ADC and noise in electronic systems

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Hi all,
first let me introduce myself.
I am a French student working for my internship on noise in electronic circuits.
I have to redesign an ADC because it is perturbated by noise. A lot of filters are used on the existing circuit, but some strange things happen and keeo fuzzy in my mind.
2 ADCs are used with the same input. The good signal is included but hidden because of the noise.
The first thing is: on the figure attached, what is this curve at the beginnig? it looks like a capacitor discharge. How to solve this problem?
Secondly, I get an other curve coming later, and shifting over the time. I do not understand at all the origin of this problem.
Do you have any idea? The best would be to get a horizontal line...
In addition, what kind of general advice you can give me concerning noise in electronic systems? grounding, type of components to use...

Lastly, I will incorporate a digital circuit into the already existing analog one for the data acquisition. What kind of precautions do I have to take for the combination of digital/analogic circuits together?
Thank you very much for your help,
Regards,
Arnaud
 

Attachments

DickCappels

Joined Aug 21, 2008
10,153
Trying to be a mind reader:

Most likely the curve at the beginning is the charging of the input capacitance of the A/D converter through the resistance of the analog multiplexer.

The one that you say is "shifting all the time", if that's the wiggling of which we can see about four cycles, that could be feedthrough from the digital portion of the converter.

In general, you will find that including a schematic along with a thorough description of the circuit and problem will help you get the best possible help. Please post the schematic.

As far as how to deal with noise and grounding, try these articles:

http://www.analog.com/library/analogDialogue/archives/40-02/adc_noise.html

http://www.hit.bme.hu/~papay/edu/Acrobat/GndADCs.pdf
 

MrChips

Joined Oct 2, 2009
30,706
You have not provided enough information.
What is missing are some real numbers.
What is the scale on the x-axis?
What is the amplitude and frequency of the signal?
What is the amplitude and frequency of the noise?
How many bits on the ADC?
What is the voltage range of the input of the ADC?
What is the conversion time of the ADC?
What is your sampling frequency?

Numbers matter a lot!
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Thank you for your fast answer. I provide here the preamplifier schematic.
Just after this preamplifier, I have 2 stages of gain amplifier with SSM2019 components.
To power supply the VJFET and the LT1115, LT1761 regulators are used.
 

Attachments

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Thank you MrChips for your answer. Sorry for my poor details. The curves are not really representative as it is used with Labview. The time is in seconds. The resolution for the acquisition board is 14. The sampling is 524288.
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
What do you think about my last question about digital and analog circuits on the same PCB? which particular precautions do I have to take?
 

t06afre

Joined May 11, 2009
5,934
What is you current sampling speed? And which ADC board do you use. Why is noise a problem here. Can you elaborate on this, like do you have any idea of the noise source. Do you use any digital filter functions in Labview also
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
My sampling rate is: 4210526.
The preamplifier is a LT1115 and the gain amplifier is a SSM2019.
The board is developped by the company in which I am working.
 

t06afre

Joined May 11, 2009
5,934
Is your sampling rate really 4210526 samples pr second or is 4210526 the total numer of samples. The sample rate 4210526 pr second was kind of odd
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Thank you for your fast answer.
Unfortunately Icannot take any picrures in the company. I can provide a photo of the board. On this PCB are the power supply regulated by LT1761 and filtered by low pass filters, a ULNA made thanks to a LT1115 op amp, and 2 stages of gain amplifiers using SSM2019.
The acquisition board is a EVAL-AD7357.Untitled1.gif
 

t06afre

Joined May 11, 2009
5,934
Have you tested the board connecting a fresh and new 1.5 battery to the ADC input. Use as short lead as possible. A battery is a very low noise source. By doing this you can say something about the background noise from the ADC system itself
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Very good suggestion but it is mandatory in the project to use a power supply. Just to test and obtain results, I will try to experiment that. Thanks a lot for your help.
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
For technical reasons, it is difficult to connect this system on batteries. Indeed, I need +15V and -15V and I cannot just reverse the + and - to get the 15V, can I?
I provide here the actual low power supply schematic.powersupply.gif
 

t06afre

Joined May 11, 2009
5,934
Just start with connecting a 1.5 volt battery to the two ADC input channels. If you want to take the experiment further. You can use two 9 volt batteries to create 18 volt battery. Then put those together like shown here to create +/- 18 volt.
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
Thanks for your help. When you say to connect a 1.5V battery to the ADC, it means that I connect this battery on the connector X9 of my preamplifier (provided on the previous page).
I agree with you for the -/+ 18 V schematic. I am trying to find what I need.
 

t06afre

Joined May 11, 2009
5,934
Thanks for your help. When you say to connect a 1.5V battery to the ADC, it means that I connect this battery on the connector X9 of my preamplifier (provided on the previous page).
I agree with you for the -/+ 18 V schematic. I am trying to find what I need.
I see now that your pre-amp is AC coupled. So a battery connected here will not give much info. But the inputs on the AD7357 Evaluation Board are DC coupled so it will make some sense to connect a 1.5 volt battery here. Just to see to get a picture of the noise
 

Thread Starter

arnaud405

Joined Oct 5, 2012
16
I will do that tomorrow and I let you know. Here is the schematic of the evaluation board with the ADC. On my board, the capacitors C1, C2, C3 and C4 are not present. That could explain the reason why I have this curve at the beginning...I will try to solder them.Untitled.png
 
Top