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.

UCC1805 Error Amplifier Question

Other Parts Discussed in Thread: UCC1805, UCC2805

I have a question regarding the UCC1805 error amplifier. I am trying to run an analysis on a design which utilizes the UCC1805-EP. Currently we are having what I believe to be compensation problems. In an averaged model the compensator works as expected, but when plugged into a transient model using TI's pspice model the supply has a large constant error.

Digging into it I made a model to test the error amplifier in the spice model. I constructed a simple inverting amplifier using it and found the amplifier output in the model to be 180 out of phase relative to an ideal op-amp stimulated by the same source. In other words when configured as an inverting amplifier it acts like a non-inverting amplifier. It does not act anything like an op amp as the data sheet says. Is this representative of the real device? If so, how do you compensate it? 

  • Chris,

    The UCC1805 is one of TI's high-reliability products, so this would be better posted to the hi-rel E2E forum here:

    e2e.ti.com/.../935

    Some one on the High-rel team should be able to help you with your questions.

    Thanks,
    Bernard
  • Bernard,
    With all due respect sir, I get the feeling your trying to blow me off. The specifications for the UCC1805 and UCC2805 are the same with the exception of the temperature rating and they share the same spice model. This is a DC:DC converter application and a DC:DC issue, it is nothing specific to high temperature operation according to the data sheet.
    When I get my hands on a part I can characterize it myself and answer this question myself but it would be nice to have information now and it is a TI product. If the model is valid it is probably a drive issue because with really light loads I'm not seeing the characteristic change. The problem is that I am seeing this change at the application load. The application requires a type 2 compensator and the values are as follows.
    The RC branch has values of 82nF & 20k
    The C branch has a capacitance of 68pF
    Oscillator RC => 10.7k & 560pF
    Fsw = 86kHz
    When comparing the UCCX805 error amp model to an ideal op amp I am seeing a 180degree phase shift in the current with the compensation network values above. With a very light load the phase shift disappears.
    If the model is valid we need to either redesign or select a different control IC, if the model is invalid then I need to get some chips to characterize and build a model. Is the error amplifier model valid for the UCC2805?
    Regards,
    Chris
  • Chris,

    I was justing pointing out that the Hi-Rel forum was the correct place to post the question on UCC180x, where you would likely receive a faster response.

    Since the UCC1805 is not in our portfolio, it would not necessarily get as prompt a reponse in this forum.


    That said, you are correct, the internal architecture of UCC1805/2805/3805 is the same, just different temp rating and electrical specs.


    Can you answer a few questions and give more details to help figure out the issue:

    - you said that when you simulate using an averaged model, all is ok - is the model taken from the UCCx805 tools page, if so, which model are you using?

    - similarly, which Spice transient model are you using? i can see 3 models availble on the website.

    - when you run the transient model, you get a large constant error - where, at the power stage output voltage?

    - what does your power stage look like, can you post a sch? Is it a flyback, buck or other topology?

    - for the compensation values and design, this will depend on the power stage poles/zeroes and the target loop bandwidth etc.

    - while the internal error-amp is a true op-amp, as mentioned on the datasheet page 12, it is severley limited in it's source and sink current capability. It can source up to only 200 uA (worst case min figure), to allow the COMP pin to be externally pulled low, and can sink up to 300 uA (again worst-case figure), thoguh the model probably assumes typical numbers of 500 uA source and maybe 2 mA sink. Can you post the sch you are using to simulate the device, and to simulate the opamp only?


    I can ask the modellign team to comment also, if you clrify which model you are sungin and add the schematics.

    Thanks,
    Bernard