• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 39 subscribers
  • Views 521 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

THS7530 Model & TINA Reference Circuit

Neil Albaugh
Genius 5430 points
Other Parts Discussed in Thread: THS7530, TMP20

The THS7530 TINA Reference Circuit works as shown in the schematic on _/-2.5V supplies but there are serious problems with that macromodel as well as the reference circuit presentation:

1. The unlabeled pin is ? Who knows what it might be-- the symbol offers no clue. Looking into the model and at what is connected to this pin in the schematic reveals that it is probably the +gain control pin. The -gain control pin is apparently referenced to ground? Vs-? Who knows-- the text on the page offers no clue.

2. Has anyone looked at the voltage on the input pins with the DC Analysis test probe? Try it-- you get -4.8V on each input. Obviously something is wrong here. The model looks suspect.

3. Try connecting the THS7530 in the single supply circuits shown in the data sheet. The amplifier stops working-- instead of the output Vocm being pulled up to mid-supply as the data sheet and EVM manual says, the outputs go to a few hundred microvolts negative! Did anyone ever verify that this model actually worked?

4. The THS7530 macromodel offers no clue that it does not model the +Vcl or -Vcl functions. There is nothing regarding the modeling of other parameters that are vital for a high speed VGA-- noise, slew rate, settling time, etc.

Frankly, it is a shame that such a good device is being treated in this shabby manner. This amplifier deserves to have a reliable, accurate macromodel rather than this minimal-effort, half-a**ed offering.

I'm interested in hearing TI's excuse for not providing (again) a decent macromodel for their amplifiers.

  • 0
    Genius 5430 points

    So far TI has not responded to the previous post regarding the problems with the THS7530 Voltage-Controlled Amplifier. Actually I am not surprised that no one wants to take responsibility for this macromodel.

    IC manufacturers seem to take a cavalier attitude toward providing their customers with decent macromodels for their products. Why spend $200k to $300k to develop a new product and then handicap it by producing a poor macromodel or not providing any macromodel at all? I think the problem lies with marketing-- these folks are, for the most part,  engineers but they haven't actually designed anything for years. Any working design engineer knows how important it is to have the proper tools and good simulation software such as TINA and good device macromodels. These tools are vital-- not jus handy!

    By the way did TI bother to produce a model for any of their analog temperature sensors? Yes, just one-- the TMP20. That is, unless I missed finding them, which is a distinct possibility given the really poor TI website.

    Bite the bullet and pay Bill Sands to produce some decent models.

  • 0 in reply to Neil Albaugh
    TI__Mastermind 40295 points

    Neil,

    Apologies for the delayed reply, and thanks for the feedback on the macromodel.

    A quick check of the macro netlist shows some ground-referenced nodes that are most likely causing the problems with single-ended supplies.
    When split, balanced supplies are used with this kind of model architecture, mid-supply coincides with ground potential, and the model works fine.
    Unfortunately, this isn't the case with single-ended supplies, and the model behavior reflects this.

    Our newer models avoid this problem, but  this model's netlist is from 2002.
    The model was created well before our corporate-wide initiative on best practices for analog models.
    We are working hard to correct some of these problems, but it will take time.

    Thanks again for the feedback.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 30 points

    John,

    I wasted some time on this faulty model with a single supply voltage.  Seems to work find with split supplies.  I have an immediate application for this part and it would be nice to simulate it under the actual conditions I will be using, which is not with +2.5V and -2.5V supplies.  Any chance for a corrected model soon.

    Thanks,

    Bruce

  • 0 in reply to Bruce Bomar
    TI__Mastermind 40295 points

    Bruce,

    Sorry to hear the model doesn't meet your needs.

    Some quick sims confirm the model doesn't work for single-ended power supplies.

    One of the contributing factors is that some of the model's internal nodes are referenced  to ground (node 0).
    This practice usually results in some signal-carrying nodes being driven into the rails when supplies are single-ended.

    Changing the internal ground reference to half supply restores the output signal, but the gain and output common-mode are still off.

    I'll keep working on this; the model does need to be fixed.

    Regards,
    John