• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 658 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.

TIPD107: Systematic spikes using either OPA333 or TLV333

Taringana Guranungo
Prodigy 50 points
Part Number: TIPD107
Other Parts Discussed in Thread: TLV333, OPA333, OPA391, OPA191

Hi

My circuit uses the TIPD107 reference design to generate 0-5uA current from 0-5V. The modifications I have are inserting a buffer at the same node as my load so that I can track the change in voltage using an ADC (real load will be a variable resistor) and I am sampling every 15secs. My simulation is giving me expected results but the issue is only coming up when I test the real hardware.

My initial setup for validating the system uses a normal resistor and I am observing some voltage spikes in the readings. When I see the spikes, there is a voltage decrease or increase of approximately 5-10% depending on the board. The spikes appear to be happening at random times and do disappear based on the input voltage for some of my boards. The magnitude of the spikes is very consistent within a given board and are in the same direction for that given board. My input to the U1 OPA333(have mostly been using the TLV333 op-amp) is stable when the spikes happen showing that the issue is not coming from a noisy input. The spikes are first observed at the Vout_opa node before they travel across the feedback loop and to the buffer op-amp. I am using resistors with 1% tolerances so the noise can't be coming from them.

Has anyone observed similar spikes/noise in their system or know where the source of the issue is?

  • 0
    TI__Guru 62700 points

    I see that you added OPA333 buffer to the TIPD reference design and use 100k Rload instead of 100ohm (1000x higher Iout gain) - OPA333 and TLV333 are chopper amplifiers with internal auto-calibration circuitry that may require matching of the input impedances between the input terminals to minimize commutation of the IB spikes across the input resistor.  For this reason, you may need to add 50k feedback resistor (100k||100k) and 100nF feedback caps - see below - or use a linear op amp like OPA391.

  • 0 in reply to Marek Lis
    Prodigy 50 points

    Hi Marek,

    Thank you for the feedback. I will try both pathways and get back to you.

    Since my load will be varying in the final application(few KOhms to a couple MOhms), I will lean more towards the use of a linear op amp. Luckily, the OPA191IDR is a drop in replacement for the TLV333IDR I am using.

    -Taringana

  • 0 in reply to Taringana Guranungo
    TI__Guru 62700 points

    I recommended a similar to TLV333, low supply voltage OPA391 while OPA191 is a high supply voltage up to 36V.  Nevertheless, they both have standard pinout and should work on 5V supply voltage you use in your application.

  • 0 in reply to Marek Lis
    Prodigy 50 points

    Replacing only the buffer amplifier did not solve the issue nor did introducing the feedback resistor and caps manage to eliminate the spike. I went on to also replace the U1 OPA333 with OPA191 and so far I haven't observed the spikes(still need to collect more data on a bigger sample size).  Would the case be maybe U1 OPA333 is also contributing to the issue or there are other places where this issue might be coming from I should consider?

    I will replace the OPA191 with OPA391 on my next round of PCB fabrication.

  • +1 in reply to Taringana Guranungo
    TI__Guru 62700 points

    IB spikes of U1 OPA333 also get converted across 100k feedback resistors into voltage spike.  However, most likely it was not a problem in TIPD107 because it used 100ohm while you use 100k Rload - 1000 times higher amplification. Thus, the best option would be to use OPA391 (or OPA191) in both U1 and U3 - this should completely eliminate the problem. BTW, simulation model does NOT include IB spikes as this would result in a very long transient simulation time. 

  • 0 in reply to Marek Lis
    Prodigy 50 points

    Thank you Marek for your help. The spikes have now gone away.

  • 0 in reply to Taringana Guranungo
    TI__Guru 62700 points

    No problem.  Good to know this solved the issue.