• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 47 subscribers
  • Views 368 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PCM56P

Paddy Burns
Prodigy 30 points
Other Parts Discussed in Thread: PCM56

Recently purchased 2 TI PCM56P devices from farnell. The devices are not behaving the same as the original BB (Burr Brown) device. For a fixed input the TI device is producing 2.247V (loaded 5.3k and unloaded O/C) the original BB device produces 2.375V (loaded 5.3k and unloaded o/c) The voltage rails are -5.03V and +4.997V.

The output from the TI PCM56P is linear with no discontinuities just seems to be operating with slightly lower gain. There is no circuitry connected to pins 14 or 15 ie they or o/c.

In the same circuit the output of BB original chips provides 2.373V and the two TI chips provide 2.247V

Is there anything that could explain this.

  • 0
    TI__Guru**** 189401 points

    Hi, Paddy,

    The guys who may provide support for PCM56 (it is NRND, so...) hang out in Precision Data Converter e2e, so let me move your thread over there.-

    -d2

  • 0 in reply to Don Dapkus
    TI__Mastermind 46290 points

    Paddy,

    To be honest I'm drawing a blank on this part, it's pretty old and NRND as Don pointed out, so tracking down any information about any process changes or any characterization changes is going to exceptionally difficult.

    I'll try to dig around, but I can't make any promises that I'll discover what has changed about the design (if anything).

  • 0 in reply to Kevin Duke
    Prodigy 30 points

    Kevin,

    thanks for looking into it.  It is as if the internal reference cct has changed. Confusing bit is happening offload ie o/c. It may also be that the trim/msb adjustment would affect this although the original data sheets did not require these to be connected. I am trying to source some devices from different manufacturer/supplier and see what way they behave.

  • 0 in reply to Paddy Burns
    TI__Mastermind 46290 points

    Paddy,

    I can't find anything that has changed about the device or it's final test plan. That said, the part is exceptionally old and has probably gone to a new process one or more times over it's life. The reference circuit itself is probably trimmed internally and over the process changes or fab changes the internal trim process may have been updated. Since the part does have external trim features it may have been decided that a revision to the internal trim process was not necessary and can be overcome by external trimming. Overall, it's very very difficult to say due to the age of the device.

    Can you tell us more about your use of the device? With it being NRND we strongly suggest that, if this is a new design, you choose another device and if you're updating an old design it may be best to go ahead and move on to a more recent device so you don't have to deal with changing the part once it hits end of life completely.

  • 0 in reply to Kevin Duke
    Prodigy 30 points

    Kevin,

    It is from a design circa 1993, so the activities are maintenance and repair.

    It is used to generate simulated transducer voltage signals less than 60 circuit boards (4 channels/ board) in total produced. We have only had 1 previous device failure prior to this occasion and fitment of the newly purchased device produced lower than expected output in uncalibrated state. We can calibrate the error out just the customer will want to know why there is such a large calibration adjustment required. The device serial input is opto coupled and output buffered so it is not subjected to spurious loads and transients.

    Thank you for your efforts as we are not new/re-design and failure is rare i will just report problem and buy some to store for future use if required.

    Paddy