1
\$\begingroup\$

Assume that a system level test setup needs to be created for a design. This shall require use of power supply with multiple outputs, signal sources that can output sinusoidal or arbitrary signals and measurement equipment that can measure frequency, spectrum or magnitude of signals at different points in the device under test.

Such a setup could be controlled from the PC if all the equipment has ports that can be used to communicate with the PC e.g serial ports.

Q1. Is this how system level test setup is created? Are there software or hardware tools out there that can be used to simplify design and implementation of hardware tests?

A software program would be one that easily integrates with power supply, signal source and measurement equipment by containing drivers to communicate with each of them so that the user does not need to bother with low level details. A hardware solution would be some device that simplify creation of such tests.

By system I mean my PCB with FPGA, ADC, DAC, Flash, RAM, Serial Port e.t.c. The purpose is to test all the functions of the FPGA board.

Only for prototype.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ The answer depends on your definition of 'system'. You should have or define your system inputs and outputs, and use the relevant equipment to generate\measure each of these. There can be high-end test equipment for special\high precision measurements and there can be custom made equipment. Usually control of equipment is done by some sort a script that operates the equipment and saves the results \$\endgroup\$ – ZelmaB Apr 21 at 11:39
  • 1
    \$\begingroup\$ GPIB devices are still sometimes used for automated testing. These are very slowly being superseded by Ethernet in 2021. \$\endgroup\$ – rdtsc Apr 21 at 12:38
1
\$\begingroup\$

There are many approaches. Depends on volume and the type of technology involved, also whether 100% coverage and long term reliability are of high importance. In some markets (a cheap gadget for teh mass market) a certain level of field failures can be deemed acceptable, it's a trade off between replacement value and multiple other factors. If you're making an altimeter for a jet fighter, obviously this is NOT the approach.

For low volumes, semi-custom work, a manual test setup where a test tech twiddles knobs and watches a scope, meter, etc can be very effective (if you have the test techs available).

Bed of nails is OK but doesn't in some cases test to the external connectors of the box.

Also flying probes testing, self test routines built into firmware - it really depends on specific scenarios.

\$\endgroup\$
2
\$\begingroup\$

Many modern bench instruments come standard with LXI these days, which uses Ethernet. Before that the standard was GPIB IEEE-1488.

One way to control disparate test instruments is with Labview. There are other ways.

Last time I did a temporary test setup for radiation testing I used some Python programs.

\$\endgroup\$
1
\$\begingroup\$

Bed of Nails testers. Worked at an assembly factory and we used those. There were some customers, Dell for example, wanted us to plug in a processor and ram and run a dos program to test hardware back in the late 90s.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.