Tweet
Afternoon with Paul Babcock
Gentlemen:
I had the privilege to share a long summer afternoon in a fascinating conversation with Paul Babcock in his garage laboratory in Spokane Valley, WA. He had the demo SERPS unit there and turned it on. It ran for at least 2.5 hours continuously operating the two light bulbs. I didn’t see any dimming, intermittent operation, or smoke in that period.
He was kind enough to share with me some of the interesting technical challenges he had to overcome in order to construct this circuit just prior to the conference. In this laboratory Paul has two benches that are dedicated to breadboard layouts of early development switching systems that form the basis of his Fire Power HID diming technology. Apparently these circuits were built about 4 years ago. Since then there have been several improved versions built that have since gone into production. Each one of these early prototype units was is composed of 4-5 circuit boards. One of these benches had a hole in it where at least two circuit boards had been removed. These missing boards were reused to form the foundation for the demonstration SERPS unit.
These prototype units were designed so that almost all of the timing parameters could be adjusted. The bench on the left (where the boards were removed) used TTL logic while the bench on the right used an industrial microprocessor. The intent of these prototype layouts was to develop modular subsystems. Paul explained the basic function of these boards. The first board dealt with incoming power processing and metering. The next card housed the heavy duty switching elements. The following card was a collection of individual gate drivers composed of a series of 24 VDC switching power supplies whose output were optically controlled. The next board contained the logic needed to direct the switching sequence with a number of potentiometers.
I was wrong about the demonstration SERPS device containing a microprocessor in one of my previous posts. It was actually operated with all TTL logic and a number of NE555 timers. It also has a few 0 degree crossing detectors to start the timing sequence. There was another board on the end of the demo SERPS that had a collection of about 20 each small 10 turn potentiometers. I assume these were there to adjust the quadrant timing parameters of the system.
Apparently the utilization of Paul’s existing prototype equipment was almost a “bolt on” component for construction of the SERPS demonstration unit. However there were some challenges. It appears that the original Jim Murray SERPS was designed with two (balanced?) loads. For single power measurements it was desirable to combine these into a single load. This took some time to develop a modification for the existing circuits. According to Paul he had to develop a new full bridge (4 switch elements) SCR switching system that utilized 8 devices in order to provide all the necessary current isolation and flow direction dynamics. It took some time to develop this retro-fit circuit and several solid state devices were lost in the process.
Paul said that the unit presented could be easily up scaled to 1 kW using the transformer that was employed. However for each incremental increase in power level more capacitors would have to be added. Since time was short he didn’t want to take the risk of making additional changes. His salvaged prototype cards had the flexibility to be adjusted to the new timing parameters that allowed the SERPS process to function. Clearly there are two power capacitor banks used in this circuit. As can be seen from the demo photo the system was housed in two plastic tubs. The first tub housed the 1 kW power transformer and two capacitor banks. The second tub contained three(3) circuit boards. These were a power switching board that had at least four switch units, an 8 channel driver board that also contained some logic and timing, and a adjustment board that had all these small 10 turn pots that appeared to be wired to the logic/driver card.
The actual individual solid state switching sub-system channel is composed of an IGBJT and an SCR in series, just like his patent documents show. It takes at least two 24VDC driver circuits to control one channel. There are also some proprietary “steering” devices that help modulate the three step control process that is used to turn one switch channel unit on or off. For low power applications (less than 200 watts) MOSFETs were used. Above that power level IGBJTs are employed. Now days the prices have dropped to the point where IGBJTs are used throughout.
I asked Paul how he envisioned this technology moving forward in the near future. He basically said that there was still a fair amount of work to do to produce a stable, practical working circuit and that no specific plans to make the technology more available to the public can be made before that work is done.
