• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

USB data acquisition

Status
Not open for further replies.

Flamer1

New Member
hello, it is a project i'm building that will be a digital data acquisition system interfaced with a pc through usb. any help at this moment will do a lot. i'm currently using pic microcontrollers, a 16-bit a/d converter, instrumentation amplifier. i'm getting a lot of noise issues, and bandwidth is horrible. not to mention, it won't re-create sinusoids

:D

thanks
 

Optikon

New Member
Flamer1 said:
hello, it is a project i'm building that will be a digital data acquisition system interfaced with a pc through usb. any help at this moment will do a lot. i'm currently using pic microcontrollers, a 16-bit a/d converter, instrumentation amplifier. i'm getting a lot of noise issues, and bandwidth is horrible. not to mention, it won't re-create sinusoids

:D

thanks
I'll throw something out here.. Do you have anti-aliasing filter on front end A/D input? This could be part of your noise problem as well as part of the inability to re-create sinusoids (what do you mean by that? they are just too distorted?)

A 16-bit A/D system is not a trivial design if you are to achieve true 16-bit performance. To keep your noise floor down, you must take great care in the parts you choose and your layout.

If you could post a schematic of your front-end (analog) with part numbers, it could be helpful. Did you do noise calculations on your A/D and INAMP? You may just have too much noise because your parts are giving you too much noise, but my experience is that it is usually a multitude of reasons.

Care to share more details?
 

crust

Member
What are your sampling parameters for your sinusoid test signal (i.e. amplitude, reference voltage, sinusoid frequency, sampling frequency)?

Also 16bits is going to be around 96dB of dynamic range, you are going to have a noise floor in most cases. Your choice of analog front-end and layout will affect the noise floor.
 

Flamer1

New Member
i had 2 separate designs so far for the front-end. one of the design considerations was i needed variable gain to accomodate different voltage input ranges from 0-150V. my first design used a simple voltage divider to divide the input by 30, then the variable inamp would raise it to a +/-5V reference. that way, i also get a 10M input impedance which i was looking forward to. the inamp was made with TC913 op-amps. the second design had variable division voltage divider at the front, and a fixed gain=1 inamp, the INA217. I'm not familiar with noise analysis calculations, could you give me some insight to that? Currently, the prototype and testing is being done on breadboard with decoupling of 10u electrolytic with .1u mylar at each power supply. what is the purpose of the anti-aliasing filter? let me know if you need to know more about the parts chosen or values

:D
 

Optikon

New Member
Flamer1 said:
i had 2 separate designs so far for the front-end. one of the design considerations was i needed variable gain to accomodate different voltage input ranges from 0-150V. my first design used a simple voltage divider to divide the input by 30, then the variable inamp would raise it to a +/-5V reference. that way, i also get a 10M input impedance which i was looking forward to. the inamp was made with TC913 op-amps. the second design had variable division voltage divider at the front, and a fixed gain=1 inamp, the INA217. I'm not familiar with noise analysis calculations, could you give me some insight to that? Currently, the prototype and testing is being done on breadboard with decoupling of 10u electrolytic with .1u mylar at each power supply. what is the purpose of the anti-aliasing filter? let me know if you need to know more about the parts chosen or values

:D
What is the frequency response range you wish to cover? In otherwords, what is the highest frequency sinusoid you wish to reconstruct without distortion? What is your sample rate? The anti-aliasing filter will reject frequencies beyond this input range. Why did you bypass the way you did?
I'm not sure about the mylar caps, you really need ceramic types. Also, if your parts are close together (within 2 or so inches), just one 10uF can be suitable for all the parts (the electrolytics are only good to a couple MHz)
How are your grounds laid out? You might have digital noise getting injected into your analog grounds/supplies.

Take a look at the noise on your supply lines (right at the IC pins) on your analog IC's? Get an idea of the frequencies present. Then look at the PSRR of your ICs (inamp, opamps, A/D etc..).. then compare that to and LSB of 16bit.. if you noise is more than that, you are likely going to have to find better parts, change your layout, clean up your supplies and grounds.
 

naskog

New Member
info

I have to build about the same project for 0 to 50 HZ
send me your circuit diagram and i will try to help you
 

Flamer1

New Member
i would like the design to have a bandwidth greater than 1kHz, preferably ~5kHz, but the A/D should be able to handle as much as 20kHz. so the maximum frequency i would like to reconstruct is ~2.5kHz

naskog: what are the specs of your project?

oh, by the way, what's a free schematic program that is easy to make custom ic's. i use pspice 9.1, but's it's a real pain labeling each pin and net and everything
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top