In the late 1960s, the LNAs (low noise amplifiers) used on satellite earth stations were very complicated. The dish was usually around 30m across, and they focused the signals onto a small sub-reflector on a tripod and then down a horn to the LNA cabin at the back of the dish. Three stages of parametric amplifier and a tunnel diode gave a total front-end gain of around 40dB with bandwidth 500Mhz and center frequency around 4Ghz. To achieve a decent noise level, these stages were packaged into a stainless-steel Dewar (vacuum flask) and cooled to around 15K by a cryogenic refrigerator using helium as the working fluid.
The LNAs were duplicated so that one could work while the other was serviced. The cryogenics needed regular overhauls to maintain the low temperature. Even a few minutes without power would cause a warm up and take several hours to sort out. The LNA team members on one station in the Arabian Gulf were mortified to occasionally see regular bursts of unwanted pulses on the LNA output along with the desired signals. Worst of all, the pulses came and went unpredictably, popping up and down on the spectrum analyzer, yet always on the same frequency. Complete and time-consuming overhauls of both LNA systems achieved nothing. It could never go on traffic until the problem was solved.
One of the team members had previously worked in avionics and suggested that perhaps the bursts of pulses might originate from a radar station. It didn't seem likely -- the dish was pointing up into the sky and looking across the Gulf. But this was the best hypothesis we had, so we looked into it. We quickly found that the nearby city airport radar worked on 2GHz, exactly half the observed frequency, and the repetition rate of the bursts of pulses did correspond to the rotation rate of the scanner. Yet, how were the signals getting into the system when the dish was pointing in the opposite direction? And why did the spurious signals come and go so unpredictably?
As we stood at the back of the dish and looked toward the airport, some 10 miles away, the penny began to drop. There was a steel sliding door on the LNA cabin. When it was slid open, the pulses appeared. Clearly, the open door allowed the radar pulses into the cabin. Yet, the sensitive gear was inside a stainless-steel flask, and everything was connected up with robust waveguide, so it still didn't make sense.
Finally, the LNA team stripped down every single piece of waveguide and found one flange joint that had been temporarily assembled without a soft metal gasket, and the bolts just finger tight. Obviously, the two mated pieces had then been picked up by someone else and bolted into place without further inspection. We fitted a gasket and tightened the bolts to the correct torque. That solved the problem. Many hours had been wasted by a simple oversight. Although all the drawings were meticulously checked and had specified the correct number of gaskets, bolts, nuts, washers etc., the policy was to "free-issue" bulk packs of such items rather than count them out as complete kits.
Given a complete kit, the extra leftover gasket might have given cause for concern and prevented the problem from occurring. The flipside of that policy is that if a gasket gets spoiled or a special nut is dropped into the desert sand, the job stops until a new one is sent from the UK. This is just one of the joys of working on an installation crew far from home. Of course, in aeronautical, aerospace, and nuclear engineering, to name but three, such strict controls on parts issued are routine and the cost of the supporting paperwork is an integral part of the safety culture.
This entry was submitted by Rod Hine and edited by Rob Spiegel.
Rod Hine, MA (Cantab) MIET MBCS, graduated from Churchill College in Cambridge, England. He has worked in satellite communications, meteorological telecomms, general automation, machine tools, and industrial control systems.
Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.
In aerospace applications, the cost of part controls is not part of the safety culture, but rather the lesser of two cost factors. It costs less to verify each part is installed, down to the fastener level, because the cost of not doing it and losing the satellite is higher. This is true regardless of manned or unmanned missions.
This expresses, also, the danger in JIT planning. One mishap and you are burned! But how many times have we put something together only to find extra parts that were either put in by mistake or added to be sure we had extras if something went wrong? Extra screws are wonderful, unless we left the screw off the brake on the bicycle we just put together. Was the waveguide not put together properly because someone assumed there were extra parts? Did someone think close was good enough? These things can cost a lot of time and money. Proper training and supervision could have prevented so many heartaches in our field.
