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.

ds900c032 feeding dac7734 and 7731 slow turn off time

Other Parts Discussed in Thread: DAC7734, DAC7731, DS90C032

I am using a line driver(ds900c032) that feeds the dac7734 and I am having trouble with the off time, or turning parts off then on and collecting data.  The power sequence on is -15v, +5v then the +15v.  Reverse in shut down.  The driver is getting warm during shut down but is not failing.  I am also using the dac7731 with the same buffer.  We did have a sequence issue on turn on originally when we where turning the +/- 15v then +5 and we were drawing to much current through the buffer.  we have to wait 2-3 minutes before we can cycle power on and collect data.  I would like that to be much faster.  I have not been able to collect data on the system yet but I thought I would post the question since I have done a data sheet review.

Any thoughts?

Thanks

Bill

  • Hello Bill,

    I apologize about the delays in responding to this thread. For some reason I did not receive a notification about this post. To expedite things as we go forward, we can exchange contact information and have some phone calls if you'd like. These are old parts but we can schedule accordingly to have at least one designer that was around in this era on the call as well.

    If you don't mind sharing a schematic, that would help me map out the design and understand how the supplies may be impacting the DACs and how the DACs might be impacting surrounding components.

    user4384251 said:
    The power sequence on is -15v, +5v then the +15v.  Reverse in shut down.  The driver is getting warm during shut down but is not failing

    I can't seem to find a datasheet for the DS900C032 to make a guess at how things are connected between that device and the DAC7734, or any supply sequencing implications concerning the interaction of these two devices, the schematic will help me understand how the DAC might be impacting the driver.

    Since you say the DS900C032 is "feeding" the DAC7734 I would guess that it's driving the reference inputs. I assume +15V is connected to VCC, -15V to VSS, and +5V to VDD, where do the VREFH and VREFL signals fall in these start-up and power-down sequences? As the datasheet mentions, if the reference is still powered while the other analog supplies have been removed it is possible for internal ESD cells to conduct which would explain the higher current consumption you see during power down.

    Are there any other symptoms or issues observed with the DAC7734 section of the design or is heat the only problem?

    user4384251 said:
    I am also using the dac7731 with the same buffer.  We did have a sequence issue on turn on originally when we where turning the +/- 15v then +5 and we were drawing to much current through the buffer.

    As with the previous case, it would be good to know where the external reference (assuming that's where the DS900C032 falls in this) falls in the power supply sequence with respect to the other supplies. Though it isn't explicitly stated in the datasheet, the absolute maximum ratings suggest that there is an ESD cell conduction path from VREF to VCC that would explain heat in the buffer if VCC is removed before the reference voltage. I am also curious if heat is the only problem you've observed concerning the DAC7731 section, or if there are any other symptoms.

    One other item I will share with you, just in case there is concern about other symptoms popping up, is that these older devices implemented different power-on reset circuitry than most of our modern devices. While the modern parts are sort of comparator based, the old products used a capacitor based POR. This means that it could require a long time for the internal POR capacitor to discharge in order to guarantee a POR event when the supplies come back up. Missing a POR for one of these parts isn't all that bad, though, because they were laser trimmed rather than "e-trimmed" through some internal OTP memory so there is no chance trim coefficients or anything along those lines will be corrupted. Worst case I can forecast is that when power returns the registers may contain old data or random data. If that's the case both of these devices have hard-ware reset pins that should allow you to reset the registers to default settings and not require a long period of time for supplies to discharge.

    The only issue I would be worried over here is the heating of the reference drivers based on the power supply sequencing including the references. Based on what you've shared so far, I'd guess that is the problem...

  • Hi Kevin

    Thanks for the help. I have attached 5 pages of the schematic that pertain to the Drivers and the DAC’s. I left the page numbers on so you can follow the signal path. Each pin has a page number and page location to where the signal was routed from or where the signal is routed to. The +/- 15v and +5v digital rails are supplied and anything labeled with an _a is a derived voltage. We do sequence the all of the supplies on turn on and turn off. I feel like we should just turn off all the supplies at the same instead of one every 10us. I also attached the data sheet for the line receiver. I do believe we are creating an esd “like” issue as described in the data sheets.

    Comments appreciated.

