ADC referense for XMEGA.

Go To Last Post
14 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

startet or moved a project fra AVR to XMEGA.

I am not see the datasheet near enough and was thinking at ADC ref can connect intern to VCC like AVR? well I do not find the way to do that,

must used externe ADC ref. if you want Vref= VCC

bad enough.

but why all intern ADC referanse for Xmega are so low? 

the 12 bit adc has lower resolution with the noise and low adc ref not 12bit reel! most 11 and worst case 10... 

Thank you for updating me!

Thierry

Thierry Pottier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

the 12 bit adc has lower resolution with the noise

I'm not sure I agree with that.

To a first order model, ignoring resistor noise and gate noise and other very low amplitude noise sources, anyway.

 

If you have a 0 to 5 V signal with 10 mV of noise and you were to simply attenuate the signal to 0 to 3 V, then the 10 mV noise would also be attenuated.

 

Noise is certainly a factor in low voltage analog circuits.

The power supply noise is important.

Just how close to the rails do rail-to-rail op-amps go?

What is Ground (0 V)? (That's a trick question for the Xmega!, It has two different Ground references, and a Ground Offset!)

 

If it is helpful to you, just do your normal analog front end signal processing design, and then scale the output on the last op-amp stage.

 

Don't get psych'd out by the lower voltage range.

 

JC

 

Edit:

Here is my last Xmega project, an EKG, typically 0 - 2 mV, with lots of noise.

The "trick" is in the analog front end signal processing, NOT in the specific micro used, or the ADC.

(This was an Xmega32E5 project)

 

 

 

 

Last Edited: Sun. Jun 5, 2016 - 03:37 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Dear DocJC.

well, I am a oll analog designer.. and the noise I mean was not from my electronic it was from Xmega32E5 himself.

I have move a design from avr 168A to Xmega32E5, it was to get quicker process + more analog + try Xmega too.

wnen I read the adc from AVR168 --> 0V analog signal give you perhaps 1 or 2 bit adc value. (avr power with 5V & extern rel 5V )

with Xmega32E5 same circuit (analog) value for 0V is 180- 190 adc (power 3V3 & 3V3 for ADC ref)

and yes not missing cap for power, analog, +++ 2 cap 100n & 10u + bead +++ , big poor gnd copper +++ 

 

2nd noise.

if you have 10mV analog side, yes it will give you same adc value if you have 3V or 5V ref because you reduce the signal.. 

 

By the way, prety design you make.

Thierry

Thierry Pottier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

power 3V3 & 3V3 for ADC ref

Remember that Xmega's AREF cannot go to VCC but to VCC-0.6Vmax or 2.7V max in this case.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thank you. 

I have just moving from AVR and was sure at they has same adc spec..

 

Thierry Pottier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It took me some doing to figure out the adc on an xmega. I keep what I learned here: http://barefootelectronics.com/x...

If you don't know my whole story, keep your mouth shut.

If you know my whole story, you're an accomplice. Keep your mouth shut. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

O V doesn't equal 0 Adc counts. The Xmegas have a built in 0 offset.
JC

Last Edited: Mon. Jun 6, 2016 - 07:04 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Torby. 

thanks usefull link...

 

DocJC.. 

ok. 

 

Thierry Pottier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
O V doesn't equal 0 Adc counts. The Xmegas have a built in 0 offset. JC

 

They call that a "feature." Seems like a bug to me 

If you don't know my whole story, keep your mouth shut.

If you know my whole story, you're an accomplice. Keep your mouth shut. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

From what I understand some of the ADC "features" like the VCC-0.7V for AREF is carried over from the UC3 range, don't know about the moronic ~200 count for 0V in.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DOCJC, where did you get your EKG AFE ( pads, etc ) from ( maybe you built it from scratch... ) ?

1) Studio 4.18 build 716 (SP3)
2) WinAvr 20100110
3) PN, all on Doze XP... For Now
A) Avr Dragon ver. 1
B) Avr MKII ISP, 2009 model
C) MKII JTAGICE ver. 1

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Indiana, what is an AFE?

The patient electrode cable is available from Sparks Fun. It is a three wire cable with "stantard" EKG patch snap on connectors. The other end is a stereo 1/8 in earphone plug, which one can cut off, if desired.

I have access to electrodes where I work, but one can buy them on line these days, also.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello Doc.

 

AFE = Analog Front End. Basically any circuitry, inc. cables/probes, leading up to the ADC input. Ok, I found the 3 wire/color cable set on Sparkfun, thank you.

 

 

Jerome

1) Studio 4.18 build 716 (SP3)
2) WinAvr 20100110
3) PN, all on Doze XP... For Now
A) Avr Dragon ver. 1
B) Avr MKII ISP, 2009 model
C) MKII JTAGICE ver. 1

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Oh man...  Should have known that.

I wear a lot of hats these days, and each has its own language and abbreviations.

Sometimes it just doesn't click.

 

Anyway, in this Thread I have a photo of my EKG project as a breadboarded project, using the Spark Fun breakout PCB for the AD8232 EKG chip, and another photo showing my PCB.

 

The AFE is basically that in the data sheet for a full bandwidth EKG monitor, slightly modified.

My front end has some additional protection circuitry, both for the patient and for the PCB, (although it isn't Defibrillator proof), and a couple op-amps after the EKG chip and before the Xmega ADC for further filtering and offset adjustment.  I am using the ADC in differential mode, so a very heavily filtered EKG signal is essentially the DC bias of the AC coupled EKG signal exiting the EKG chip.  This feeds the Neg input of the differential ADC.  The EKG signal feeds into the ADC positive input.  The EKG can then go above and below the bias voltage fed into the negative input, and the entire DC bias can wander up and down, and one can use the Xmega's ADC's internal gain for the ADC. 

 

I'm not at home this week, so I don't have access to the schematic right now.

 

Truly the AD8232 does all of the hard work!

My last EKG, using an instrumentation amp and a few other op-amps took a full circuit board just for the EKG signal processing before the ADC.

 

JC