- Ask a related questionWhat is a related question?A related question is a question created from another question. When the related question is created, it will be automatically linked to the original question.
We're using an MSP430F5336, which has the ADC12 module in it. We're powering the chip from 3.3 V, and using this for AVCC as well. We use AVCC as the ADC reference. When we tie an ADC input to ground, it doesn't read zero counts. Instead, it reads a very high count, close to full scale. Bumping the input voltage up slightly using a power supply, we see it begin to behave at a couple of tenths of a volt. What could cause this? We need to know where to look. I've pasted a code snippet below which shows the ADC setup. I should note that I'm the hardware guy, so if the firmware below contradicts anything I've said above, please tell me!
// ADC12CSTARTADD_0 - start with ADC channel 0
// ADC12SHS_0 - use ADC12SC for sample and hold source select
// ADC12SHP = Sample/Hold Pulse Mode
// ~ADC12ISSH = The sample-input signal is not inverted
// ADC12DIV_0 = ADC12 clock divider divide by 1
// ADC12SSEL_0 = ADC20 clock source select - ADC12OSC
// ADC12CONSEQ_3 = Repeat Sequence of channels
// ADC12BUSY - input
ADC12CTL1 = ADC12CSTARTADD_0 | ADC12SHS_0 | ADC12SHP | ADC12DIV_0 | ADC12SSEL_0 | ADC12CONSEQ_3;
// ~ADC12PDIV - Predivide by 1
// ADC12TCOFF - turn temperature sensor off
// ADC12RES_2 = 12 bit (13 clock cycle conversion time)
// ~ADC12DF = Binary unsigned data format
// ~ADC12SR = Sampling rate 200 ksps
// ~ADC12REFOUT = Reference output off
// ~ADC12REFBURST = Reference buffer on continuously
ADC12CTL2 = ADC12TCOFF | ADC12RES_2;
// enable ADC
ADC12CTL0 = ADC12ON;
// ADC12SHT1_3 = 32 ADC12CLK cycles sample & hold time (ADC8..ADC15)
// ADC12SHT0_3 = 32 ADC12CLK cycles sample & hold time (ADC0..ADC7)
// ADC12MSC = Multiple sample and conversion. Valid only for sequence or repeated modes
// ~ADC12REF2_5V = 2.5V (Only if ADC12REFON)
// ~ADC12REFON = Reference off
// ADC12ON = ADC12 on/enable
// ADC12OVIE = Overflow interrupt enable
// ADC12TOVIE = ATimer Overflow interrupt enable
// ADC12ENC = ADC12 conversion enabled
// ADC12SC = Software-controlled sample-and-conversion start
ADC12CTL0 = ADC12SHT1_3 | ADC12SHT0_3 | ADC12MSC | ADC12ON | ADC12OVIE | ADC12TOVIE;
Thanks,
Scott
I'd like to add some additional info to this. It turns out that there's actually an offset/wrap-around happening. I've attached an image showing actual vs. ideal. Please ignore the fact that the y-axis scale is a multiple of the actual resolution. What matters is the shapes of the two graphs.
**Attention** This is a public forum
