,

How did you programmed LMP91300? I was not able to do it. Could you please send schematic for pin connection to programmed?

@: mehmettaspinar@endustriyel.com
Best regards,
Mehmet