    Billds90c032.pdfdac_sch.pdf

  • every 100ms.... sorry
  • Bill,

    Unfortunately power supply sequencing can be one of "those" problems on the board and system level since you're integrating so many different things with so many different requirements.

    In the interest of trying to directly address your goal of reducing time in power cycle events, what happens when you do just remove all of the supplies at the same time? Which of the DS90C032s get hot? All of them? Just the drivers associated with one of the DACs? Are there any symptoms that the DACs display as well? If you can share some oscilloscope captures, maybe of the DAC supplies discharging along with one of the input signals that the DS90C032 is driving to the DAC we can better understand what the root cause of the issue is and optimize the sequence.

  • WE took some data on the supply rails and I have attached the plots.  It looks like the time constant of the rails is to slow for the turn off sequence.  Do you think this will damage the devices?  We have not changed the fall time sequence yet.

    Thanks

    Bill O'CRise_Fall_Time.pdf

  • Bill,


    Does this question relate to any particular device in your system? Often if there are problems with power sequencing, we'll mention it in the datasheet. However, it would be helpful if you could state what particular device you're worried about. That way we could give comments based on the history of the device.


    Joseph Wu
  • I thought I was attaching the data to my previous threads?
  • ds900c032 feeding dac7734 and 7731 slow turn off time
  • ds900c032 feeding dac7734 and 7731 slow turn off time
    Prodigy 50 points
    user4384251
    user4384251

    I am using a line driver(ds900c032) that feeds the dac7734 and I am having trouble with the off time, or turning parts off then on and collecting data. The power sequence on is -15v, +5v then the +15v. Reverse in shut down. The driver is getting warm during shut down but is not failing. I am also using the dac7731 with the same buffer. We did have a sequence issue on turn on originally when we where turning the +/- 15v then +5 and we were drawing to much current through the buffer. we have to wait 2-3 minutes before we can cycle power on and collect data. I would like that to be much faster. I have not been able to collect data on the system yet but I thought I would post the question since I have done a data sheet review.

    Any thoughts?

    Thanks

    Bill
    DAC7731, DAC7734Edit tags
    Prodigy 50 points
    user4384251
    user4384251

    May 28, 2015 5:26 PM
    Like
    More
    Flag as spam/abuse

    All Responses
    Answers Only

    Mastermind 28660 points
    Kevin Duke
    Kevin Duke
    Jun 1, 2015 9:01 PM

    Hello Bill,

    I apologize about the delays in responding to this thread. For some reason I did not receive a notification about this post. To expedite things as we go forward, we can exchange contact information and have some phone calls if you'd like. These are old parts but we can schedule accordingly to have at least one designer that was around in this era on the call as well.

    If you don't mind sharing a schematic, that would help me map out the design and understand how the supplies may be impacting the DACs and how the DACs might be impacting surrounding components.

    user4384251
    The power sequence on is -15v, +5v then the +15v. Reverse in shut down. The driver is getting warm during shut down but is not failing

    I can't seem to find a datasheet for the DS900C032 to make a guess at how things are connected between that device and the DAC7734, or any supply sequencing implications concerning the interaction of these two devices, the schematic will help me understand how the DAC might be impacting the driver.

    Since you say the DS900C032 is "feeding" the DAC7734 I would guess that it's driving the reference inputs. I assume +15V is connected to VCC, -15V to VSS, and +5V to VDD, where do the VREFH and VREFL signals fall in these start-up and power-down sequences? As the datasheet mentions, if the reference is still powered while the other analog supplies have been removed it is possible for internal ESD cells to conduct which would explain the higher current consumption you see during power down.

    Are there any other symptoms or issues observed with the DAC7734 section of the design or is heat the only problem?

    user4384251
    I am also using the dac7731 with the same buffer. We did have a sequence issue on turn on originally when we where turning the +/- 15v then +5 and we were drawing to much current through the buffer.

    As with the previous case, it would be good to know where the external reference (assuming that's where the DS900C032 falls in this) falls in the power supply sequence with respect to the other supplies. Though it isn't explicitly stated in the datasheet, the absolute maximum ratings suggest that there is an ESD cell conduction path from VREF to VCC that would explain heat in the buffer if VCC is removed before the reference voltage. I am also curious if heat is the only problem you've observed concerning the DAC7731 section, or if there are any other symptoms.

