Advertisement

Analog Angle Blog

Virtual Eval Through the Eyes of an ADC

As I mentioned in my last blog I would like to take some time in this blog to introduce you to the new online simulator tool offered by Analog Devices called Virtual Eval. Of course, as usual, I’ll give my insights on the tool from the perspective of an ADC although the tool has been built to support products beyond high speed ADCs. For a moment let’s hop into the famous ‘way-back machine’ and recall a few blog posts that I did early last year discussing ADIsimADC. You can find those at Analog Devices Design Tools: ADISimADC Web-Based Simulation Tool, and at Analog Devices Design Tools: ADISimADC Performance Metrics, and also at Analog Devices Design Tools: ADISimADC Frequency Folding Tool . Interestingly I finished up that series of blogs on ADIsimADC just before IMS2015 last year and here we sit just a little after IMS2016 this year. But I digress, let’s get back to discussing Virtual Eval. I mention these previous blog posts because the foundation for the ADC simulation portion of Virtual Eval has its basis in ADIsimADC, but is much more powerful and refined tool. I’ll be focusing on the simulation tool from an ADC perspective; for more on Virtual Eval make sure to check out an in depth look over on Analog Dialogue in the article An Engineering Walk Through Virtual Eval, ADI’s Online Data Converter Product Evaluation Tool. Before we dive into using Virtual Eval for the AD9680 we will first take a look at an example with an ADC that is a bit simpler.

Let’s start with where we can find Virtual Eval. Currently it exists in its beta form at Virtual Eval. When you click this link you’ll notice that the first screen that opens up prompts the user for a category selection. These categories include: All Converters, High Speed ADCs > 10MSPS, Precision ADCs < 10 MSPS, High Speed DACs > 12 MSPS, and Integrated / Special Purpose Converters. For our purposes here I have selected High Speed ADCs > 10 MSPS. For this introduction we will use the AD9643 as an example since it is an established and relatively simple high speed ADC product offering.

Product Category Selection Page in Virtual Eval

Product Category Selection Page in Virtual Eval

Once the product is selected it opens to the main view for the AD9643. The product settings for the AD9643 are shown on the left of the screen and the default view selection is set to show the AD9643 FFT results (for some products a block diagram view may also be available). There are also two links on the bottom right of the screen. One is to allow the user to quickly access and download the data sheet for the AD9643. The other is to allow the user to order the evaluation board for the AD9643 for the opportunity to use actual hardware and compare simulation results with measured results if one so desires. Last but certainly not least is a selection for help next to the view selection. This directs the user to a video tutorial on operating the Virtual Eval simulator tool.

Main View for the AD9643 in Virtual Eval

Main View for the AD9643 in Virtual Eval

Now let’s run a simulation and take a closer look at the results. In this example I have selected an input frequency that matches one of the listings in the AD9643 AC specification table. If you refer to the data sheet you will find that the following performance specifications are given for the AD9643-250.

AD9643 Typical Performance Specifications at 185 MHz Input Frequency

AD9643 Typical Performance Specifications at 185 MHz Input Frequency

As you can see from the simulation plots (note the two plots showing noise and distortion performance) in Virtual Eval the performance given by the tool is very close to the specified values in the data sheet. The SNR given in Virtual Eval is within 0.2 dBFS of the actual measured value. The SFDR (which is dominated by the 2nd and/or 3rd harmonics) is within 2 dB of the measured typical value given by the data sheet. The one difference comes in with the worst other. In the simulation the fourth harmonic is shown as the worst other harmonics or spur at 102 dBc while the worst other typical is 92 dBc in the data sheet. Once levels of harmonics drop below 90 dBc it becomes harder to measure as well as harder to simulate. Since the 2nd and 3rd harmonics typically dominate the spurious performance either of these levels is pretty insignificant. Even in the case where the 2nd and 3rd harmonic are out of the band of interest, spurious levels below 85-90 dBc are generally sufficient for most applications.

FFT Results View for the AD9643 in Virtual Eval

FFT Results View for the AD9643 in Virtual Eval

FFT Results View (Distortion Performance) for the AD9643 in Virtual Eval

FFT Results View (Distortion Performance) for the AD9643 in Virtual Eval

The last thing we will take a look at is the ability of the Virtual Eval simulator tool to include the effects of external input jitter on the ADC input clock. The default setting is for 60 fs of jitter which is pretty low. This is on par with the level of jitter expected from a signal generator such as the Rohde-Schwarz SMA100 with the low phase noise option. In a real system however, jitter levels can be nominally higher (loosely you can say phase noise since one can be calculated from the other and are essentially two different sides of the same coin). If we assume we are using a device such as the AD9523 to clock the AD9643-250 then we can approximate the jitter as about 200 ps. If we put that into the Virtual Eval tool and run the simulation we now see that the SNR is 69.135 dBFS.

FFT Results View (Jitter Impact) for the AD9643 in Virtual Eval

FFT Results View (Jitter Impact) for the AD9643 in Virtual Eval

Let’s again use the ‘way-back machine’ and go back to April of 2014 in my blog ADC Noise: The Clock Input & Phase Noise, Part 3 – Test Setup and find that under similar conditions with the AD9523 providing the clock to the AD9643-250 we had an measured SNR of 68.848 dBFS at 140.1 MHz. How cool is that? Our Virtual Eval tool came pretty close to predicting the actual value of the SNR when approximating the effect of external input jitter of the AD9523 on the AD9643 noise performance. From the examples provided here you can see that Virtual Eval can be used to provide a pretty accurate prediction of the real performance of an ADC. In the case where budgets are tight it really makes Virtual Eval an attractive tool to help select a product such as a high speed ADC. Stay tuned as we continue looking at Virtual Eval and expand to a more complex example. We will look once again to the AD9680 and show how Virtual Eval can be used to predict the performance when using the DDCs.

2 comments on “Virtual Eval Through the Eyes of an ADC

  1. Cillyaussem
    June 22, 2016

    nice post dis

  2. jonharris0
    June 22, 2016

    Thank you Cillyaussem!  I hope you find it useful.  The Virtual Eval tool is a very cool and useful tool!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.