ADS1220: returns full scale data in Two-wire RTD

Hi Samuli,

The differential capacitor (10nF) in the anti-aliasing filters for the analog inputs is 10 times smaller than the common mode capacitors (100nF). 

TI recommends that the differential capacitor CDIFF be at least an order of magnitude (10x) LARGER than the common-mode capacitors (CM1 and CM2), because mismatches in the common-mode capacitors can convert common-mode noise into differential noise. This is mentioned in section '9.2.1.2 Detailed Design Procedure' of the data sheet.

From the schematic, I'm assuming the IDAC excitation current is being configured for the AIN3 pin. Have you tested if there is a voltage drop across the reference resistor to confirm the IDAC current is being set properly?

From what you have shared, it seems like the IDAC current is being set to '250uA'.

If this is the case, using a 1.65kΩ reference resistor for a ratiometric measurement, the voltage across the reference resistor will only be ~0.4125V.

When using an external voltage reference, the minimum Vref voltage should be 0.75V, otherwise you are operating the device outside of the recommended operating conditions:

Try increasing the value of the IDAC current, or the reference resistor so that the external reference voltage applied is within the values stated in the data sheet.

It is also always a good idea to probe the SPI communication lines with a logic analyzer to make sure that the digital communication to the device looks okay.

Best Regards,

Angel

Hey Angel,

We swapped the capacitors from 10nF to 1μF as you recommended.


Here is our new connections:




We also tested the voltage across the 1.65k reference resistor and with CFG2_IDAC_500UA we measured 0.85V. We also measured with different IDAC values to be sure that the IDAC is working. For exampe with 1000uA we measured 1.69V.

Yet still, with these changes we only receive full-scale values. 


Here is console prints of our test program output(Configuration registers, Raw ADC data)

Hi Samuli,

Are you issuing a START/SYNC command to start a conversion after setting your register configurations? 

How are you determining when a conversion result is ready?

Rather than console prints, can you provide captures of the digital communication of the write/reads to the ADC registers and when doing reads of the ADC data?

Best Regards,

Angel

adc_troubleshoot.zip

I've attached a zip with images of the requested captures.
It contains attempts with both single-shot and continuous mode and the images are numbered in order in which they happened.

We are issuing START/SYNC when using single-shot mode after setting the registers. This is followed by a 1 second delay, after which we read the 3 bytes.

In continuous mode we are reading 3 bytes with RDATA command every second after setting the configurations.

So currently we are only using delays to make sure the conversion result is ready when we read the bytes.

In every case the 3 bytes read are 7F FF FF.

Hi Samuli,

I'll take a look at these captures and get back to you tomorrow.

Best Regards,

Angel

Hi Samuli,

I still need to look at these captures in more detail but should be able to get to them by EOD tomorrow.

Best Regards,

Angel

Hi Samuli,

I went through the digital communication captures in detail, thanks for sending these, these are always very helpful when troubleshooting.

You can write to the configuration registers and read them back correctly which verifies the communication to the device is working.

One thing to point out, is that you sent two folders, one for continuous-conversion mode, and one for single-shot mode, but in both sets of captures, you are setting the device to continuous conversion mode.

Both sets of captures you are writing 0x64 (0110 0100b) to configuration register 1, which sets it to continuous conversion mode:

To set the device to single-shot mode keeping the same settings otherwise, you would write 0x60 to configuration register 1.

In the continuous conversion mode captures the register settings are the same as the single-shot mode captures, but this time they are correct as you are intending to set the device to continuous conversion mode.

In the continuous conversion mode captures however, you are not issuing a START/SYNC command after setting the register configurations. This command is needed to start ADC conversions:

If the command is not used, there is no ADC data to be read when you issue the RDATA command: 

The IDAC, input MUX, external REFP0-REFN0, and most of the settings are set correctly for your circuit shown in the schematic capture.

One thing that stood out however, is that in Configuration Register 2, you are enabling 60-Hz rejection:

This should only be used with the 20-SPS data rate setting in normal mode, but you are setting the data rate to 175SPS in configuration register 1:

This may be causing the issue.

Either set the data rate to 20SPS if 60-Hz rejection is desired, or write '00b' to the FIR filter field if a data rate other than 20SPS is desired:

Also make sure to issue the START/SYNC command before doing ADC data reads after setting your configuration registers, and that you are setting your desired configurations properly (single-shot mode for example).

It might also be a good idea to monitor DRDY after issuing your START/SYNC to ensure ADC data is ready before doing an RDATA.

Hope this helps!

Best Regards,

Angel

Hey Angel,

We modified the configuration registers as you suggested:

Register 0: 0x61
Register 1: 0x60
Register 2: 0x45
Register 3: 0x82

We are not sure if we "need 60-Hz rejection", so we set it off and left the data rate to 175 SPS.

