Confused with K60 ADC

[keithtang]

I am very new to Freescale 32-bit architecture. I am reading the document K60P144M150SF3RM Rev 2 Dec 2011, and get very confused with the ADC configuration in the K60 chips, hope someone can clarify it to me.

While I understand there are 4 ADCs, 4 Differential and 24 Single-Ended channels, I am confused with:

1. the term "interleaved" in the document. Does it mean "mutually exclusive" or "multiplexed"?

2. the relationship between the pins (ADCx_SEn, ADCx_DPMn), and the channels (ADCx_ADn). I don't find any detailed block diagram illustrating the interconnection of these. For instance, I am confused:

        (a) if ADC0_SE11 and ADC1_SE11 are connected internally?

        (b) why are there multiplexed channels such as ADC0_SE7A and ADC0_SE7B? The MQX library does not seem to multiplex it at runtime.

3. the built-in temperature sensor. Do I have to physically assign a pin to it? My guess is no, but I need verification.

Thanks.

[mjbcswitzerland]

Hi

1. I don't know what exactly interleaved means but SW controls the ADC input actually used so I would suggest interpreting it as either/or. Maybe both ADC0 and ADC1 can use the same input at the same time (same signal of source) but I don't think that there would be reasons to do this.

2.

(a) ADC0 and ADC1 are different AD converters and so their pins are unique. ADC0_SE11 and ADC1_SE11 are different pins and not connected internally.

(b) Some ADC channels can be connected to two different physical pins, controlled by the ADCx_CFG2[MUXSEL] bits. When the software doesn't specifically control this the default (a) pin will be used.

3.

The internal temperature channel is a dedicated internal one and is not connected to any external pin.

Regards

Mark

[iafg]

RE - Interleaved:

It is possible for two ADC channels to sample the same input signal on certain pins.(channels).

This COULD be useful to achieve a higher sample rate than a single channel would be able to attain. The timing of the start of conversion can be controlled using the precision time delay block, so a second channel can start conversion when the first channel is 50% done.

I am not certain of the complexity on handling the data after it is converted.

This is an unusual use-case, but I have heard of people considering it to try to achieve faster than 1MSPS sampling using the K60.