Former RMC staff: Errol Wallingford
Years at RMC: 3 Sept 61 – 3 Sept 86 Positions: Lecturer, Assistant Professor, Associate Professor
Ottawa native, Errol Wallingford graduated in 1946 (Grade 13) from Glebe Collegiate High School. He received his BSc from Carleton College (formerly the old Ottawa Ladies College) in 1953.
Following a 35 year teaching career at RMC he then went to Annapolis Maryland for two sessions.
His leisure activities in retirement are primarily sailing in the summer in a Laser dingy on Sydenham lake. Skiing in winter all over Canada and playing bridge at anytime.
For the last 4 years he and his wife have taken cruises with another one planned for this year.
Tell us a little of your background prior to arriving at RMC.
While waiting for returning servicemen to go through college I worked for 3 years at the main branch of the bank of Nova Scotia in Ottawa as teller number 6,4,3,2 and 1 before accepting a 4 month research position at the Canadian Signals and Research Development Establishment on Montreal Road in Ottawa. This was based on a recommendation from my brother in law who worked there, after telling them of my qualifications as a radio ham (Call sign VE3BEN) and extensive work on all types of early crystal sets and radio receivers. The task was to try to change a multi-stack, multi-turn variable inductance tuner in a military receiver from exponential to linear operation across the full frequency range.. By extensive testing and winding my own high Q inductors, along with capacitors, a nearly perfect linear tuning was achieved. Many considered this to be theoretically impossible but they were thrilled with the results.
Where did you do your undergraduate degree?
At Carleton College I graduated with a BSc in 1953.
What did you do following graduation?
Following graduation I spent 4 years at DND Inspection Services testing many types of military equipment to their specifications. This included radar sets, communication receivers, waveguide equipment, battery packs, generator sets etc. Tests were for vibration in 3 directions, operation in an explosive environment, hot and cold cycling, susceptibility to external interference, or quietness in not creating conducted or radiated interference. Equipment had to be repaired before proceeding whenever it failed. Microwaves and screened room measurements of electrical susceptibility and radiation were my specialties.
When Waterloo College Associated Faculties started I accepted a position as a physics lecturer teaching in their co-op electrical engineering program. I stayed for two years until I was told I needed a Master’s degree to continue.
Went to Ottawa U where I completed the required courses for a Masters or PhD and received my MSc in physics in 1961.
You arrived at RMC in September 1961. What do you recall about the erly days?
At RMC I was delighted to be chosen by the department to supervise one of the first candidates for the new MEng program. This was followed by several others. It was an honour to have so many young gentlemen wishing to work with me. I did my best to find suitable challenging projects for them.
I am deeply indebted to Dr John Wilson for showing me so much about microprocessors. My work with him was interesting and productive and helped considerably with many research endeavours, as well as in the publication of many papers both with him and several other colleagues in the future.
What were some of the more interesting studies and projects that you were involved?
A particularly interesting application was based on the work by Patel and Hoft. They produced two papers showing that if a full-wave bridge is used in lieu of the much more common half-wave bridge that it was possible to produce an essentially harmonic free waveform to drive a 3 phase motor. When using a microprocessor 3 phase, square waves are easily generated and the even and the third harmonics and their multiples are automatically removed. This leaves only the 5, 7, 11, 13 and 17th harmonics requiring removal. Higher harmonics are of much lower amplitude and can be safely ignored.
Using the calculus of variations they solved the exact problem and a practical solution for a discrete waveform with zeros appropriately placed to the nearest 1 degree of rotation of the motor. The only problem was that no one seemed to be able to implement their solution.
The required solution was immensely intriguing so it was rigorously pursued. I found a microprocessor solution within 6 months. It consisted of about 40 discrete speeds of Patel and Hoft waveform which could be controlled and commanded at will by a master microprocessor. All attempts to publish this paper failed as the then simpler but dirty waveforms of the half-wave bridge dominated and still hold true to this day except for my applications. In disgust I wrote and published a patent on this work for DND who paid the patent fees. The first implementation was done at RMC using a government grant and Janos Kovacs from Hungary. The implementation was successful by using microprocessors with a fixed number of clock cycles for each command and by adding no-ops as required to maintain synchronicity. Thus changing from 17 to 18 degrees of rotation was identical to changing from 359 back to zero. The required waveform occupied successive degrees of memory and was accessed sequentially.
What other projects stick out in your mind?
Another intriguing project was one done with Dr Claude Bordeleau. It used the synchronous techniques shown above to generate a triangular wave to drive a radioactive source, in lieu of an external triangular generator, to produce a MossBauer display. Prior to this change MossBauer spectrometry took several days to get meaningful results.
