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.

Some hardware questions about the ADS1248

Other Parts Discussed in Thread: ADS1248

I'm working on a wave energy project that requires digitzing the following sensors (all at 10 Hz,  50 samples/sec in aggregate):

-- Two  type T thermocouples converting -20 to +100 Celsius, ADS1248's internal diode will be used as the "cold junction".  Software will perform the cold junction compensation.

-- One three wire RTD, converting -20 to +100 Celsius

--  Two pressure sensors that produce 0 to 10 volts

The proposed connection of the hardware and configuration of the ADS1248 follows:

The two thermocouples will be converted differentially.  An external differential RC filter will be used on each thermocouple to reject noise from nearby, kilowatt range DC to DC converters.  One side of each thermocouple will be biased to mid-supply.  The internal reference will be used for conversions. 

The RTD will be connected as described on page 57 of the ADS1248 datasheet except that REFP1 and REFN1 will be used to measure the voltage drop across RBIAS caused by the RTD's excitation current.

The 0 to 10V input from the pressure transducer will be divided by four (0 to 2.5V) and then converted single ended.  The internal reference will be used for conversions.  The resistive divider will be followed by a unity gain buffer to drive the ADS1248 inputs.  Is this buffer necessary? I'm concerned about any high impedance source interacting with the ADC input switching currents to produce measurement errors.

 What is the best way to isolate inputs from the ADC input switching currents?

What suggestions do you have for protecting the inputs of the ADS1248?

What is the recommended method to deal with ununsed analog inputs?

Thanks again for your help,

Wayne

 

  • Wayne,


    There's a lot to go over in your comments, but I'll make a few comments.

    1. If you use the ADS1248 internal temp sensor (in the system monitor) for the cold junction, just be aware that there may be some error. This sensor is based on the difference of voltage of two diodes mismatched in size and current. Based on bench measurements and final test data, the internal temperature sensor error is about +/-6 degC (for about 3 sigma of deviation). If you have questions about it, I would recommend measuring the temperature an alternate way and comparing the result with what you get from the ADS1248.

    2. Are you using the VBIAS function to bias the thermocouples? This is what they were designed for.

    3. What type of DC to DC converters are you using and how are you using them. I find that switching converters can throw off a lot of noise so you have to be careful with the design and layout.

    4, The buffer for the pressure transducer may not be necessary. The front end of the ADS1248 uses a gain amplifier (which is chopped for reducing 1/f noise), so the input impedance is pretty high. I would note that with the gain amplifier, it's less of an input impedance, and more of an input bias current to deal with. In the datasheet, we say that the typical differential current is about 100pA. However, Table 11 discussing the analog input current may more descriptive.

    5. If you're concerned about the input switching currents, using a capacitor across the input as a charge bucket to draw from is pretty helpful. It may depend on the sensor you're looking at and the impedance afteward, but about 10nF is sufficient to help with the input.

    6. If you're concerned with an overvoltage event from having a larger supply, the input can take about 10mA continuous as listed in the absolute max table. Some series resistance might help with that. Also, some BAT24 Schottky diodes from the input to  AVDD and to AVSS may shunt current to the supplies when the input goes above the supply. If you look at the ADS1248EVM users guide schematic, it shows some protection for the reference input. You should be able to apply this to the inputs as well.

    7. For the unused inputs,  I don't think there's any harm in floating them, but if they aren't going to be used, I'd just ground them to AVSS.

    If you have any other questions, you can post them here. If you have a schematic for me to look at, let me know and I can contact you directly.


    Joseph Wu

  • Joseph

    The purpose of the project that I'm working on is to generate power from wave energy.  There is a 5 KW DC to DC converter aboard that charges sealed lead acid batteries from the electrical energy generated by waves.  Ultimately, the system will keep position without being anchored and serve as a filling station for autonomous underwater vehicles to dock and recharge their own batteries. 

    To badly misquote a certain computer company(RIP) slogan, "The system is the DC to DC converter".   When you ask about DC to DC converters, I should ask "Which one?".  

    Seriously, the sensor and data logging electronics are mounted in a separate enclosure a few meters away from the power inverter (the real source of noise).  Communication with the power inverter is done through an isolated RS-232 connection.  Both systems do not share a common ground.  Feed through capacitors are also used on the power inputs to each enclosure.

    In response to your reply:

    1.   We're measuring temperature as way of gauging the efficiency of our power conversion.  The relative temperature of sea water with respect to two gas chambers in our system gives us this info.  If I need more accuracy, I will use a two wire RTD as the "cold junction".  Thanks for the warning.

    2. VBIAS it is.

    3. On the sensor board itself, there is a DC to DC converter for converting the nominal battery voltage into +5.5V.  The +5.5V is fed to a LC filter and an LDO before being delivered to the ADS1248 as +5V.  There are plenty of ferrite beads in strategic locations and a common mode choke on the input to the DC to DC converter.  Sadly, there are other DC to DC converters for isolating several RS-232 interfaces.  Certainly, it isn't an ideal environment for any low level analog signal conditioning.

    4. & 5.  After scaling, the pressure sensor's output is about 8 mV/PSI (0 to 2.5V range).  If I just use a divider without a buffer, I would want the voltage generated by the 100pa  "input bias current" flowing through the resistive divider to be significantly less than 1 mV.    Seems like anything less than 100K will work.  Hopefully, I have understood your comment.

    6.  We need to make temperature measurements very rapidly using some 1 mil thermocouples with a 3 ms(!) time constant.  Due to the Rube Goldbergian mechanical arrangement of the generator, these fine thermocouple wires and the 500V output from the generator must pass within 10 cm of each other.  The ocean is a nasty place for anything man made; I want to protect the ADS1248 inputs in the event that the thermocouple wires are shorted to the generator output without comprising the accuracy of the temperature measurement.  I was thinking that a Gas Discharge Tube might be a good approach (very little leakage current until the breakdown voltage is reached).

    7. If I only use the internal reference, what should I do what the unused reference inputs? Ground them?

    Thanks for the fast and detailed response, I'm amazed at the great service!

    Wayne

     

  • Here are just a couple of further comments:

    I'm not sure about using the ferrites and inductors in the supply path, a linear regulator will be fine by itself. In the ADC there is a significant amount of switching (from both the sampling and the digital section) that the changes in current will cause some extra supply noise with the inductors. Generally, I leave them off.

    For the input protection, I've never used gas discharge tubes. For the level that you're concerned with, I don't have any good answers. We do make some ESD protection diodes used to protect against high-voltage surges, but in your case the voltages may last for a more prolonged period of time and I don't think the ESD protection would help.

    For the extra reference input, instead of grounding this, I would float the extra input. I believe there are cases where the internal reference can be tied to the reference input and for this case, it's better just to leave them open.

    I'll start taking a look at the software questions soon.


    Joseph Wu