ADS1259 Clocking Out Data

Hi Edmond,

You are correct - keep DIN low while clocking out the data (i.e. sending 0x00 over SPI) to retrieve the data. Your sequence should work.

A few possible causes I would look into:

• Make sure /CS is being held low during the entire read data transaction. I believe most MCUs won't have a problem with this. Typically, you have to set the GPIO pin to control /CS in your code. Make sure your peripheral is not operating in a "frame sync" mode that toggles /CS after every SPI byte sent.
  
  
• With the MSP430, you'll typically wait until the SPI read/write operations have occurred and you check by reading a flag. You may need to check if you are reading the correct flag so that you are not waiting on the wrong operation to finish before you can send your next byte.
  
  
• I don't think this is an requirement with the ADS1259, but other ADC's require that you read all data bytes, including any CRC or checksum bytes that follow the data. Terminating data read out early can cause problems with those kinds of devices. If you are not reading the checksum byte, I recommend turning it off in the register settings - it should also be off by default with the ADS1259; therefore, I doubt this would be an issue.

Best Regards,
Chris

I just saw your edit... have you tried reading back the register settings?

The data rates assume you are applying the nominal 7.3728 MHz clock. With an 8 MHz clock your actual data rates will be slightly faster.

Regards,
Chris

Hi Chris,

I have tried reading back from the registers of the PGA280 and the ADS1259 perfectly fine using the read and write commands. I also have monitored the data ready lines to make sure that the they are being toggled correctly when I send it the start command. The issue I currently have is that when I try to assign a variable to my read function, it says that it is zero. I can see that when I send 0x00 over SPI, the data is being sent over the DOUT but I am unable to retrieve it. My intuition is that by the time I do the read, the time window to do so is already over. 

EDIT:

Here are some pictures I took of what I have so far:
http://imgur.com/80gtPnQ (the top is my clock signal and the bottom is my DOUT) Note: I am also in the read data continuous mode via sending 0x10 so these are the 3 0x00 bytes I am sending in order to get the data back. 

http://imgur.com/MAs9ltS (the top is my clock signal and the bottom is my ADC chip select line when writing the 3 0x00 bytes)

http://imgur.com/PvnsCvA (the top is my clock signal and the bottom is my data ready line) The last clock signal there is me sending a read stop continuous 0x11 since I only wanted a single sample.

Here is also a snippet of my code that I am using to send the 3 0x00 bytes and attempting to retrieve the data back:

selectADS_CS(1,0); //set the specific chip select to 0
SpiChnPutC(1,0x00); // Write 0x00 (Output to keep ADC transmitting)
dummy1 = SpiChnGetC(1); 
SpiChnPutC(1,0x00); // Write 0x00 (Output to keep ADC transmitting)
dummy2 = SpiChnGetC(1);
SpiChnPutC(1,0x00); // Write 0x00 (Output to keep ADC transmitting)
dummy3 = SpiChnGetC(1);
selectADS_CS(1,1); //set the specific chip select to 1

Something interesting I noticed is that when I try introducing 2-3 variables, the function call gets locked up and CS never goes high again.  However, when I just set 1 variable, the variable passes through but every time I try to read the dummy byte, it is always 0.

Regards,

Edmond

Hi Edmond,

What is the datatype of "dummy#" and what are you returning from the SpiChnPutC() function (and what is its datatype)?

The scope shots look pretty good... I do see some small crosstalk, probably a result of capacitive coupling of the digital traces. Just in case the scope is not showing the full bandwidth of the crosstalk, you might try slow down the digital edges. You can do this by adding series resistance to you digital traces or additional capacitance near the digital input pins. You might have to cut some traces and scrap off some soldermask to make space for these components.

OR

If your digital traces are connected via jumper wire to the ADS1259EVM, then make sure the connections are short and twist your wires around the ground wire to reduce the current loop area and lower the inductance of the jumper wires. This has solved many communication problems I've run into.

Unfortunately, I have no experience programming with the PIC32MX795F512L. If my suggestions above don't help, you might try searching for help on the Microchip website.

Regards,
Chris