• State TI Thinks Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 1249 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADC1175: Jitter problem when sampling sinewave and squarewave

Baris Erdogan
Prodigy 180 points

Part Number: ADC1175

Hello,

I have a project that continuously sampling ADC In pin on ADC1175 20Mhz but facing a lot of jitters on waveform. I thought problem may be regarding bare wire coming from microchip's clock and connecting ADC's clock pin. And decided to make more investigation and used Inverter before ADC's clock pin by suggestion of Customer Support. But jitters did not go even got worse. Number of jitters increased. 

I attached the screenshots (case.rar) when i used clock signal from microchip without any logic IC (like inverter) and with inverter. How can i get jitters reduced?

Thank you

Regards 

Baris ErdoganCase.rar

  • 0
    TI__Genius 12780 points

    Hi Baris,

    Is this on a custom PCB, or is this being done on a breadboard/protoboard? Can you please share your ADC schematic? Is the ADC enabled and sampling when you are taking these scope shots? Can you please take a scope shot of the clock input (20 MHz) when the ADC is sampling, and also when the ADC is disabled?

    Best Regards,

    Dan

  • 0 in reply to dBrock
    Prodigy 180 points

    Hello Dan,

    please see answers below,

    Is this on a custom PCB, or is this being done on a breadboard/protoboard?

    protoboard

    Can you please share your ADC schematic?

    Please see attachment

    Is the ADC enabled and sampling when you are taking these scope shots?

    ADC was enabled and constantly sampling when these screenshots taken from my waveform analyzer application running on windows 7

    Can you please take a scope shot of the clock input (20 MHz) when the ADC is sampling, and also when the ADC is disabled?

    I have a scope has only 20Mhz bandwidth, taken shots attached but not meaningfull

     

    I also attached ADC Vin pin Scope screenshots revealing that pin is generating a sparks. may be a capacitor issue on one of the analog power pins is main reason or comparator in ADC1175 IC itself.

    I forgot to tell about choke merging the Analog power and Digital power pins, I used inductance of 4.7uH for that

     

    Thank you

    Regards

     

    Baris ErdoganCase_2.rar

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    One thing I see from the schematic, is the VRT is connected to 5V and VRB is connected to ground. This is too large of a sampling window for this device (2.8V max difference between VRT and VRB), so you should change either VRT or VRB to be in compliance. I'm not certain what kind of errors you could see due to this violation.

    What does your sample data look like? Are you able to get codes or FFT? Just double checking, but the sample clock (pin 12) is 20 MHz, what is the analog input frequency (pin 19)?

    Best Regards,

    Dan

  • 0 in reply to dBrock
    Prodigy 180 points

    Hello Dan,

    VRT and VRTS connected each other and two of them connected to 5V. I missed updating that point. So, as ADC1175 datasheet mentions in Figure 19.

    I measure VRT approx 2.09 V, VRB is very close to 0 V.

    Sample data looks like what is in Before_Inverter_1.png file (and _2, _3 respectively) in "case.rar" attachment. 

    I also took sample with scope from ADC Vin and run FFT and attached "case_3.rar" .

    "FFT_1.png - FFT_3.png" when sine waveform triggered, FFT_4 and FFT_5 when spike triggered when test signal is "878 Hz Sine wave" Scope measures frequency 27-28 hz on ADC Vin pin i think because scope spoils oscillation.

    Thank you

    Regards

    Baris ErdoganCase_3.rar    

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    I am trying to piece together the files that you have shared, but I am not sure what problem I am trying to address at this point.

    The FFT_1-5 png files look the same to me. Are you referring to the spike I circled in the pic below?

    I had asked earlier about if you were using a protoboard since noise can easily be coupled from a "noisy" ground plane. This is especially true with a microcontroller on the same ground. Can you powerdown the microcontroller/oscillator and see if you are still noticing this "glitch"?

    If I am not addressing the problem you are trying to solve, can you please redirect me?

