• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 49 subscribers
  • Views 934 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.

LM258 non-inverting input randomly going to Vs

Alan Stockdale
Prodigy 30 points
Other Parts Discussed in Thread: LM258, TL072

Hi,

I am using an LM258 in a simple unity gain configuration. Vs is 12V.   V- is tied to Vout, and a voltage divider signal is input to V+.  The signal itself is from a probe in a pull up configuration - pull up resistance is 1Mohm to a 1V supply, and when a contact is made to ground on the low side, the path is also about 1Mohm (through a low conductance fluid).  Ground is common.

Most of the time this circuit works fine, but on occasion, the V+ pin will go to about Vs, which is 12V.  This then produces and output of about Vs.  This has happened on more than one chip.  If I disconnect the signal input, V- stays high.  If I power cycle the circuit, things go back to normal for a time, and then at some seemingly random point in the future, the same thing will happen.

Only one side of the chip is used.  Do I need to ground the pins on the unused op-amp?  I do not have decoupling caps, as we are just using this a crude threshold trigger, so we are not overly concerned with noise.  Could some sort of power spike have damaged the circuit to cause this intermittent problem?  Could it be a heat issue (circuit is in an enclosed space that gets warm (30-40C?).   I wasn't sure if this was a known issue, or if some of the above precautions would make it go away.   Thanks in advance for any responses.

  • 0
    TI__Guru*** 139956 points
    Alan,

    The LM258 has a input bias current that flows out of the device. If the input is not connected then it will float high by itself.
    Tell me more about the input divider. What values of resistance are used.
  • 0 in reply to Ron Michallick
    Prodigy 30 points

    Ron,

    Thanks for the response.  I thought the divider was 1M pull-up, to a 1M resistive pathway - this is for a conductive point level sensor used to detect the fluid level of a cell culture - so the second 1M is through a fluid.  The cell culture changes conductivity as the culture grows, so 1M is an estimate.  But I have since discovered that there is what I think is a static charge build up on the probe itself, thus skewing the DMM measurements I was taking.  So if the lower side of the divider was really 1M is not clear.

    On the LM258 issue - I now believe that I was getting various symptoms of latch-up on the IC.  There are some motors/solenoids that are on the same power supply, and any spikes from these I think pushed the IC into latch-up.  One symptom i think was the Vs showing up on the non-inverting input.  Another symptom that I did not mention was that we had a couple of these burn up - looks like Vs got shorted to ground.  I have been looking for good external component configurations to deal with latch-up, so if you have any resources you could link to, that would be great.  

    This is all somewhat moot now though.  We are trying to avoid electrolysis in the fluid, so we are moving to an off the shelf conductive level probe transmitter that uses and AC signal.  So I will decommission the LM258 circuit soon.

  • 0 in reply to Alan Stockdale
    TI__Guru*** 139956 points
    Alan,

    With 1M input resistance the input bias current will be a factor. I have never heard of a LM258 latching up before. It wouldn't be easy to make that happen. For less input bias current, TL342 for low voltage and TL072 for high voltage are good candidates.
  • 0 in reply to Ron Michallick
    Prodigy 30 points

    Ron,

    Thanks for the suggestions.  The only thing about these LM258s- I have had at least 4 of these just burn up - where is the current coming from if not latch-up?  1V driver on neg input - very high impedance - I can't imagine that kind of current could be produced other than from the 12V Vs to ground.

    Thanks,

  • 0 in reply to Alan Stockdale
    TI__Guru*** 139956 points
    Alan,

    Do you know which pins are shorted?