BY Walter Shawlee 2
Originally published in Avionics Magazine (September 2000), reprinted by permission.
Testing for safety in an aircraft environment is not such a simple thing. Sometimes the tests are a whole adventure in themselves, taking longer to implement and run that the design may have taken. Anyone who has ever tried to do the DO-160D DC power supply tests for transient, audio and over/under voltage categories (Sections 16-18) has probably experienced the pain and despair of trying to do these tests effectively and in a repeatable manner. This is especially true during the design and test phase, when it is often useful to have this capability right at the work bench to get the gremlins driven out of the circuitry at an early stage, while the system is still easily amenable to change at low cost.
The original transient and audio interference technique is essentially passing DC through a very large non-saturating transformer, and then driving the secondary winding of the transformer with sufficient signal to induce AC ripple and audio interference that has the desired wave shape. This sounds simple, but is much harder to do than it appears, especially for fast transients, which involve some fairly scary relay driven pulse networks. The massive coupling transformers and other paraphernalia illustrated in the RTCA (Radio Technical Commission for Aeronautics) DO-160D specification are fairly inconvenient and cumbersome for widespread use in a design department, unless you favor the Castle Frankenstein look. The illustrations are very useful to understand the test itself, but cry out for more modern techniques at test time.
Study of the test routines shows that they are really steady state DC levels, with specific waveforms, ramps or sequences applied, and could certainly be implemented that way. Many power supplies have external programming capability, and with a bit of thought, some can be made to do the test sequences. Things start to go badly with this scheme when fast transients and (egad) polarity reversals enter the tests. There is a device that actually does all of this very cleverly, and can be a huge benefit both for qualification testing and development. Conceptually, a high powered audio amplifier could do this, but most are AC coupled, and can’t support the steady DC levels required, or the ugly duty cycles.
The hybrid device that can do it can be thought of as a massive Operational Amplifier, with high power DC capability. It has existed for a long time from both HP and Kepco, and is a useful marriage of a power supply, adjustable DC reference and fast amplifier. HP’s units were lower power (100Wor less.), and were just called Power Supply/Amplifiers (very catchy name, no doubt marketing spent weeks on it). If you are lucky enough to find one of these elusive and uncommon devices, their part numbers are HP6824A through 6827A, and they can deliver BIPOLAR power under external control, meaning they can swing both positive and negative, a mandatory requirement for the DO-160D test sequence. Needless to say, these very useful devices are now discontinued from HP, now Agilent. Attempting to generate the sequences via the newer technique using IEEE-488/HPIB bus is not always so successful, due to damped system responses, and lack of bipolar capability in the available supplies.
Fortunately, the real test killers come from power supply maker Kepco, and are called BOP’s, Bipolar Operational Power Supplies. These (still available) monsters can supply very high power levels. This is the ideal weapon of choice for all around abusive power testing, and I found that the BOP 100-4M (100V/4A) works for most small applications, and has been my trusty companion for many years. This BOP can supply up to 100V at 4A, which comfortably handles most transients in the basic DO-160D schedule (category B, +78/-22V) with ease. It is especially useful for testing to destruction, when you need the ability to see exactly how much safety margin you have, and want to alter parameters on the fly to explore this region carefully. This is a horrifying exercise with the older pulse transfer method, especially in terms of repeatability, because of the technique’s dependence on uncontrollable characteristics of the power supply and test item.
I use my BOP with a companion Tektronix TM506, stuffed with two DVMs, a pulse generator, sine wave generator and dual channel scope. This is hooked together so that I can see exactly what my steady state DC is, read my RMS AC component, and observe the waveform while exciting the system with either pulses or sine waves. What incredible fun if you are inclined this way. It is also hugely instructive for many designers and engineers as a way to see how fragile and imaginary their “input protection” circuit really was. With a few adjustments, you can be pounding in pulses, huge AC ripple, or anything else you want, all easily controlled and recorded, and most importantly, duplicated later. Here’s a picture of my own BOP station, happily pumping out massive AC on top of a normal DC bus. If you marry up an arbitrary waveform generator to the BOP, you can literally pre-program the entire ramp/pulse and interference sequence and just hit a few keys to make it all unfold.
Avionics systems really need this type of testing, especially those with an audio aspect, as power supply induced audio interference can be ruinous to many otherwise attractive circuits, and be full of unacceptable noise and interference. Digital circuits may also experience upset or incorrect power up/down sequencing on the transient and power interruption tests, and may prove completely unstable. The slow power up sequence is especially important for neophyte designers, whose switching power circuits may fail catastrophically, or cause system over-currents that trip the feed breakers. Suffice it to say, one can learn a lot in just a few minutes, and explain many of those “unexplained” system problems being reported from the field. Problems you were absolutely sure were impossible until just a few minutes ago.
Using the BOP is simple, just dial in the steady state voltage you want, and use the voltage “Input” to send in your pulse, transient or audio waveform to modulate that steady state. The input responds to either polarity, and is DC coupled, so very slow ramps, or other slow shifts are fine. You can also set current limiting, or if your application requires it, current modulate the steady state output. There is a lot of flexibility here, and a wide range of models up to 400 Watts is available. To learn more about BOPPING, visit this web site: https://www.kepcopower.com/bop.htm
Because of the very good frequency and high power capability of the BOP, you can also use it to drive a shaker table, or use it to exercise massive loads like heaters or motors, and have it act as a general purpose buffer for anything up to about 10Khz. There is also a related item, called an OPS (Operational Power Supply) which can go to 5,000 volts, and can be very handy for generating calibrated static pules, and other intentional damage signals. I just picked up a 1KV unit surplus for $100 to add to my arsenel.
In any case, if you are like me, you may not dance all that well, but anybody can learn to BOP. All for now.
You can contact Walter at firstname.lastname@example.org or visit http://www.sphere.bc.ca.