Announcement

Collapse
No announcement yet.

Limitations of the SG

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Limitations of the SG

    Hi All,

    I have been working on this thought for a very long time.....what are the limitations and why do they exist. I have gone over and over this problem and keep running into the same results. After burning up several transistors and experimenting with different coil sizes it comes down to this one thing. We are limited by the transistor technology. We can build coils of any size and we can produce some very impressive negative energy but directing it has always been the big problem. JB has tried rotating armatures with brushes but the break of the dc input and the connection of negative energy output burns the hell out of the contact areas. They just do not last very long.

    JB then went to using MJL21194 transistor and it worked well.......providing we do not put too much power thru it. The transistor is the weakest link. The MJL21194 has a current rating of 15amp and the FJL6920 has a current rating of 20 amp. We are limited by the amount of current we can move thru these components. This means that we are limited by the amount of power input and exhaust thru this little devise. We can use any size coil we want providing we do not exceed the limitations of the transistor we are using. Around 130 feet long of #18 wire is the maximum size we can get away with without exceeding the transistors power handling capabilities.....at 12 volt. If we exceed 12 volt @ 130' then we are taxing the transistor very hard. Those wanting to exceed 12 volt must use a shorter coil length to limit the amount of power so we do not burn the transistors. We can also use many coil windings litzed together providing each single coil lead is on its own transistor and we do not tax the transistor too hard.

    At the end of the day the transistor is the weak link in the SG. Until someone comes up with a much higher "fast switching" current rating transistor then we are limited to 130' of coil wire per strand.

    Bud
    Do not procrastinate! Make something happen...even if it is wrong. Once begun half done!

  • #2
    Hey Bud, you're correct in saying that we are limited by the SOA of the transistor. However I have run parallel 130' #18 windings (8 filar coil into 4 transistors) at 24v and still got pretty good results. The transistors remained cool.

    But it is important to note that the draw current is the AVERAGE current. The peak pulse current is many times higher than the average current.

    John K.

    Comment


    • #3
      Hi Bud,
      Your observation is very accurate as far as pointing to the transistor as being the weaker link. Couple points however, if I may.
      1) Choosing the appropriate base resistance will determine how much current you will allow to flow thru your transistor. in other words "heat". But you already know that
      2) 130' of #18 wire is the recommended spec. It's a good place to start but I don't believe it is a hard and fast rule. I like to see it as a trade off between inductance and impedance. Your wire/coil need to be sized to produce a decent magnetic field but still need to have an appreciable impedance to collect the negative energy. As you can see those are contradicting requirements.
      Add to that the limitation of switching (i.e transistor) and you get your design space.

      I don't pretend to have all the answers, so I welcome all feedbacks.
      NoFear

      Comment


      • #4
        Hi Mike,

        I have done quite a bit of experimentation with higher voltages....first 12v... then 24..... then 36. When using a 130' long coil weather multiple strand or single and using the MJL21194 transistor, the heat gets to be an issue at the higher voltages. I have seen many of the transistors fail....some right away.......others many hundreds of hours in service. I guess what I have meat to say above is to keep the voltages down @ 130' or shorten the coil length to go with more voltage in reference to the MJL21194. Stay on the safe side, away from the curb, with coils and high voltages.

        When using the FJL6920 transistor I have found that these can take quite a serious beating in comparison to the MJL21194. The FJL6920 has a 5 amp advantage over the MJL21194. These seem to be way more forgiving and you do not need to put the neon bulb in the circuit.

        However when any of these transistors fail there is no warning! You might see a 10 degree rise over ambient and then all of a sudden they either short or fail to trigger. It seems the negative energy is very hard on them and with very little heat. The same exact thing we see when the negative energy works on our control relay circuits. Wires getting burned off and contacts being eaten away with very little heat.

        After classifying around 200pcs. MJL21194's I have seen very wide test results using the test rig in the intermediate SG book that I built. Now in comparison I have classified 300pcs. of the FJL6920 and the results were very narrow and the production of these transistors seems very impressive......very little variation has been found. The FJL6920 seems to me the best bang for the buck when pushing the limits.

        Bud
        Do not procrastinate! Make something happen...even if it is wrong. Once begun half done!

        Comment

        Working...
        X