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.

tsw1200evm/ads58c48 signal capture problems

Other Parts Discussed in Thread: ADS58C48

Hello forum,

I'm working with the TI ADS58c48EVM attached to TI TSW1200EVM to evaluate the ADS58c48's performance for a project.  I followed the ADS58c48EVM User's Guide, pgs 6-8, but I don't think the data isn't being captured properly. 

My setup:  Clock is 200MHz sine wave, 1Vptp @ sma clock input of ADS58c48 (J7).  Clock is being sent thru BPF, with Fo=200MHz, & Fc=20MHz.  Analog signal is 19.98291016MHz (using coherent freq. calculation in software) sine wave, 2vptp @ sma channel B (J6). Analog signal is sent thru BFP, with Fo=20MHz, & Fc=10MHz.  I'm using a benchtop power supply @ 6.125v to power both boards.  Jumpers have been set for this as well.  Clock and analog signal comes from an Textronix AFG3252 Arbitrary Function Generator.

TSW1200EVM is setup using LVDS DDR inputs, ADS sampling rate of 200M, ADC Input Freq. Fc is set to 9.98291016MHz, 16384 number of samples.

ADS58c48EVM is setup as per page 6 of user's guide. with no gain and no SNRBoost.

Please excuse the stupid questions, as I am a noobie to using your evaluation boards/software, and the documentation could have more info about the software.

Doesn't the time domain page show what is being captured by the adc, ie shouldn't it show the sine wave I'm sending as analog input? The single tone FFT never shows what I would expect, especially the AC results are way off specs in the data sheet.  I can never get the carrier amp to -1dbFS, even when I increase the amplitude of the signal.  Also ENOB is always very low, never gets up to 11 bits.

I've checked the voltage (power) to the board, and the cmv.  The cmv is approx. 955mv between the transformers and the ac coupling caps.  I can see with scope both signals at the inputs of the adc.  I've also tried the same setup with a duplicate set of boards I have with similar results.  Pictures will be at the end for your reference.

Could clock jitter be the cause of all this?  Clock jitter (RMS) for Tektronix function generator is spec'd at 100ps.  Are there any steps I can take to get this working?  We really want to be able to use the ADC for a project, but I need to prove it can perform to datasheet specs before we move forward.

Thanks in advance for all your help.  Pictures are below for your reference.

TSW1200EVM/ADS58C48EVM Setup

Time Domain page zoomed out.

Time Domain page zoomed in to see the signal.

Single Tone FFT page showing bad results from capture.

Scope showing 200Mhz clock signal, with FFT (at bottom) showing the large 200Mhz frequency component with other frequencies much lower.

  • Hi,

    quick question - have you used the pull down menu item in the TSW1200 GUI to select 2's complement option?  The default for most EVMs is offset binary but a few EVMs power up in 2's complement.  Inverting the most significant bit disturbs the FFT greatly.  Also - turn on the overlay unwrap waveform in the time domain.  If the setup is correct then this overlays a single period of a sine wave over the sample data, but if the 2's comp/offset bin is wrong then this sine wave is 'chopped in half' with the top half of the sine on the bottom and bottom half on the top.

    Regards,

    Richard P.

  • Richard, you truly are a genius!!  Thanks for your suggestion :-)  I needed to change the menu item in the TSW1200 GUI from offset binary to 2s compliment option and everything is working as expected!!!  I was struggling with this for a couple of weeks thinking it might be a clock jitter issue.  I should have came here much sooner.  This forum is a great service to all us engineers!!

    Thanks again Richard for the expertise and patience.

    Todd Moore