PSOC 5LP Internal crystal ADC max resolution

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
PaSv_1652016
Level 2
Level 2
10 likes given 5 likes given First like received

Can i use PSOC 5lp  internal clocking for   effective 18 bit  resolution ADC.
Or i  need use ECO ?

Clock Source The ADC can be clocked by a source internal to the ADC component, a source external to the component but internal to the chip using a standard clock component or UDB, or by a source external to the chip. The internal or external clock selection is made using a radio button. When external clock is enabled, a clock input pin is displayed on the ADC schematic symbol. External clocks must have 50-percent duty cycle. The internal clock is guaranteed by design to have the correct duty cycle. Clock stability is important for achieving low noise operation. One of the effects of jitter is substantial spreading of the signal. These are clearly shown in the following FFTs. The signal-tonoise ratio (SNR) of the ADC can be significantly improved with the use of an external clock. Delta Sigma Analog to Digital Converter (ADC_DelSig) PSoC® Creator™ Component Datasheet Page 20 of 50 Document Number: 001-79377 Rev. ** One measure of ADC performance is the distortion caused by INL, DNL, and timing errors. This distortion is measured by driving the ADC with a “perfect” sine wave then evaluating the output using an FFT (fast Fourier transform). Excess INL results in harmonic distortion with “lines” at multiples of the input sine wave frequency. Excess DNL results in an overall increase in the noise floor. Timing errors (frequency jitter) in the clock to the ADC result in spreading of the indicated fundamental from a nominal single line in the FFT. The PSoC internal clock has more frequency jitter than an external crystal oscillator. The upper plot shows the noise and spread in the FFT at the fundamental from the internal clock. The lower plot shows substantial reduction in the spreading of the fundamental as a result of the reduced frequency jitter from the external crystal oscillator.

pastedImage_0.png

0 Likes
1 Solution

PaSv,

Thank you for your specs.

You list: "0.1% ~16-18 bit".  What is the min and max readings you need to support.  This will determine that capability of your system's FSR (Full Scale Resolution).

You list:  "my device will operate in extremely noisy environments (HiGh EMI, and  up to 200kV spark )".  Do these factors influence the signal you are measuring or the environment the PSoC has to isolation from?  If these factors are in the signal, it is much better to provide a windowed average scheme across many samples.   If these factors can influence the PSoC measurement and you need to isolate it from it, there are board layout techniques that should be employed.  For example, you can use a 4-layer board with power and GND planes.

You list: "full diffirential ADC Ultra High Voltage current shunt measure".  Of course you will need input isolation from the high voltage.  It sounds like you are trying to measure motor currents in a very voltage system.  My experience with motor currents is that the signal is extremely messy.  Even with BDLC drive circuits, you get eddy current fluctuations, PWM drive induced noise, load variation-induced noise, rotational bearing friction-induced noise for starters.  If it is a brushed motor design, there is armature commutation switching noise.

For this type of design, it is always best to take the "long-view" approach to current measurements.  The "long view" is to take highly averaged readings to try to minimize these noise influences.  If this is the case, an external XTAL is not going to help much.

Len

Len
"Engineering is an Art. The Art of Compromise."

View solution in original post

7 Replies
Len_CONSULTRON
Level 9
Level 9
Beta tester 500 solutions authored 1000 replies posted

PaSv,

You've provided very a interesting question.

What accuracy do you need for your application?

For example:  If you were going to use the ADC to sample audio data at 44.1KHz, where each sample counts.  Then you might want a external XTAL using the ECO to minimize jitter-induced SNR distortion.

If you are reading a fairly constant battery source that doesn't change very offset or very fast, you might want to use the internal oscillator and do some windowed averaging.

The external XTAL is definitely your more expensive (but more accurate) option.

Len

"Engineering is an art.  The Art of Compromise."

Len
"Engineering is an Art. The Art of Compromise."
0 Likes
What accuracy do you need for your application?

-> -> tends to  ->  0.1% ~16-18 bit  (full diffirential ADC Ultra High Voltage current shunt measure , high speed process , but measure about ~1000 sample\sec )
my device will operate in extremely noisy environments (HiGh EMI, and  up to 200kV spark )
and I understand that I do not get these results, but I have to eliminate all possible factors.

0 Likes

PaSv,

Thank you for your specs.

You list: "0.1% ~16-18 bit".  What is the min and max readings you need to support.  This will determine that capability of your system's FSR (Full Scale Resolution).

You list:  "my device will operate in extremely noisy environments (HiGh EMI, and  up to 200kV spark )".  Do these factors influence the signal you are measuring or the environment the PSoC has to isolation from?  If these factors are in the signal, it is much better to provide a windowed average scheme across many samples.   If these factors can influence the PSoC measurement and you need to isolate it from it, there are board layout techniques that should be employed.  For example, you can use a 4-layer board with power and GND planes.

You list: "full diffirential ADC Ultra High Voltage current shunt measure".  Of course you will need input isolation from the high voltage.  It sounds like you are trying to measure motor currents in a very voltage system.  My experience with motor currents is that the signal is extremely messy.  Even with BDLC drive circuits, you get eddy current fluctuations, PWM drive induced noise, load variation-induced noise, rotational bearing friction-induced noise for starters.  If it is a brushed motor design, there is armature commutation switching noise.

For this type of design, it is always best to take the "long-view" approach to current measurements.  The "long view" is to take highly averaged readings to try to minimize these noise influences.  If this is the case, an external XTAL is not going to help much.

Len

Len
"Engineering is an Art. The Art of Compromise."

This solution should measure current from for the scientific equipment up to 5A in pulse under up to 150kV.
i have battery powered module connected via Fiber Optic chanell.
Battery powered module include precision low induction shunt .
Shunt connected through Isolated amplifier and DCDC transformer module . DCDC i use for isolate High Voltage noice ground from ADC\mcu grounds.
line   from  shunt to Isolated amplifier have  protection  element and full differential design.
I chose 2-layer board beacouse  i have long differential line with protection  component,  filters  , I limits R .

From qfn64 PSOC5 i  have only ADC input + digital out  to FiberOptic
For all PCB regions i use ground planes

pastedImage_0.png

0 Likes

PaSv,

I like your isolation method.  Thanks for sharing it.

What is the accuracy of your isolation amp?  This piece might be the limiting factor in your accuracy goal.

I found the best way to minimize most of your accuracy error sources is to perform a calibration for gain and offset.  These values can be stored in EEPROM.   If you find in development that you have non-linearity errors, you can create a table of gain and offsets across the input range.  Sadly, this will slow down your calibration cycle.

Len

Len
"Engineering is an Art. The Art of Compromise."
0 Likes

PaSv,

It doesn't look like the board will hold 150kV difference between shunt and fiber optic connector. What is estimated voltage drop from shunt + amplifier to PSoC to fiber optic to the ground? What isolation Opamp did you use?

/odissey1

0 Likes
HIt doesn't look like the board will hold 150kV difference between shunt and fiber optic connector.

it's not needed.  Module have long wire  from shunt  and    isolation  only  for  noice  ground   isolation. And fiber optic\shunt module inside shielded metal box powered from Li-Ion battery. Fiber optic long  wire   used for translate  digital signal from Shunt to long  distance safe place for people

What isolation Opamp did you use?

AMC1301
http://www.ti.com/lit/ds/symlink/amc1301.pdf

pastedImage_5.png

0 Likes