    One other item I will share with you, just in case there is concern about other symptoms popping up, is that these older devices implemented different power-on reset circuitry than most of our modern devices. While the modern parts are sort of comparator based, the old products used a capacitor based POR. This means that it could require a long time for the internal POR capacitor to discharge in order to guarantee a POR event when the supplies come back up. Missing a POR for one of these parts isn't all that bad, though, because they were laser trimmed rather than "e-trimmed" through some internal OTP memory so there is no chance trim coefficients or anything along those lines will be corrupted. Worst case I can forecast is that when power returns the registers may contain old data or random data. If that's the case both of these devices have hard-ware reset pins that should allow you to reset the registers to default settings and not require a long period of time for supplies to discharge.

    The only issue I would be worried over here is the heating of the reference drivers based on the power supply sequencing including the references. Based on what you've shared so far, I'd guess that is the problem...
    Kevin Duke
    Analog Application Engineer
    DAC7731, DAC7734
    Like
    More
    Flag as spam/abuse
    Prodigy 50 points
    user4384251
    user4384251
    Jun 3, 2015 4:33 PM

    In reply to Kevin Duke:

    Hi Kevin

    Thanks for the help. I have attached 5 pages of the schematic that pertain to the Drivers and the DAC’s. I left the page numbers on so you can follow the signal path. Each pin has a page number and page location to where the signal was routed from or where the signal is routed to. The +/- 15v and +5v digital rails are supplied and anything labeled with an _a is a derived voltage. We do sequence the all of the supplies on turn on and turn off. I feel like we should just turn off all the supplies at the same instead of one every 10us. I also attached the data sheet for the line receiver. I do believe we are creating an esd “like” issue as described in the data sheets.

    Comments appreciated.

    Billds90c032.pdfdac_sch.pdf
    DS90C032, DAC7734Edit tags
    Like
    More
    Flag as spam/abuse
    Prodigy 50 points
    user4384251
    user4384251
    Jun 3, 2015 4:38 PM

    In reply to user4384251:
    every 100ms.... sorry
    DAC7734Edit tags
    Like
    More
    Flag as spam/abuse
    Mastermind 28660 points
    Kevin Duke
    Kevin Duke
    Jun 3, 2015 6:05 PM

    In reply to user4384251:

    Bill,

    Unfortunately power supply sequencing can be one of "those" problems on the board and system level since you're integrating so many different things with so many different requirements.

    In the interest of trying to directly address your goal of reducing time in power cycle events, what happens when you do just remove all of the supplies at the same time? Which of the DS90C032s get hot? All of them? Just the drivers associated with one of the DACs? Are there any symptoms that the DACs display as well? If you can share some oscilloscope captures, maybe of the DAC supplies discharging along with one of the input signals that the DS90C032 is driving to the DAC we can better understand what the root cause of the issue is and optimize the sequence.
    Kevin Duke
    Analog Application Engineer
  • This is a follow-on to this thread for the DAC7734/7731.

    e2e.ti.com/.../426699

    We'll close this thread and keep replies to the original.


    Joseph Wu
  • 5228.dac_sch.pdfI have attached the schematics data sheets and voltage rail turn on and turn off measurements.

    My issue is when we turn off the unit and turn it on again fast (less than 1 min) the DAC is not working correctly or its output is not responding.  When we wait a couple of minutes and turn the board on it responds the way we expect it. 

    The supplies are turned on in one hundred ms steps and the same in turn off.  -15v then +5v then +15v is the on sequence. .  We turn it off in the exact sequence which might be the issue. 

    My question is if the turn off voltage sequence causing the DAC or the receiver "lock up" issue on a fast turn on?  Also are we damaging the parts  with this sequence?

    u7(pg12)  feeds u32(pg19) is an example of what we are concerned with.

    Thanks

    Bill O'C4401.ds90c032.pdf

  • Hello Bill,

    For your convenience, I have merged your posts regarding both DS90C032 and DAC7731 into one thread.
  • Bill,

    I wouldn't anticipate that any of the devices are being damaged through this sequence unless the reference voltage is still powered after the analog rails have dropped. In that case there's some potential for stress on the ESD cells and the analog supply rails themselves. As far as the analog rails and digital rail that are shown in the plots you shared go, everything looks fine. Do any of the rail consume excessive current in this process? What do the pins that are connected to the line driver and the reference input look like during these sequences?