Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 1221 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.

LM8261: Current source

BenoitBosquet
Prodigy 210 points
Part Number: LM8261
Other Parts Discussed in Thread: OPA890, OPA690

Hello,


I currently use LM8261 in an improved Howland current pump design. Current pump is supplied by +/-5V sources and is designed to output 100kHz sinusoidal 4mA pk-pk current. It is aimed to be injected in a human body (end application is for bioimpedance, load range equivalent to 300 to 1500Ohms). It works and seems to be stable without capacitors in the feedback networks. LM8261 offers rail to rail output with low current consumption. Output impedance is about 70kOhms from my simulations and practical measurements.

I went through several documents and papers stating that higher slew rate and higher GBP lead to higher output impedance. Claimed output impedances can be up to 1Meg. I’m not sure stability has been assessed in all studies. Stability topic has been covered in some posts on your forum (Thank you to Tim Green. He posted the following interesting document “To Design and Apps Current Sources_AFA.ppt”).

I went through AN1515 application note. It covers op amp selection and provides some references but GBP and SR of listed references are not really higher than the one I use.

I then used the TI op amp parametric table to find low power rail to rail wideband op amp. Unless I incorrectly used the filters, I didn’t find low power wide band rail to rail op amp in +/-5V power supply range. From my investigation I finally picked OPA890 and compromised on the rail to rail feature. In simulation current pump with OPA890 provides about 250kOhms. However it requires capacitor in the negative feedback to be stable. Then output impedance decreased to 47kOhms with a 5pF in the negative feedback.

If I incorrectly used the parametric table, do you have in your portfolio op amp similar to LM8261 but with slew rate and bandwidth similar to OPA890 or OPA690?  

What makes LM8261 stable without feedback capacitors? Is it related to its Unlimited Cap Load feature?

Please let me know if from your experience you have any advice, suggestions to design a high output impedance current source (output current from 100kHz to 150kHz).   

Thank you in advance for your remarks. 

  • 0
    TI__Guru 62915 points
    Hello,

    A circuit's closed loop-output impedance (Zout) is dependent on the open-loop output impedance (Zo), the circuit feedback factor (Beta), and the open-loop gain (Aol) performance of the circuit: Zout = Zo / (1+Aol*B). So lower values of Zo, or higher values of Aol*B result in lower closed-loop output impedances. On that note, the LM8261 output impedance should be much lower than 70kOhms.

    The stability concerns for the OPA890 and OPA690 without feedback capacitances are from to the op amp input capacitance interacting with the feedback resistances to form an additional pole in the loop-gain which reduces the circuit phase margin. Since the BW is so high for these products it's pretty easy for that pole to fall within the usable BW causing the stability issues. A feedback capacitor cures the stability issues by adding a zero to the loop-gain to cancel the pole. If you use high feedback capacitors with the LM8261 you'll eventually drive it unstable as well.

    About your circuit, could you share your schematic and any simulations you've run so we can see what you've built? Does the current source need to be bidirectional or is it always sinking or sourcing the 4mAppk?
  • 0 in reply to Collin Wells
    Prodigy 210 points

    Hello, 

    Thank you for your reply. I have attached corresponding Tina/spice simulation file (HCP_LM8261_Z1.TSC) of the bidirectional current source. Below is the picture I obtained from the simulation. 

    I have also attached the equivalent using OPA690 (HCP_OPA690_Z1.TSC). When C5 is removed then output impedance is much more higher about 1,07Meg. When C5 is added in the negative feedback loop for stability then output impedance is about 25k (with C5=10pF), 50k (with C5=5pF).

    Best regards, 

    Benoît, 

    HCP_LM8261_Z1.TSCHCP_OPA690_Z1.TSC

  • 0 in reply to BenoitBosquet
    Prodigy 210 points

    Output impedance of the Improved Howland is described on page 2 of this document:

    e2e.ti.com/.../Improved-Howland-Article.pdf