ADS1148: correct communication sequence?

Joseph,

many thanks for your answer.

I changed my code according with your suggenstions.

Now, I do the following steps:

• Initialize the SPI communication with the clock High when communication is Idle (data clocking at the falling edge)
• hardware reset of the device by putting low the /RESET pin for about 6us
• pull High the START pin
• Send the RESET Command (0x06) to the device
• Write the configuration registers: in this case, i want to measure the external VREF between REFP0 and REFN0 pins
  • /CS low
  • first, I send the 2 WREG bytes
  • then, I send the 15 byte of the configuration registers
  • /CS high
• Send the SELFOCAL command (0x62)
• 20us delay
• Toggle START pin:
  
  • START low
  • 10us delay
  • START high
• while cycle to wait the DRDY pin going low
• /CS low
• Send through SPI 0xFF (NOP command)
• SPI receive byte function and save the byte into a variable
• Send through SPI 0xFF (NOP command)
• SPI receive byte function and save the byte into a variable
• /CS high

Here there is the behaviour  of the CLK, DRDY, STARD and CS signals:

Looking the shot, it seems that the START pin is toggled BEFORE the communication, but I think it is from the following cycle of operations . I can't show in the same shot the propper START pin going low and high.

Below I show the DIN and DOUT in respect with the CLK

On the DOUT pin I can see something, perhaps the NOP bytes, but on the DIN pin I'm not able to see anything..

May I forgot some passages or did any mistakes?

Thanks again,

Andrea S

Andrea,

Make sure the SCLK dwells low (that SCLK starts low and clocks up for a pulse). It is not set up that way according to the scope photo.

It might affect the data clocking out and clocking in register values.

Joseph Wu

Joseph,

1. yes, you are right about the behaviour of the SCLK when idle: I modified the SPI initialization, and the new scope photo of the communication is shown here:

2. but now I have a doubt: I'm trying to measure an external VREF (about 1,5V) between REFP0 and REFN0 of the ADS1148; in order to write the configuration registers, I write the MUX1, IDAC0 and IDAC1 registers, at the address 02h, 0Ah and 0Bh rispectively. To do this, I send the WREG command and after that the single byte through SPI, and repeat it for three times, in this way:

/CS is low here
ucArrayToSend[0] = 0x42;
ucArrayToSend[1] = 0x00;
ucArrayToSend[2]  = 0x35;	// 00110101	(MUX1 - ADDRESS 02h)

for (i = 0; i < 3; i++)
 ADS_1148_SendByte(ucArrayToSend[i]);  // this is the SPI send byte function
//--------------------------------------------------------------------------
ucArrayToSend[0] = 0x4A;
ucArrayToSend[1] = 0x00;
ucArrayToSend[2] = 0x01;	// 00000010 (IDAC0 - ADDRESS 0Ah)  IDAC1=50uA
for (i = 0; i < 3; i++)
	ADS_1148_SendByte(ucArrayToSend[i]);  // this is the SPI send byte function
//--------------------------------------------------------------------------
ucArrayToSend[0] = 0x4B;
ucArrayToSend[1] = 0x00;
ucArrayToSend[2] = 0x3F;	// 00111111 (IDAC1 - ADDRESS 0Bh) //IDAC1 connected to AIN3, IDAC2 disabled.
for (i = 0; i < 3; i++)
	ADS_1148_SendByte(ucArrayToSend[i]);  // this is the SPI send byte function
//--------------------------------------------------------------------------
/CS go high here

After that, I want to start the communication (toggle START pin, wait for DRDY to go low, send RDATAC command) and receive the measured VREF voltage an the REFP0-REFN0 pins, but if I trigger the oscilloscope on the MISO signal (DOUT), I can see data that don't make sense. My VREF (REFP0-REFN0) is about 1,5V (measured with a multimeter), but the reading of the oscilloscope doesn't match with it. Perhaps the configuration is wrong? Have I to configure other registers?

my steps in this step are:

• pull the START pin low and high
• while cycle to wait the DRDY to go low
• /CS low
• send RDATAC command through SPI
• send 0xFF through SPI
• receive a byte from SPI
• send 0xFF through SPI
• receive a byte from SPI

Could anyone tell me what I'm doing wrong?

