Post Go back to editing

problem with the continous mode of AD7738

I cannot read the status and data registers in the continuous mode of AD7738. The single mode is OK. On page 23 of the data sheet in fig.12 the value of the mode register for the continuous mode is wrong. Couldn't find errata for the AD7738.Any idea?

  • Hi Peter,

    I've checked the scope shots you've sent.

    I noticed that when you pulled /CS pin low, your DIN pin was kept high for some time before it went low. The duration when you kept DIN pin high was over 100ns. This may mean that you were actually exiting continuous read mode before you even started reading. As indicated in the continuous read mode section in page 23 of the datasheet, The AD7738 will stay in continuous read mode as long as the DIN pin is low while /CS pin is low, and pulling DIN pin high for 100ns would exit continuous read mode.

    With this, I would suggest that DIN should be kept low during reading.

    Can you try doing this and update me of the results?

    Thanks and Best Regards,

    Chris

  • Hi Peter,

    Regarding on where I've read that the DIN should stay low while the /CS  stays low, it is found in the Continuous Read(Continuous Conversion) Mode Section on page 23 of the datasheet.

    Hope this helps.

    Thanks and best regards,

    Chris

  • Thanks Chris,

    I came to the same conclusion myself. What we’ll do next will be to read all the 24 bits with one chip select low and DIN low all the time while the CS is low.

    I’ll let you know the result. But where have you read that the DIN should stay low while the CS stays low?  All the data sheet says is (page21 of the data sheet of AD7738)”

    “The DIN pin should not be high while reading 24-bit data in Dump Mode”

    And we know that Continuous mode is automatically the Dump Mode.

    Even more: in the timing on this same page they show the CS pulsing and with all the SPI bus requirement it’s natural that the CS should be a little bit larger than the

    Time needed to read or write the data because that’s why they call it a CS- to enable the component to listen to the data traffic. Anyway because I came to the same conclusion myself and you state the

    Same reason as the possible cause for my problem, I will test with DIN low while the CS is low.

    This time it must work…

    Thanks and talk to you soon again.

    BR,

    Peter

  • Thanks Chris,

    Find attached 2 more scope pictures. You can see that the DIN is low when the CS is low after the chip enters the continuous mode.

    Basically at this point the DIN is a mirror image of the CS. There is no delay between these 2 signals – they are exactly the same in time (only after the Ready pin goes low indicating that

    The conversion is complete and I can read the first status and data register content).

    Still no effect- no data come out from the Dout output of the chip.

    For now I’ll stop my investigation of this issue.

    Of course if you come up with some suggestion what could be done to have the continuous mode functioning you can e-mail me but I tried everything. Just let me know if by any chance you

    Know another customer using successfully this continuous mode. I really need it because I have to read the data from many, many channels…

    And thanks again for your help. I won’t bother you with questions any more.

    Have a nice day!

    Kindest Regards,

    Peter

    attachments.zip
  • Hi Peter,

    As I was looking at your scope shots, I saw that the DIN went high for more than 100ns and that might have caused your device to exit the continuous read mode.  I've pointed that in the attached pic.

    The part enters continuous read mode after the last rising edge of SCLK when writing 48h. Bringing DIN high after this write would exit the continuous read mode. DIN must be kept low after the write of 48h to keep it in continuous read mode.

    I have attached scope shots of a working set-up using one of our Evaluation Boards for AD7738.

    As you can see, DIN is always kept low so that a write to the part can be avoided while it is reading. Also, this would keep it into continuous read mode as the DIN is never brought high for 100ns.

    I hope this helps.

    Thanks and Best Regards,

    Chris

    attachments.zip
  • Hi Peter,

    It's a pleasure to be of help. We are now working on revising the datasheet.

    Anyways, I hope this would help. Please update me of the results.

    Thanks and Best Regards,

    Chris

  • Thanks again Chris!

    Well, the picture you sent me should be in the data sheet of the part instead of Figure 22 on page 25 of the AD7739 data sheet.

    (I test the AD7739). It shows clearly the DIN is in tri-state after the 48 defining the continuous mode. In this same picture DIN goes low with the CS only

    Later on when the CPU reads the status bit (see the attached PDF). Plus the DIN pulses high and low in sync with the pulsing CS while reading the status register and the data reg.

    I’ll test with this new setting and let you know the outcome.

    It was great that you spent on my problem so much time and even tested this elusive mode.

    Please, please, let the technical writers from AD know that the data sheet is not perfect.

    Probably more people will spend additional time in trying to use the chip in such a mode.

    Just my 2 cents.

    Have a nice day!!

    BR,

    Peter

    continuousmodedatasheetAD7739.pdf
  • Hi Peter,

    You're welcome.

    I'm glad that it's already working.

    Thank you and Best Regards,

    Chris

  • Hi Chris,

    I inverted the DIN in order to have this signal low in normal state and I got the continuous mode

    Without changing anything else. I have to work on the details but clearly the chip responds now.

    You helped me and I highly appreciate your help.

    Thank you!!

    Peter

  • Sorry, that e-mail was not for me.

         I wanted to know if does an another ADC exist like AD7783, but a

    little bit faster and with less "bit". Vor example 14-16bit  but also

    with current source for RTD Temp.measuring

         Thanks if you also want to answer me

         With best regards

         Bogdan Sosic