• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 501 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DDC112: Output is stuck at max value

Christer Bengtsson
Prodigy 100 points
Part Number: DDC112
Other Parts Discussed in Thread: DDC264

Hi,

I have designed a 4-channel picoammeter based on two DDC112s as a student project. The instrument behaved well enough until it was connected to a beamline slit system. One of the channels got stuck on max output, and has stayed there ever since. I suppose the corresponding input must have been damaged. It behaves as before in test mode. What kind of damage could make the output stay at max? 

/Christer B

  • 0
    TI__Mastermind 49146 points

    Hi Christer,

    Are you using external integrating cap mode? 

    What is the input current range?

    I am checking with design team for possible conditions that might lead to what you are seeing.

  • 0 in reply to Praveen Aroul
    Prodigy 100 points

    Hi,

    I only use range 7 (87,5 pF) and have no external capacitors.

    Input range 344,1 pC, integration time 8 ms => current range 43 nA.

    A bias current of 20 nA splits the range to ca. +22/-20 nA.

    /Christer B

  • 0 in reply to Christer Bengtsson
    TI__Expert 4590 points

    Hi Christer,

    In normal operation (no test mode), what do you have connected to the input? Can you try with the input floating but still in normal operation?

    Thanks!
    Edu

  • 0 in reply to Eduardo Bartolome
    Prodigy 100 points

    Hi,

    I assume floating means empty. With nothing connected the fourth channel still gives full scale output. There is still the on-board bias current of course. 

    Earlier I have made test measurements using a standard 1,5 V battery and a 5 Gohm resistor. Later I replaced the battery with a function generator. The amplitude was never above 4 V.

    But it was when the instrument was connected to a (negative ion) beamline slit system that something happened to the fourth channel.  

    /Christer B

  • 0 in reply to Christer Bengtsson
    TI__Expert 4590 points

    Hi,

    Yes, for floating we mean nothing connected to the input, at all. So the bias is actually something connected. You can start by measuring the voltage at the input of the DDC in normal mode (including the on board bias). That may give us a clue. Is your "bias" simply a resistor to some voltage? If so, it'll tell us how much current is going in.

    The other clue would be to disconnect everything (leave the pin of the device floating, i.e., still soldered to the board but nothing connected to it) and check what is happening. 

    Our theory is that one of the input back to back ESDs is damaged. What a negative ion beamline slit system does? I mean, how much current or voltage you get?

    Anyhow, overall, what is the reason to try to find what is damaged? Is it simply to try to prevent from happening again? Do you expect the instrument to work with the negative ion beamline slit system?

    Regards,
    Edu

  • 0 in reply to Eduardo Bartolome
    Prodigy 100 points

    Hi,

    The bias current is created as described by Figure 5 of Application Bulletin Creating a bipolar input range for the DDC112.

    Voltage 4,096 V, 200 Mohm resistor (1%) => 20 nA

    I only have a simple multimeter but I managed to measure this:

    Between input and AGND: 21 mV (is this normal?)

    Between "input" of 200 Mohm resistor and AGND: 4,09-4,10 V

    I will not try and disconnect the bias resistor at this point. 

    The slit system is used for aligning a negative ion beam in an accelerator. When the beam is not aligned, ions hit (four) metal plates and create a negative charge on these. The charge can be measured as a small current. The metal plates are floating and the surrounding vacuum system is grounded. The plates are supposed to be at virtual ground and very near actual ground. I was told the current would be no more than a few nanoampere but the instrument gave a full scale reading at 22 nA for two channels, including the damaged one. 

    I have done a lot of other measurements as I said before and the instrument worked good enough. I am asking so that I can learn from mistakes and also provide an explanation in my exam paper. I don't think the instrument will be used. Maybe there will be another student project building on my design.

    If one of the protection diodes is destryoed I assume the capacitor keeps its charge which will give a full scale output at every AD conversion. Is that a correct understanding?

    /Christer

  • 0 in reply to Christer Bengtsson
    Prodigy 100 points

    Did the measured values help?

    /Christer

  • 0 in reply to Christer Bengtsson
    TI__Expert 4590 points

    Hi Christer,

    Sorry for late reply but somehow didn't get a notification that you had replied...

    Anyhow, 21mV looks quite high. I don't have info for the DDC112 on input offset voltage vs current, but in something like the DDC264 (see figure 16) it doesn't go beyond couple of mV. My guess is that there may be some internal leakage current flowing into the input which saturates the integrator and as such the input virtual ground is not, anymore. But the direction of the flow to produce that voltage should be into the device, so, I somehow don't think it is one of the input diodes shorted to ground (the current would be on the opposite direction). There may be some internal structure in parallel with your external resistor. 

    If you disconnected the input or add more current to the input to see if you can get it out of saturation, we could guess more, but not sure we can help any further anyhow. Usually these things require failure analysis (opening the IC...)

    If the input diode was "destroyed" open, then you would not see anything, in general. If "destroyed" as a short (to ground) you may see saturation in either direction but I don't think you would see the 21mV. Regardless, the input cap (I guess you mean the parasitic?) would get discharged into the device (otherwise you can't see the output saturated). So, no, the cap can't keep the charge and just keep giving positive saturation. There are resets, etc. There needs to be an input current flowing into the device... or more drastic malfunction (like the ADC is simply stuck).

    Regards,
    Edu

  • 0 in reply to Eduardo Bartolome
    TI__Mastermind 49146 points

    Hi Christer,

    I have closed the post. Can you please post your response in a new thread if this thread is locked?

  • 0 in reply to Eduardo Bartolome
    Prodigy 100 points

    Thanks for the reply. I did not refer to input capacitance but the feedback capacitance. But I had it wrong. I can follow your answer. It seems unwanted current is getting into the device. I applied a larger periodical input signal but the output was still saturated. I have finished my project work now. Maybe I will try and build an improved instrument further on.

  • 0 in reply to Christer Bengtsson
    TI__Expert 4590 points

    Excellent. Good luck on your career! :)