We also checked the DRDY pin:

When sending the START/SYNC it goes from low to high for a short time and then back low.
After that, reading data with RDATA sets the DRDY signal high.

All we can get is 7F FF FF still.

Hi Samuli,

Samuli Oja-Kaukola said:

When sending the START/SYNC it goes from low to high for a short time and then back low.
After that, reading data with RDATA sets the DRDY signal high.

Correct this is how DRDY should behave to indicate when new ADC conversion data is ready to be read.

What you are doing now seems correct so I'm not too sure what the issue may be that you are only getting full-scale readings.

Have you only tried testing one device?

The device may have been damaged at some point so it would be a good idea to try a different ADS1220 to see if the results are different.

You can also try a different configuration with the ADC such as shorted inputs to GND or mid-supply using the internal reference, or a single-ended measurement using the internal reference. To see if you can get ADC output readings that are not full-scale (7F FF FF). 

These should give something other than full-scale, so if otherwise, the device may be damaged.

Best Regards,

Angel

Hello Angel,

I meet the same issue in ADS1220, and i found if disconnect the PT100 while ADS1220 is running, connect the PT100 to ADS1220 again, the result goes well.

it's very strange, you can check the log below, thanks.

ADC Code: 8388544
RTD Resistance Value: 824.993713
Temp: 9223372036854775807.9223372036854775807

ADC Code: 8388544
RTD Resistance Value: 824.993713
Temp: 9223372036854775807.9223372036854775807

ADC Code: 8388544
RTD Resistance Value: 824.993713
Temp: 9223372036854775807.9223372036854775807

ADC Code: 1493824
RTD Resistance Value: 146.914093
Temp: 122.247627

ADC Code: 1513664
RTD Resistance Value: 148.865325
Temp: 127.435073

ADC Code: 1515840
RTD Resistance Value: 149.079315
Temp: 127.998191

LEON

Hi Leon,

Thanks for sharing this. 

It may be worth it to try this and see if it fixes the issue.

Best Regards,

Angel

I found a similar issue discovered some time ago where with an RTD measurement and the device configured, the first conversion was always wrong. 

Following a POR, a RESET or POWERDOWN command, the first conversion following START/SYNC was always in error.  In the case of the ADS1220 there is an automatic conversion that takes place following POR and RESET.  The action however differs for the POWERDOWN command that turns off the analog circuitry.  When leaving the POWERDOWN state the START/SYNC command is issued to wakeup the device.

When the START/SYNC command is issued, the IDACs are turned on which requires time for the IDAC current path to settle for both ADC inputs and reference inputs. 

Have the following ADC conversion results after the first full-scale ADC reading been checked?

Best Regards,

Angel.

Hello Angel,

We followed the command sequence as POR, RESET command, START/SYNC command sent one by one, and the wrong value will continuously output in this condition( tested for 10 minutes at least). when i disconnect PT100, the output value is still 8388544, after i reconnect PT100, the value goes well.

I also measured the voltage while the wrong value condition, REFP is around 1.65V, and the voltage between VIN0 and VIN1 is the normal value around 55mV.

Does there have any register or something others that i can check for further analysis.

Thanks.

LEON

and current register setting is below.

ADS1220_0_REGISTER: 0x66
ADS1220_1_REGISTER: 0x04
ADS1220_2_REGISTER: 0x65
ADS1220_3_REGISTER: 0x70

Hello Angel,

Thank you.

and the circuit we used is 3-Wire RTD as below.

LEON

Hello Angel,

I might got the root cause in this issue.

i found if i config IDAC1 and IDAC2 simultaneously, the error will come out.

but if i config IDAC2 first, then delay at least 3 seconds(the time might be related to the CAP and RES used as below). and config IDAC1. it works fine.

R REF = 1.65KΩ, R F3 = R F4 = 1 kΩ, C DIF2 = 100 nF, and C CM3 = C CM4 = 10 nF.

The log of config ADC register:

//Config IDAC2

[09:33:19.141]Temp:10
ADS1220_3_REGISTER: 0x10

delay 4 seconds.

//Config IDAC1

[09:33:23.544]Temp:70
ADS1220_3_REGISTER: 0x70

and if i config IDAC1 first, delay 3 seconds, config IDAC2, it is still wrong in value.

in the circuit, we can see that we are using external ref, it means that we may need setup external REF first with IDAC2, after the external REF is ready, enable IDAC1 for PT100, then we can do the ADC sample and get the right value.

Can you help to have a check internally, is it the root cause for this issue?

thanks.

LEON