With the synchronized change displays were visible immediately and if the apparatus was faulty it would show up in just a few hours instead of several days. I presented this finding at a conference in Hawaii and published it with Dr Bordeleau in a leading journal.
Did you do any work with other NATO Forces?
From August 1978 to June 1979 I was an exchange professor at the US Naval Academy.
This was a fascinating year as well as a very productive one. I struck up an immediate rapport with Dr Tony Sarkady who was an expert in signal processing. In turn I was a skilled programmer. He showed me auto-correlation, cross-correlation, power spectral densities etc. and I programmed them on a 16 bit microprocessor using fast-Fourier transforms. We attended many conferences and presented or published several papers, some of them with other interested faculty. Signal processing was used to supplement courses taught at the Naval Academy and at RMC on my return. One paper was on my simple implementation of Walsh transforms. On Aug 5, 2009 Dr Jetse C. Reijenga from department of Chemistry from Eindhoven Netherlands sent me an email saying he will use my computer generated Walsh transforms to present a paper on Separation Science. He was impressed with the simplicity of the implementation.
How about your sabbatical? Where did you go? What did you do?
My final year at RMC was a first and only sabbatical. This year was spent at Process Technologies Limited in Oromocto, New Brunswick. I was tasked with finding a means to power plasmas efficiently in a post reaction chamber which would considerably reduce the emissions of poisonous gases into the atmosphere during the coating of wafers used for the manufacturing of semiconductors.
A literature search was not helpful as a theoretical model for a plasma from IBM was dimensionally incorrect. They also claimed plasmas to be reactive hence frequency sensitive. All Process Technology personnel were afraid of high voltage from the radio frequency generator so this gave me complete freedom to find an appropriate solution. I started with wrapping metal straps around a fluorescent light bulb removed from its sockets. Next a plasma was attempted inside a metallic chamber with a glass top which permitted viewing the plasma generated and through which the poisonous gasses were passed prior to release in the atmosphere. A very big problem was that all commercial vacuum sealed connectors for entering the chamber failed immediately or very shortly after the plasma was fired up as the plasma deposited conductive coatings on the connector shorting out the power input. Other experimenters used these post reaction chambers but typically used a pair of external straps wrapped around the chamber which required large amounts of power to sustain a plasma.
Finally my colleague suggested a spark plug. A spark plug without a resistor had its ground electrode removed and its hot electrode extended with a small metal rod welded to the central insulated point to extend it. A glass rod was placed around the extension arm which was screwed onto an inverted metallic cup. I soldered a short piece of ½ inch copper water pipe to the external part of the spark plug which was threaded into the chamber and added a commercial connector for attachment to the generator. This worked well and plasmas finally could be sustained for about a month before dismantling and removing the accumulated poisonous contents. All these early experiments required large amounts of power to generate or sustain the plasma. A limited range auto-transformer was available for matching to the plasmas impedance but its range was much too low. I went to Radio Shack and purchased several feet of similar wire to add to the autotransformer and output several additional taps from it at uniform intervals.
The auto-transformer was next matched to the much higher impedances of typical plasmas. This reduced input power significantly. Also the plasmas were found to be frequency insensitive. I derived a new model for the plasmas from Maxwell’s equations which was remarkably similar to the earth’s ionization layer. This paper was given to senior personnel for their use. I do not know if it was ever published. I retained a copy of this paper but all other documentation was surrendered to the company when I left.
Was there a patent?
I wrote a patent in the name of the president, vice president, myself and my colleague.
A commercial model of the post reaction chamber was made. My colleague repeatedly wished to replace the spark plug with something else. This I would not allow. Many changes were attempted. In the production model the only change I allowed was to turn the post reaction chamber upside down as it worked equally well up or down. The first customer was IBM the second was ATT. They were both surprised when I showed them that plasmas were insensitive to frequency.
What about your time after retiring from RMC?
After retirement from RMC I worked for Artech Corporation based in Falls Church Virginia for two sessions within 2 years. My work station was at the David Taylor Research Center in Annapolis Maryland. I was recommended for this position by my good friend Dr Anatol Sarkady of the US Naval Academy.
Were there any special challenges?
I was hired to try to find flaws in welds in steel plates which could be up to three inches thick using a pair of ultrasonic probes placed at an angle to the surface. One probe was a transmitter, the other a receiver. Prior to my arrival a husband and wife team from England spent a year trying to find a successful solution in vain. He had a PhD in physics and she had a PhD in mathematics. A quick look at their proposed solution shocked me in that it defied both the laws of reflection and refraction. I was on my own to find a new solution.
