1. How does the Intel script differ from Polar's Atlas solution?

The script is just that — a script that will need to be programmed into an environment, or used for manually post processing the data after a measurement is made. Polar's solution is mathematically the same*, but fully incorporated into an easy to use environment combining both impedance test and SET2DIL test. Like CITS the Atlas delivers accurate measurements in an easy to operate environment which can be set up and used by non electronics technicians and operated by production line operators.

The Intel script is based on MathCad — the Polar solution was completely written in C++ by Polar — it does not need or use MathCad or a MathCad runtime environment. Functionally the Atlas SET2DIL software is designed to be identical to the Intel script.

Intel's script post processes csv data files that are output from the DSA8200. It does not set up or drive the TDR to obtain those files. The user is required to set up the TDR channels with the appropriate TDR / TDT setup and all of the triggering, timebase and VA settings. The TDR has to be driven outside of the script i.e. manually or by other means to acquire the csv files. The output from the Intel script is another csv file that has to be imported into, say, Excel to graph and interpret the data (if that is required).

*The Atlas also adds a test that monitors system bandwidth to ensure that the cables probes and TDR are suitably capable of measurement at the required frequency. This gives added confidence to the end user.

2. How soon can we deliver a probe set?

4 weeks after receipt of order. i.e. if a prospect had placed an order at the beta release time they would be receiving them in under 2 weeks. We do have initial quantities in stock now, but time should be allowed for shipping, and dependent on how many get ordered we would still project delivery at 4 weeks. However the probes ARE available to order.

3. How do we define the system bandwidth? Is it -3dB?

The system mathematically extracts the frequency domain information from the time domain response, the upper cutoff frequency is where the Polar math is unable to resolve the higher harmonic information from the noise floor of the system. The DSA8200, and Polar supplied Atlas probe and cables have a bandwidth of 20GHz; the critical determinator of bandwidth will be the design of coupon that you receive, either from your OEM or from the CGen Plus coupon Generator. We are fine tuning CGen plus at the moment before release. Initial coupon bandwidth is the responsibility of the original coupon design authority.

The system bandwidth reported in the Atlas loss display is essentially the point at which the time domain -> frequency domain FFT breaks down into 'noise'. In itself the FFT is just math — it will cheerfully crunch numbers out to thousands of GHz even if its input does not contain the requisite data to justify such. At some point due to the board geometry and other factors the stage is reached where the system is dividing noise by noise and generating — noise.

It is not the -3db point; it is an indication of where the combined system of TDR, cable, probe, launch into the DUT and DUT itself degenerate into producing data that can no longer be discriminated from the noise floor.

4. What are typical test times for SET2DIL measurements?

Under 2 seconds.

5. How does Atlas communicate with the TDR?

Atlas communicates with a compatible measurement system (TDR) via a TCP/IP Ethernet connection; a static IP address is required for the measurement system.