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.

DAC8760 load specifications and design considerations

Other Parts Discussed in Thread: DAC8760, DAC8760EVM

Hello,


I am designing an automated measurement & characterization system for various resistive sensors and the DAC8760 seems to be an ideal choice for the application since it can provide a current output with suitable voltage compliance and a variable voltage output which might be useful for additional functions.

The sensors I have to test display a resistance between approx. 50 Ohms and 60 kOhms, depending on the sensor type.
The low resistance sensors require a max test current of approx 22mA, the high resistance sensors require a test current around 500uA and there is also a third type with intermediate characteristics. Therefore as you understand I will need to cover a broad range of required current and voltage.

If you could provide some further details and recommendations to the questions below it would be very helpful.

1) Are there any limitations concerning the load impedance? Will it work ok for my application? As far as I saw the datasheet specifications are based on 300 Ohm loads only. Could you provide some data for the current output as a function of  the load impedance?

2) I suppose it would be preferable to use both an external Rset and a boost transistor for my case since I will probably have a very large power dissipation on the IC otherwise. Are there any specific NPN transistors that you would recommend as a pass element?
Perhaps you could also provide some hints to improve the response and minimize offsets and stability issues due to the external transistor?

3) I would like to insert a low value (10Ohms) external sense resistor to measure the current applied. Would you recommend a differential/instrumentation amplifier or would it be better to use a chopper amplifier in order to minimize errors across the current range?
If you can suggest some suitable parts and an optimized schematic it would of course be appreciated.

4) I would like to alternate between connecting the current output and the voltage output of the DAC8760 to the device under test. It will not be required to change with a high frequency, only when the system initializes.
Do you think it would be preferable to implement this using a changeover relay or using 2 discrete JFETs or some other way in order to minimize noise etc?

Thank you for your time!

  • Anastasios,

    Anastasios Moschos said:
    1) Are there any limitations concerning the load impedance? Will it work ok for my application? As far as I saw the datasheet specifications are based on 300 Ohm loads only. Could you provide some data for the current output as a function of  the load impedance?

    Parametric performance (i.e. linearity, offset, gain, etc.) should remain mostly the same regardless of the load - assuming we leave out any effects of any heating caused by different loads. The primary issue to consider in terms of functionality is the loop compliance voltage, the high-side supply voltage AVDD needs to be 2V higher than the voltage developed by the output current passing through the load. In order to drive a 60kOhm resistor with 500uA of current you will require at least a 32V AVDD supply voltage. The other case, 50ohms with 22mA, would only require the minimum AVDD supply voltage.

    Anastasios Moschos said:
    2) I suppose it would be preferable to use both an external Rset and a boost transistor for my case since I will probably have a very large power dissipation on the IC otherwise. Are there any specific NPN transistors that you would recommend as a pass element?
    Perhaps you could also provide some hints to improve the response and minimize offsets and stability issues due to the external transistor?

    Given the wide range of loads / output currents you need in your application, doing something to manage heat when driving the small loads is a good idea. Using the external boost transistor configuration is one possible solution assuming you use the largest acceptable thermal pad you can in order to realize the optimal thermal resistance profile of the transistor - if you look at most of these datasheets the manufacturer will supply two figures, one that assumes a smaller "thermal pad" and another with larger. 

    The DAC8760EVM shows an implementation of the external boost configuration with a FZT651TA and a compensation circuit that should deliver stable operation over all conditions.

    Another option is documented in a TI Design from some of my colleagues called TIDU709. This option uses a buck-boost converter along with a feedback circuit that "adaptively" adjusted the high-side voltage (AVDD) in order to provide the lowest required high-side voltage to drive the corresponding load and therefore optimize power/heat dissipation. 

    Anastasios Moschos said:
    3) I would like to insert a low value (10Ohms) external sense resistor to measure the current applied. Would you recommend a differential/instrumentation amplifier or would it be better to use a chopper amplifier in order to minimize errors across the current range? 

    If you can suggest some suitable parts and an optimized schematic it would of course be appreciated.

    As mentioned in the first question, the biggest thing to keep in mind is the compliance voltage for the current output to continue driving the load in a linear fashion. If you insert 10ohms in series with the load this will eat into the maximum load impedance you can drive with a constant supply. In your case you still have headroom to raise the AVDD supply voltage by a bit to compensate for this.

    Choosing an amplifier depends on the accuracy requirements you have for the current measurement and the operating temperature range of your system.

    Anastasios Moschos said:
    4) I would like to alternate between connecting the current output and the voltage output of the DAC8760 to the device under test. It will not be required to change with a high frequency, only when the system initializes.
    Do you think it would be preferable to implement this using a changeover relay or using 2 discrete JFETs or some other way in order to minimize noise etc?

    There is an application note that describes software and hardware compensation schemes for sharing the voltage and current outputs on a single pin. The document explains how to do it using the internal PMOS but the boost configuration would look the same, just including your additional sensing circuitry etc. There's a TI Design that goes into detail on the analog approach as well.

    http://www.ti.com/lit/an/sbaa199/sbaa199.pdf

    http://www.ti.com/tool/tipd119

  • Thanks for the quick reply and the suggestions, I hadn't seen that there were complete schematics for the eval. module. I'll study them further and let you know if there are any points requiring clarification.


    Concerning 1) there is a point that is still unclear:
    I don't think that there will be a situation where I'll need more than 32V or 24mA on the load and the external transistor & Rset will alleviate the power dissipation on the IC so I suppose that won't be a problem.
    However, according to the TIDU709 that you mentioned there is a specification that the load should not exceed 1kOhms (Page 3, Table 1, 3rd row), which is somewhat confusing. Is that a specification of the HART protocol or is there some limitation of the DAC device?
  • Anastasios,

    They / we often treat these reference designs as "mock" products - so in this case the reference design is specifying that the performance discussed in the document is based on driving a load between 1 to 1000 ohms. The only functional implication of the device is as discussed in my previous post.
  • Aha, that makes sense. Thank you for everything! I'll proceed with the design and let you know how it goes. :)