• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 49 subscribers
  • Views 2282 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.

LF353 offset level changes from part to part

Pedro Casas
Prodigy 40 points
Other Parts Discussed in Thread: LF353, LMV932, OPA2377, LMV612, TLV2372, TLC272, LM358

I have get 600 LF353 from Mouser electronics. 

I have used the LF353 to make a simple voltaje buffer (unitary gain or follower amplifier) for interfacing a sensor with electronics. The low bias current and high imput impedance makes this part an ideal one for this application. The LF353N does exactly what is supposed to, but it is adding an offset to the sensor signal.

The real issue I am having is not the offset itself, but the fact that the offset level changes from one part to another. It does not matter if it is on a PCB or a protoboard, using the same other components and changing just one LF353N to another one, the DC offset may change as much as 300 mV! We are working with Vcc = 5V

We shall produce over 1000 units, but we can not really trust on its behaviour because of that variable DC offset. Would it be something I am missing on this basic design?

  • 0
    TI__Guru*** 139906 points

    Pedro,

    VIO at 25C can be as high as 10mV. The VIO gain is 1, so the maximum shift will be +/-10mV.

    The common mode input range is the likely problem. The input need to be at least 4V to ensure a valid input range.
    What voltage range does the sensor provide?

  • 0
    TI__Mastermind 49836 points

    Hello Pedro,

    5V supply is on the low side for this device. Most likely you are running into the common mode limit. The input common mode range does not extend to ground like newer devices, but the opposite, it's common mode extends to the V+ rail, which is typical of the older FET input devices.

    The legal input common mode range is from V+ to 3V above V- (see the Input Common Mode Voltage Range line in the table)...So the "legal" input range would be +3 to +5V, and anything from 0 to 3V is "illegal" and will see errors.

    When outside the common mode range, the offset will increase and the gain/BW will decrease. The offset shifts you are seeing is what I would expect for any input voltages below 3V. I assume you are seeing the shifts on the low side of the range.

    If you want the output to reflect the full 0-5V input from the sensor, then you need a "Rail to Rail" input device. Since it looks like you need a DIP package and FET/CMOS input, I would suggest the LMC6482AIN or TLV2372. If package is not an issue, the LMV612, LMV932 or OPA2377.

    Always test the circuit on the bench before ordering parts....surprises in production can be expensive!

    Regards,

  • 0 in reply to Paul Grohe
    Prodigy 40 points

    Thanks a lot Ronald, Paul,

    The sensor supplies a voltage range from 0 to 5V, but in this application it woul typically be from 0 to 1.5V top. On this configuration, the offset is around 1.4V on 60% of the devices, and around 1.7V on the remaining.

    I found this behaviour very strange, but honestly I really have not even considered the common mode range. I tested about 10 LF353 with good results before ordering, and convinced myself this was the part I was needing. 

    Fortunatelly just replacing the LF353 with another one (as the TLV2372) will do the trick... or is there something else I must consider?

    Once again thanks a lot and regards,

  • 0 in reply to Pedro Casas
    Prodigy 40 points

    Something I missed is the component price. Both the TLV2372 and LMC6482AIN are considerably more expensive than the LF353. Is there any other cheaper option I could consider?

  • 0 in reply to Pedro Casas
    TI__Mastermind 49836 points

    Hello Pedro,

    Those were the lowest cost devices for DIP R-R input devices. R-R devices are going to cost a little more.

    The LF353 is a 30 year old device. DIP packages are going the way of the dinosaur - so requiring DIP package is limiting your choices. No new devices are released in DIPs.

    Surface mount and SOT are the popular packages now.

    *IF* you do NOT require the full 5V input range, and can live with a 0 to 4V range, then you do not need a R-R input and have more devices to select from.

    If you can live with the 0-4V range, and still need DIP package, see the TLC272 - it should be even cheaper than the LF353.

    If you can get out of the DIP package and move to SOIC or MSOP, then you have many, many more choices.

    Regards,

  • 0 in reply to Paul Grohe
    TI__Guru*** 139906 points

    So input needs to be 0 to 1.5V
    What is VCC?

    Do you need a low input bias current? If not, we have several low cost "single supply" DIP op-amps.
    LM358 for example.

  • 0 in reply to Ron Michallick
    Prodigy 40 points

    I do need low imput bias current, and high input impedance. What I mean for VCC is the positive power supply. As you have said, I am working with a 5V single power supply.

    The input doesnt need to be from 0 to 5V, but from 0 to 4 or a little less will do it.

  • 0 in reply to Pedro Casas
    TI__Mastermind 49836 points

    Then your cheapest choice is the TLC272CP or LMC6032IN for CMOS input devices in DIP.

    Regards,