MSP430F135: MCU ADC failures

Hello Bernie,

Thanks for your detailed post.

Bernie Ryan said:

Why won't this app allow me to insert my scope pictures of ADC glitch?

Can I post the glitch analysis as a PDF file?

You can insert images using the "Insert/Edit Media" button below.

You can insert .pdf files using the "Insert File" button shown below.

Bernie Ryan said:

We have been encountering a problem on about 10% of our detector boards where the ADC in the MSP430F135 MCU will returned corrupted measurements. This will happen at random times when temperature cycling the boards during a system burn-in process.

Is the device ever subjected to temperatures higher than those specified in Section 5.1 in the datasheet? If so, the device can be permanently damaged.

Bernie Ryan said:

We are also using the internal reference of the MCU set to 1.50 volts with the recommended filtering capacitance values of Fig 5-15 of the MCU data sheet.

Does changing the internal reference buffer mode (ADC10SR) from 1 to 0 change the behavior?

Bernie Ryan said:

Does this sampling glitch appear to be typical for the MSP430F135 parts? All the PCBA display a similar glitch when doing ADC conversions.

This could potentially be caused by errata ADC11 that is present on Rev L devices. Do you see this behaviour in non-L devices? Besides ADC11, there are several other ADC-related erratas for these devices that could be causing the issue.

Regards,

James

Regarding your questions

1st

No, device is not subjected to higher temperature in burn-in.

Soldering is done using reflow method by outside assembly house

2nd

Comment doesn't appear to applicable to the 12 bit ADC variant of the MCUMSP430F135 ADC glitch problem.docx

3rd

We have Rev N device on our boards showing issues so ADC11 errata would appear not to apply

I tried to insert PDF file and not sure it worked correctly, will try againMSP430F135 ADC glitch problem.pdf

Thanks for the additional details. Can you share your ADC configuration code? It would be helpful to know things like ADC12CLK frequency, SHT0x sampling time, etc.

With regards to the observed glitch, Section 17.2.2 in the user's guide describes how the ADC12 module uses the charge redistribution method. When the analog inputs are internally switched, the switching action may cause transient inputs on the input signal. Ideally, the transients decay and settle before conversion or else they could cause errant results. The transients are most likely caused by the in-rush current flowing through the mux (up to 2kOhms) and charging up the input capacitors (can be up to 40pF). Typically, the higher the external source impedance, the higher the transient will be. Assuming the op-amps output is low impedance, the transient amplitude shouldn't be too large.

What DC offset voltage are you using? Combining this offset voltage with the "glitch" amplitude, is there a possibility that the maximum ADC input is exceeded, perhaps damaging the device requiring you to replace it? Can you measure or test this?

Regards,

James

Answers to your questions

1. configuration code/ADC set-up

//...... SNIP ...............

// Clock setup

DCOCTL = (DCO0+DCO1+DCO2);
BCSCTL1 = (RSEL0+RSEL1+RSEL2);

//...... SNIP ...............

// Port 6 --------------------------------------------------------------------
// P6.0 not used
// P6.1 not used
// P6.2 not used
// P6.3 AN0 - analog input CH0
// P6.4 AN1 - analog input CH1
// P6.5 not used
// P6.6 not used
// P6.7 not used
// port 6 selection
//  7 6 5 4  3 2 1 0
//  0 0 0 0  0 0 0 0   1=output
P6SEL = 0x18;
P6DIR = 0x00; // all inputs
P6OUT = 0x00; // inputs low

//...... SNIP ...............

// ADC Setup

ADC12CTL0 = ADC12ON+REFON+SHT0_5 + MSC; // Setup ADC12, ref., sampling time

ADC12CTL1 = SHP + CONSEQ_3;           // Use sampling timer, repeating sequence of multiple channel

ADC12MCTL0 = INCH_3+SREF_1;             // select channel A3 <==AN0
ADC12MCTL1 = INCH_4+SREF_1+EOS;                  // select channel A4 <==AN1
ADC12IE = 0x01;                       // Enable ADC12IFG.1

Delay(100);       // Delay for reference start-up
ADC12CTL0 |= ENC;                     // Enable conversions
ADC12CTL0 |= ADC12SC;   // Start conversion - once

2.  Signal amplitude to ADC

I should add that the ADC glitch appears the same on both "good" and "bad" boards.  Glitch amplitude varies greatly from 50 to 200 mv.  Not sure why so high as calculated glitch should be less than 10 mv.  Assume 1.5v DC input to ADC and 2k multiplexer impedance and 0 volt on SAR caps would give current of 0.75 ma, then assuming a 10 ohm source impedance glitch amplitude should be 7.5 mv.

Over voltage seem unlikely as ADC input is driven by op-amp on +/-2.5v supply, ADC input is unipolar positive signal derived from photo diode, op-amp saturates at 1.6 volts, glitch amplitude is 150 to 200 mv on top of 1.6v max signal, doesn't appear over voltage should be an issue as MCU is ona 3.3 volt supply.

I've noticed in some app notes that the inclusion of a anti-aliasing filter with a large cap can help absorb the SAR transient.  We need a 159 kHz bandwidth so a filter could be made from a 22 ohm and 0.043 uF cap.  I'll try this experiment to see if the glitch can be reduced.

Finally on my scope picture of the disconnected P6.4 ADC input, the glitch start appears to lead the the P6.3 input sampling glitch, are we seeing the input mux sampling the input before the actual conversion starts?