I have found that many times the liberal application of a torque wrench or a little but of alignment can solve a wide variety of problems. Oft times we search for solution to an apparently difficult problem before we make sure the system under consideration is at it's "should" state.
Just as an aside, early LNA's cost between $40k-$200k. Today 15degree K noise performance can be had for several hundred dollars and is a fist size device bolted to a flange, no cooling is required! Progress!
Nice point bdcst! After several years on those big earth stations I moved on and only went back to them later as a consultant for refits to extend their working life another ten years or so. By then for every 30m dish there were a dozen smaller dishes with ambient LNAs and so much simpler and cheaper to build and run. Many of the early generation of big dishes were scrapped by the end of the 1990s. Of course, it was the advent of more sophisticated satellites that drove those changes, with localised footprints (hence much greater EIRP) and higher frequencies (hence smaller dishes). But it was all good experience and I got the chance to work on everything from LNAs and TWT transmitters to digital encoders and servo tracking systems, all of which stood me in good stead later on.
Back to the point of the tale, earth stations were a very competitive field so we always struggled to balance the commercial pressure to get the station on air and earning revenue, and getting the job done right. The station in the Gulf was only our second big dish and it was the very first entirely to our own design so we learnt a lot of very hard lessons in a few hectic months. Nevertheless, the local Sheikh was very pleased because we got ours on traffic before any other station in the Gulf. Subsequent stations evolved to a more or less standardised design and those projects went much more smoothly. I believe a few are still working 40 years on.
I thought this was a fascinating story. In regards to part verification - I think it should be a part of any "kit" endeavor as a built in cost. Even something as mundane as student kits for an electronics course. I have a friend that teaches online and apparently the company they use to send parts to the students does not think part verification is important. Unfortunately it has really hurt the students who are taking a time-compressed course to have to wait on part replacements because they received an AC input voltage PLC instead of the DC one which is used in the course and they can't conduct the labs according to the course timeline.
Nancy, that is a great point. At the shop where I apprenticed every toolmaker made it a habit to put all parts in a box as they disassembled a die for maintenance. After the work was done having a screw or dowel left in the box was cause for disassembly until correct assembly was verified. We had a clown in the shop who thought it was clever to throw an extra part in the box of parts when a tool was being worked on. One old journeyman did not find the humor and a threat to meet the guy on the parking lot after work ended that behavior.
Fred, not only a torque wrench, but a proper tightening sequence can make a huge difference. One shop where I worked made oil pans for a number of vehicles, including a large earth mover. We had to pressure check for leaks and our customer sent a chart illustrating the proper sequence and torque. If you just started at one bolt and went around the pan torquing screws down, it would leak. But loosening the screws and retightening in the proper sequence would usually produce a good seal. I say usually because occasionally improper sequencing would so wrack the pan that they would have to be returned to production and be restruck in the die.
Tool_maker, that reminds me of that old Michael J. Fox movie - Doc Hollywood. He crashes his Porsche in a small rural town. When the home grown mechanics give him back his car after completing repairs, they hand him a little white box. Michael's character asks what is it for? The mechanic shrugs and says "I don't know, but I always have some parts left over..."
From Dell / Intel® New Paradigms in Design Work Scott Hamilton, vertical market strategist for Dell Precision workstations, 5/2/2013 5
Early in my career, I worked as a draftsman and remember the days of drawing on vellum with numbered pencils and Mylar with plastic lead. This was a fun experience in the sense that I ...
I've been using workstations for more than 10 years and love finding ways to get more performance from my system. With demanding professional applications that require more power each ...
A lasting memory from my first job as an engineer in an auto assembly plant is standing on hard concrete at six in the morning, vending-machine coffee clutched in hand, listening to ...
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This radio show will show what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.
To save this item to your list of favorite Design News content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
Tweet This
7 
