Joining the dots
February 2012: We thought it was a cakewalk as we built a flawless RISAT. A very cute elephant, obeying all the commands and behaving as expected. We heaved a sigh of relief when it was put in a 9 m dia. 40 m high, giant thermo vacuum chamber of ISAC. We started thermal cycling.
The bad news came in dead of night. The main chain output was corrupted by an oscillation somewhere close to 850 MHz chain at a temperature of 14 degree C or lower. I asked for repeating the measurement and it was repeating exactly the same behaviour. We switched over to the redundant chain. Phenomenon is the same but the cut off was at 13 degree C. We had never seen such ghostly behaviour. The system was misbehaving like a switch. Very well behaved above a temperature and lunatic below that temperature. We zeroed on to a package called Frequency Generator (FG) which has a 850 MHz output.
The elephant and the blind men
We asked for a testing of a spare third FG package lying at Space Application Centre (SAC). It showed the same behaviour but at a cut off temperature of 18 degree C. Everybody concluded that RISAT will be grounded. Success has many fathers. Failures have many armchair critics. Failure draws big attention. The person who solves it gets big credit for saving a system from imminent death. So many fellows joined the fray. I was hearing from Bangalore – so many parallel meetings, so many simulations were being tried out at SAC. One particular big man, who always looked down upon anybody working in frequency band lower than optical band, also joined the fray. I knew so many blind men were running amok to describe the elephant. I decided to fly down, catch hold of the situation.
One thing, nobody can find fault with me, is my ability to assess people’s real worth. I zeroed upon a lady engineer, who used to be cornered by her bosses, for being very bright. I gave her the job. She reasoned out that at a particular temperature and below, the flat cover on the package gets distorted, opening a pathway to connect the cavity housing 850 MHz circuit to an adjacent blank cavity. By design mistake the circuit is operating very close to instability. Additional RF cavity in very neighbourhood, was driving the circuit to oscillation. She suggested changing four resistors in the circuit. RISAT was back on track.
All those experts were disappointed at such easy solutions and made me stand before houseful of gatherings to explain why simple solutions should work!! Her immediate bosses went out of the way to spoil her career, corner her in the team. She somehow could be extricated from the quagmire.
There was another incident where I had to struggle with my own folly. We tested RISAT active antenna with the payload elements on satellite with 5 m umbilical chord. In fact, we had been carrying out the test from SAC itself, the chord being the permanent fixture. But we had to finally integrate the active antenna with the satellite. We removed the chord and connected with satellite. We switched on the payload and to our dismay we found that 5 out of 12 tiles got powered on. Rest were stubborn, not responding at all.
The failure of RISAT quickly became a hot topic in ISAC, within no time. When I was leaving ISAC for lunch, the CISF jawan consoled me at the gate that I should get it right in my next version of RISAT!!
So many people came to give me strength as well as to slyly enjoy my predicament. I wanted to buy time. I found out that connectors have been mated without cleaning them. I asked for demating the connectors, cleaning them with alcohol and remating them. I knew the job is so laborious that by the time we were through, all those onlookers would be tired out, giving me time for cool thinking.
In between, then Associate Director of ISRO Satellite Centre (ISAC) also came, took me aside and asked me quite sternly about the rationale of cleaning the connectors with alcohol. I told him in straight face that alcohol is good enough to loosen the toughest of the nuts. He gave a bizarre look at me and quickly went away.
After the crowds thinned, I figured out that 1 MHz clock running on 30 m harness is the culprit. The horrible shape of the clock on oscilloscope was the pointer.
I decided to apply RF technique of stub matching to digital signals. We roughly calculated the open ended twisted cable to be used as stub cable. We put stub in each antenna panel, trimming them in steps of a centimeter to arrive at perfect clock shape. RISAT flew with stub cables. Next day I got huge compliments from my colleagues for using RF technique to solve digital problem. Hitherto probably nobody used. I still cherish the patting I received from Dr. Vasantha, undoubtedly a master in digital technology.
Why do I narrate these stories? The real life issues are vexing last mile problems. You are left with very few options, hardly any room to manoeuvre. At this stage the issue of the problems, never envisaged, threatens to derail the work done over years. This is where the ability of understanding the problem by imagination and joining the dots comes into play. Unfortunately our education system is designed to increase our knowledge as series of information. But our children are not encouraged to nurture their right brain too, to develop soft skills, strengthening imagination. Imagination is the key to the ability of joining the dots of dissimilar and disjointed observations.
Many of our programmes did not get satisfactory conclusions or did not meet user expectations, as promised, simply because the teams lacked those brilliant minds to solve last minute problems, which will occur in any case. After all, we cannot envisage all the possible modes of failures, however best we may try. But solving them, as and when we encounter them, is the key to success.
Tapan Misra is a globally acclaimed distinguished scientist, who has headed India’s Space Application Centre and also served as Advisor in Department of Space. He has contributed significantly to India’s Space Programme.