    Best Regards,

    Dan

  • 0 in reply to dBrock
    Prodigy 180 points

    Helo Dan,

    Yes exacly the glitch that you circled is my issue. Since i could not send you a video of the oscilloscope screen, i tried to explain the problem by sending consecutive shots of the case. The microcontroller is which is sending the sample data to my computer and waveform analyzer app via usb.

    Let me take apart two grounds and then take shot again.

    Should i merge microcontroler's ground and adc1175's digital ground?

    Regards

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    Yes, please give this a try.

    Just so I understand, the microcontroller is sending the ADC1175s digital data to your PC which is then running an oscilloscope application to view the wave form? You are not able to view the analog signal on the vin pin?

    Best Regards,

    Dan 

  • 0 in reply to dBrock
    Prodigy 180 points

    Hello Dan,

    I am able to measure adc1175 Vin pin by my waveform application and my oscilloscope as well. But both of them show same glitch when i trigger the waveform. I need to remove glitch from Adc1175's Vin pin. Glitch is sourced by my poor pcb design or adc1175 itself.

    The shots that have green colored background is from my waveform application which is being filled  by adc1175 and and usb connection(via microcontroller)  to the computer's usb buffer

    Thank you

    Baris Erdogan

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    Were you able to resolve the glitch issue? Did separating the planes help?

    Best Regards,

    Dan

  • 0 in reply to dBrock
    Prodigy 180 points

    Hello Dan,

    New shots attached, I removed all shaping networks(resistors, OPAMPs, Caps etc) before ADC's Vin pin and, I applied sine wave directly to Vin pin  with 20 Hz and 406 Hz frequency. Seperating grounds after and before. Strange thing is that after seperating grounds it seems to get worse to me. 

    Is it OK, although I seperate ADC's Analog Grounds and  Microcontroller's Grounds, but ADC's Digital Grounds and Microcontroller's Grounds same, when i test cable connectivity of Microcontroller's Ground and ADC's Analog Ground, and i get beep signal, as if all grounds already connected each other?

    I also tried to reduce clock frequency by dividing by 2 with Flip-Flop (10 Mhz) and applied to clock pin but nothing has changed.

    It feels like when ADC is sampling, trying to get over the potential barier and get over, and after very short time  it has to get over another potential barier and get over and this behaviour produces that spikes.What about you?

    Thank you

    Baris Erdogan

    Case_4.rar

  • 0 in reply to Baris Erdogan
    Prodigy 180 points

    Hello Dan,

    Did you ever have chance to investigate my update and my new shots?

    Thank you

    Baris Erdogan

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    Thank you for your patience as I acquired all necessary items to perform tests in my lab. Using the ADC1175EVM, we can better isolate the ADC1175 to understand how it behaves in a controlled system.

    Please see the attached document where I test three different analog inputs with a 20 MHz onboard oscillator serving as the sample clock.

    ADC1175EVM_analog input test.pptx

    Unfortunately, there is not much else I can do to help track down the exact issue with your board. My suggestion for you is to order the ADC1175EVM, and perform your own tests to better understand how the device will behave in your design. Or, you can look at how the EVM is designed, and copy this into your project.

    Best Regards.

    Dan

  • 0 in reply to dBrock
    Prodigy 180 points

    Hello,

    Does "the place you highlighted with yellow" mean that you took them off from the board during test?

    Are these graphs produced by pure sample data or does your application (HSDC Pro) take average of the sample data?

    Thank you

    Regards

    Baris Erdogan 

  • 0 in reply to Baris Erdogan
    TI__Genius 12780 points

    Hi Baris,

    The highlighted area in the upper left corner is where I applied the analog input sine wave and the capacitor next to it is where I took the oscilloscope measurement for 1 MHz sinewave input.

    The other highlighted area near the middle is the 20 MHz crystal oscillator.

    This is not averaged data, but just a single 65,536 capture.

    You can view the ADC1175EVM design files here http://www.ti.com/tool/adc1175evm 

    Best Regards,

    Dan