I spent a few weeks playing with the probe pair and was gratified to see so many signals coming out of the receiving probe. I assured the US Navy that a solution should be possible once the various signals were analysed and properly interpreted.
Did you find that solution?
Eventually I attempted using a metal screen to reduce the number of signals. Much to my surprise one of the largest return signals disappeared before the metal screen got to the main beam. My antennae theory immediately became relevant. I recalled that every radiator had not only a main beam but side lobes as well. It was the first side lobe of the main beam of a sinx/x distribution which was travelling through the water path striking the surface of the steel and reflecting up into the receiver. Discontinuities below the surface used the main beam which penetrated the steel then refracted down to hit the (void) target which reflected it to the surface and then refracted at the top surface of the steel to hit the receiving probe where it was recorded. Knowing the ultrasonic speed in water and in steel and the geometry of the system; the time of flight of the main beam to the target and return, and time of flight of the much shorter pure water path of the first side lobe provided a measure of the depth of the target.
How did you determine flaw sizes in the steel?
Flaw sizes in steel were to be determined automatically, if possible, in each of the 3 perpendicular dimensions to within 1 mm. This was accomplished by scanning the probe pairs, moving them say 1mm in the x direction while traversing the target, then moving it 1 mm in the y and repeating the x scans until the entire target surface is traversed. Two or three different spacings between the probes was found to be optimum for best signal to noise ratio dependent on the depth.
When traversing a flaw many more hits than the size of the flaw are recorded.
By using a set of 18 known samples and based on its depth below the surface and the known flaw size in each direction the apparatus could be meticulously calibrated.
A fully automated version was finally made in my second session at the David Taylor Research laboratories and I had an independent engineer place several of the 18 samples under the scanner at random. It successfully identified each and every sample.
Did you get a patent?
A patent in my name and my research supervisor was made for the US Navy. I was also asked to review ultrasonic results from the Washington laboratories and MIT. Some of these involved various plastics which were thought to have special properties when measured with a simple ultrasonic probe. The special effects could easily be due to the ignoring of the main beams side lobes bouncing around off the walls of the specimens.
What are you doing these days?
I finally officially retired to my current residence on the north shore of Sydenham Lake – near Kingston. [email protected]
I still do a “little work”with Tony Sarkady of the US Naval Academy.
I designed a very simple humidity sensor using only 3 components; the sensor, a computer controlled switch to reverse the polarity of the sensor between readings and a microprocessor. An added temperature sensor and EEProm for storage of temperature and humidity results permitted the status of perishable cargoes to be monitored throughout shipping to see if proper storage was maintained in transit.
Although you are “retired” – would it be fair to say that you still are involved?
Yes, I’ve worked also with Hugh and Mitch Kerr on dual 64 speed binary transmission control for an armed personnel carrier. Wrote software and solved problems like making smooth shifts by delaying the hydraulic control to the smaller faster shifts until the larger bigger and slower shifts were nearly completed.
I continued to work with my implementation of the Patel and Hoft waveform for 3 phase motors. I modified a Trans2 golf cart to run off of a 7.5 hp electric motor using 120 volts of DC (ten 12 volt) sealed lead acid batteries. I now know that the motor only operated at 5 hp but this was sufficient to climb steep grades and provide a range of 30 km at about 40 kmh.
Later I purchased a 1997 Geo Metro and had it converted to electric. A stock 15 hp 2 pole 3 phase motor was installed. This used a correct voltage of 168 volts DC (fourteen, 12 volt) lead acid batteries. I made several patents to control this through 4 shifts of the 5 speed manual transmission. It could attain 100 kmh on the flats. But the 800lb of battery in the back seat made hill climbing slower. I presented papers on this at EVS 17 in Montreal and EVS20 in Long Beach California.
It is interesting to note that the inverter on this vehicle never became hot, proving Patel and Hoft system works. The motor’s front and rear bearings had some heating but only because the cooling fan on the totally enclosed motor was removed for a speed sensor.
I used two small venting tubes for cooling each bearing. They were directed to the vacant radiator location. This is in sharp contrast to the EV1 made by GM. It used liquid cooling for both the motor and the inverter.
Were you ever on an executive committee that pertained to your expertise in this field?
I was president of the Electric Vehicle Council of Ottawa for about 3 years.
What have you done lately?
More recently, using simple microcontrollers recommended by Tony Sarkady, I electric powered a riding Tractor Lawn Mower.
Low cost inverters may be made using simple mosfets. It is considered possible to run variable speed 3 phase motors of 1 or 2 hp off of 120 volts AC. I have a 1 hp motor currently working off of the 120 volt AC at reduced power.