Spokane1
Gentlemen:
I had the privilege to share a long summer afternoon in a fascinating conversation with Paul Babcock in his garage laboratory in Spokane Valley, WA. He had the demo SERPS unit there and turned it on. It ran for at least 2.5 hours continuously operating the two light bulbs. I didn’t see any dimming, intermittent operation, or smoke in that period.
He was kind enough to share with me some of the interesting technical challenges he had to overcome in order to construct this circuit just prior to the conference. In this laboratory Paul has two benches that are dedicated to breadboard layouts of early development switching systems that form the basis of his Fire Power HID diming technology. Apparently these circuits were built about 4 years ago. Since then there have been several improved versions built that have since gone into production. Each one of these early prototype units was is composed of 4-5 circuit boards. One of these benches had a hole in it where at least two circuit boards had been removed. These missing boards were reused to form the foundation for the demonstration SERPS unit.
These prototype units were designed so that almost all of the timing parameters could be adjusted. The bench on the left (where the boards were removed) used TTL logic while the bench on the right used an industrial microprocessor. The intent of these prototype layouts was to develop modular subsystems. Paul explained the basic function of these boards. The first board dealt with incoming power processing and metering. The next card housed the heavy duty switching elements. The following card was a collection of individual gate drivers composed of a series of 24 VDC switching power supplies whose output were optically controlled. The next board contained the logic needed to direct the switching sequence with a number of potentiometers.
I was wrong about the demonstration SERPS device containing a microprocessor in one of my previous posts. It was actually operated with all TTL logic and a number of NE555 timers. It also has a few 0 degree crossing detectors to start the timing sequence. There was another board on the end of the demo SERPS that had a collection of about 20 each small 10 turn potentiometers. I assume these were there to adjust the quadrant timing parameters of the system.
Apparently the utilization of Paul’s existing prototype equipment was almost a “bolt on” component for construction of the SERPS demonstration unit. However there were some challenges. It appears that the original Jim Murray SERPS was designed with two (balanced?) loads. For single power measurements it was desirable to combine these into a single load. This took some time to develop a modification for the existing circuits. According to Paul he had to develop a new full bridge (4 switch elements) SCR switching system that utilized 8 devices in order to provide all the necessary current isolation and flow direction dynamics. It took some time to develop this retro-fit circuit and several solid state devices were lost in the process.
Paul said that the unit presented could be easily up scaled to 1 kW using the transformer that was employed. However for each incremental increase in power level more capacitors would have to be added. Since time was short he didn’t want to take the risk of making additional changes. His salvaged prototype cards had the flexibility to be adjusted to the new timing parameters that allowed the SERPS process to function. Clearly there are two power capacitor banks used in this circuit. As can be seen from the demo photo the system was housed in two plastic tubs. The first tub housed the 1 kW power transformer and two capacitor banks. The second tub contained three(3) circuit boards. These were a power switching board that had at least four switch units, an 8 channel driver board that also contained some logic and timing, and a adjustment board that had all these small 10 turn pots that appeared to be wired to the logic/driver card.
The actual individual solid state switching sub-system channel is composed of an IGBJT and an SCR in series, just like his patent documents show. It takes at least two 24VDC driver circuits to control one channel. There are also some proprietary “steering” devices that help modulate the three step control process that is used to turn one switch channel unit on or off. For low power applications (less than 200 watts) MOSFETs were used. Above that power level IGBJTs are employed. Now days the prices have dropped to the point where IGBJTs are used throughout.
I asked Paul how he envisioned this technology moving forward in the near future. He basically said that there was still a fair amount of work to do to produce a stable, practical working circuit and that no specific plans to make the technology more available to the public can be made before that work is done.
Spokane1
I hope they make it big and get plenty of green backs to fund their operation before the bell rings for them.
so it is time to get in on. Well I am sure they would like to see their unit excell and yet are very wise men about what it takes to stay out of the way.
to review.
. I had never thought about it much till John shows how to build an SG OSC various ways and so I sampled the operation of this invention for many hours hooking to batteries.
Comment