• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 190 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.

THS3001: THS3001 gain and phase margin at low gain.

steve smith1
Intellectual 290 points
Part Number: THS3001

Hi looking at the THS3001 datasheet it says for a gain of +2, -1 Rf = 680 Ohms so Rin=340 Ohms. When I simulate this it gives a phase margin of only 7.5 degrees and a gain margin of 1.4dB so barely stable. When I increase Rf to the recommended 1k value I get a phase margin of 24 degrees and a gain margin of around 5dB which is also marginal. On the bench I know Rf =500 Ohms and Rin=200 ohms causes instability which agrees with simulation. My question is the simulation model giving incorrect results and datasheet recommendations correct, or is the simulation model correct and the datasheet recommendations wrong ? If you have corrected versions of either please let me know ! (Used Michael's method of voltage controlled voltage source between summing node and in- to verify). I need a definitive answer because my actual circuit currently uses Rf=750 and Rin=300 which per datasheet should work fine but in simulation gives like 1degree phase margin and as we intend to build boards in quantity don't want parasitic oscillations so need the correct information. Thanks.

  • 0
    TI__Genius 12055 points

    Hi Steve,

    We always recommend applying information found in the datasheet, so the values suggested in the datasheet are your best bet. We create models to try to simulate what is in the datasheet not the other way around, and there are instances where models show promising results but when implemented the real-life device behaves differently. The latest model for this device is the one currently found on TI.com, however we will always recommend using the values highlighted in the datasheet, as with any model, not every performance and application scenario is captured.

    Best regards,

    Ignacio

  • 0
    Intellectual 290 points

    Thanks Ignatio,

    What I found concerning was that using the recommended values shown on the datasheet the model didn't return a reasonable gain/phase margin. Do you know what the gain and phase margin is for the recommended values on the datasheet ? Also how accurate are the datasheets ? Are they done by the component designers or technical writers ? I ask because I've come across a number of errors on other manufacturers datasheets !

    Finally I understand TI uses behavioral models derived from the datasheet  so there are some inherent limitations even if the datasheet is 100% accurate, but can you possibly let me know which  op amp  parameters are generally reasonably well modelled and so dependable in simulation and which are not and can only be measured on the bench with an eval board ?

    Thanks,

    Steve

  • 0 in reply to steve smith1
    TI__Genius 12055 points

    Hi Steve,

    The model for this device is from 1999 and is designed using a different modeling template. We do not have much information on this model therefore it is hard to say what the model is good at recreating and what it is not. The device itself is also relatively old and the designers for this device are not around, however the recommended values did come from the designers themselves and was based off of simulations and were chosen to provide the optimal configuration for this device. Therefore, as with any implementation of our devices it is always a good idea to prototype on a physical board as even the best models do not simulate the device under real world conditions. This device does have an EVM available that can be found in the product home page.

    As for your second question, the older models, like the one for this device, it is hard to really answer that question considering the very little information available. However, for the latest models we do highlight what the model was designed to simulate in the netlist of the device. I will add that although we do try to recreate AC performance of the device and we highlight the model's ability to do so, this is a simulation, and many other factors affect the real-life performance of any device. The DC parameters for the models are reasonably well modelled. Other performance specs like overload recovery are very difficult to model accurately as the device's transistors are in a unique state so it is expected that the model sometimes does not model these types of performance specs all that accurately. However, like was mentioned you will not get a clearer answer to how the device will perform like a real-world circuit.

    Best Regards,

    Ignacio