Serious I-Excec problems with ADS1248

Hi Tomas,

I do apologize that you are having trouble with our ADS1248.
As I did not see this issue before, could you please be a little bit more specific about the conditions when this happens, what you do before it happens, schematic and overall configuration?
If you don't want to share this information here in the forum I can give you my email address so you could send it to me directly.

On another note. Unless you have a bipolar supply for ADS1248 you can't measure single-ended signals referenced to GND with it.

Regards,

Joachim

As I mentioned it happens trandomly, so I cant say what happens before, since it might take 100 hours or 2 minutes before it happens.

Our design is quite simple, we use 4-wire PT100, upper line is fed directly from I-Excec, clamped to Agnd AVdd with BAT54, second line to input 1 via 100k/10nf filterand clamped to Agnd and AVdd.

Third line is the sam, but to input 2.

Fourth line to Agnd via a 2k2 resistor, (I have recently seen the design where the reference inputs is used, and that will be changed at the next revision).

Again this line is clamped to AVdd and Agnd.

Regarding the single-ended inputs, we feed Vref into one of the inputs.

When we first discovered this problem, we changed the code so we start the IDAC's before we take tyhe measure, but that does not help, thew only way so-far to solve the problem, is to do a reset of the 1248, and re-initiate it.

As I wrote earlier, this problem only affects the I-DAC's , since the single ended channels is still working, the internal V-ref should be ok.

I thought ther shode be a "code-tag" for posting code, but I cant find one, so I cant post my Init-function

Hi Tomas,

thank you very much for the further information.

You can actually attach files to a post. It is not really obvious how I have to admit.
When you reply to a post, you see three tabs "Compose | Options | Preview" on top of the quoted post. If you go to "Options" you should be able to upload files there.

What supply voltage are you using for the ADS1248, what current did you set the IDAC to (I assume 1mA) and what gain are you using?
Actually I am surprised to see that you place such big filter resistors on the inputs. They introduce additional thermal noise and any differential input current flowing into/out of the ADC inputs will manifest in an additional offset voltage across the 100k resistors.

Do you currently use the internal voltage reference to measure the Pt100?
As you mentioned we would recommend to use e.g. the voltage across the 2.2k as the reference voltage (connected to REFPx, REFNx) for the measurement to have a ratiometric approach.

Do you mean you feed the VREFOUT to one of the inputs to double check if the internal voltage reference is operating correct?
As I said before you should not get correct readings then, as you can't measure single-ended signals unless you use a bipolar supply.

How do you actually test if the IDAC has shut down?

If you could upload any code example or schematic extract that would probably be helpful, especially how you do the single-ended measurement.

Regards,

Well, they are sitting in a very harch enviroment. Next to a 40kW inverter, so I figure that a little extra filtration is not a bad idea.

Anyway.

AVdd = 5 V from a NS(TI nmow i guess) Linear LDO 24V feed.

DVdd = 3,3V from a simple switcher.

4-layer board, different planes fo AVss and DVss, standard stack, signal, Vss, Vdd and signal.

I use 1 mA to excite the PT100.

I use the internal VRef for all measurments.

I am feeding Vref via 10k into one of the AD-inputs and use that one incombination with de singleended inputs, i.e I do a differential measurment betwen Vref and the single-ended signal, workas out quite well actually.

The way I test it is to measure the voltage over the PT100, and I also basically have a 0 reading on the PT100 inputs.

PGA is set for 8, and I use 20 samples/sec.

I have 5 AD-converters set up in paralell, sharing Reset and Start, individual SS and the DReady is ORed together.

Hense I start them all at the same time, do a single conversion, then fetch data from each of them.

I have two PT100 sensor per AD and three singelended channels, with the remaining input as reference for the singleended channels.

As I mentioned, when this happens, it is only the differential channels using the I-DACs that is affected, i read proper data from my "single ended channels" hence the internal reference must be working correctly, the same with the multiplexer, the PGA and the actual AD+ SPI, It is ONLY the I_DACS that are affected, either they shut down or the internal MUX/Switches stopps working. 

If I can have your e-mail I can send you pdf's of the relevant schematics, as well as the c-files we use

Thanks