Here is the scope photo of the communication:

Many thanks!

Joseph,

Thank you for your exhaustive answer: now I succeded in measuring the VREF between REFP0 and REFN0. I did the steps thant i described in my last post, but with the following differences:

• I send the RDATA command instead of RDATC command
• I wite the configuration registers every loop cycle: it seems it is not sufficient to write them one time at the start. Could you tell me why? It is the expected behaviour?

As you suggested, I will now try to change the configurations, and try to measure a signal connected to the AIN0 and AIN1 (that is my main purpouse), disabling the internal reference and using the previously measured VREFas voltage reference.

Thanks again!

Andrea

Joseph,

as I wrote in my last post, I tryied to measure from the AIN0-AIN1 input (I have a resistor connected simulating a PT1000 sensor). I followed the same steps as I did previously with the VREF (REFP0-REFN0) measurement, but I obtain values from the conversion that doesn't have any sense.

I'm writing the configuration registers as follows:

(MUX0 - ADDRESS 00h)  <--  11000001      My measure is between AIN0 and AIN1

(MUX1 - ADDRESS 02h)  <--  00100000     I use internal oscillator

(SYS0 - ADDRESS 03h)  <--  00100000     I set the PGA to 4  and the data rate to 5SPS

(IDAC0 - ADDRESS 0Ah)  <--  00000010   The IDAC is set to 100uA

(IDAC1 - ADDRESS 0Bh)  <--  00111111    The IDAC output is AIN3

At this point, I think there should be something wrong in configuration shown, but I don't really understand what. Can you see something that doesn't match with my measure on the AIN0-AIN1 input?

Thanks again,

Andrea

Joseph,

this is my schematic:

• The reference resistor is 15Kohm, I selected an IDAC value of 100uA, so the Vref (REFP0-REFN0) should be about 1500mV (and I can read it correctly with a multimeter).
• I thought the IDAC current should be routed to AIN3 because, looking at the schematic, that current flows through the PT1000 and generate a voltage differences that I can measure between AIN0 and AIN1. Is it correct?
• For example, a resistance of 1,7Kohm is connected to the inputs, simulating a PT1000 value. Measuring the voltage at the inputs with a multimeter, I can read a voltage of 160 or 170 mV. Running the program, the ADS1148 reads a value of about 0xC6A3 (50850d). I think it is not correct, because this value corresponds to a voltage of: Vin=Vread*VREF/(PGA*32767) = 50850*1500/(4*32767)=581mV , that is different from 170mV I measure on the board. Perhaps there is any error in the formula I used?

Joseph,

you are right! I was measuring a negative voltage at the input! I changed the MUX0 register setting the AIN0 as positive input and the AIN1 the negative one, and now the value the ADS1148 reads is a positive and right one. Thanks!

Just another last question. As I mentioned before, I am now able to measure the potential difference on my PT1000, and it is almost the same as the one I measure with a multimeter on the board (OK). Since I would like to measure a temperature, I calculate the resistance of the PT1000 as

R(PT1000) = (AIN1-AIN0)*100uA  (I selected a 100uA IDAC),

but I obtain a resistance value that is not really the same as the one measured on the PT1000:

At ambient temperature:

• R(PT1000) measured with a multimeter: 1096 ohm
• R(PT1000) calculated by the ADS1148: 1050 ohm

Since the (AIN1-AIN0) is well measured from the device, I think that there would be an error in the current, perhaps it is not exactly 100uA. Can it be possible? Which can be a possible cause of this error?

I would like to solve this problem, because in my system it means some °C celsius of difference.

Do you have any suggestion or idea?

Many Thanks,

Andrea