Selfrunning cold electricity circuit from Dr.Stiffler

[hoptoad]

Hmmm, to develop a theoretical understanding of "something" prior to experimental evaluation, or to derive a theoretical understanding of "something" after experimental observation?;   that is the question Horatio!

Hmmm, the old, which came first, chicken and egg routine again.

Personally I'm waiting in great anticipation for the chick to hatch!     KneeDeep!

This continues to be an interesting thread with intriguing surprises. Cheers all from the Toad Who Hops 

[Loki67671]

Progress update;
Someone lost their AM radio  Removed the AM loop from the radio and inspected it. Horribly assembled by god knows who. Cheap, no way to compete with this stuff if people are willing to buy this garbage. The winding is deformed by wax of some sort. I don't know if it was deliberate or not. Doesn't matter. Will fire up with my heat gun and reform. Now I am having a bit of difficulty with obtaining the tuning range of colpitts base inductor that the Doc calls for, as measured with an LCR. I did manage to get a 4.5uH to 9.5uH put together on a phenolic form. It tunes with a standard ferrite slug. I'll see if I can make that up by adjusting the cap around the 400pf mark. We'll see what range I can get going out of this. I have to clear a space in the basement for firing this thing up. Don't want to upset the powers that be by blasting off indiscriminate RF emissions, at least not too far  . I'm subbing in an NTE-123AP genpurp x-sistor to try first for my oscillator as it crosses to the called out one. Anyone know how critical the collector inductor design is? Are we talking about large self inductance or just 22uH of some type? First light....soon, hopefully tomorrow................Jim 

[DrStiffler]

@ALL

Gentlemen!

Try not to get all worked up again. Here are some thing that will help your confusion.

Power In: It is damn hard to measure (period). On my SEC devices you can filter th hell out of the input and still see artifacts >50MHz. I have been working with RF for 40+ years and never have had a problem similar to making these simple circuits quiet. That is the thing that is most interesting and (maybe) why no one has seen this effect before. Bad, noisy circuit, filter, bypass or throw it away. Conventional EE does not want noise, spikes, jitter, bleed, whatever you want to call it. EE wants stable calm circuits that can show no noise or artifacts on the  power rails.

Yes you can quiet these circuits, but that is not the point in measuring input. The lab that (I think) will test this circuit uses a major piece of equipment that has a band width of 20-30khz for True Power. Guess what, SEC will never show OU this way. You need a very expensive digital integrating oscope, something with a sample rate >10MHz, then and only then will the facts appear.

I do not want to make anyone mad or say they are wrong, but someone please show me one of my circuits (working) with a clean set of wires feeding back to the power supply??? Please, I am willing to learn even this far into the game.

Now for output. I have to admit it has to be Heat and this is far from easy or simple, even though the concept is simple. Building a device that gives correct results is damn hard.

You lock away your circuit in a container, hoping the tuning does not change, calibrate the whole instrument with a resistor and DC known current and voltage, determine the 'Offset' for the device, let it site for many 10's of hours to again reach equilibrium, power up your circuit and hope for the best. No adjustment when its in the box, just cross your fingers.

In my lab the device that (hopefully) will be lab tested showed an OU of 2.158 after an 8 hour run. This may sound great, but in the first hour of testing showed an OU of 2.663, it should not have gone down, it must go up or stabilize. So I have fingers, arms and everything crossed. My lab cell (crude) but following all requirements for calorimetry may be in some way hosed and I do not know it. Yet power out in a SEC, because of the high impedance on the output is best left to a calorimeter.

I think I am okay, but maybe the next few weeks will really tell. I do not think I have that much wiggle room.

My device is shown below, 1.5' foam around circuit and water container, cotton batting and two layers of bubble wrap.

[DrStiffler]

@All

From that most exciting first hour, here is how the computer looked.

Q = c m dt * 1.1622E-3 = W/Hr

   Calorimeter Calibration Offset   3115.000
   Temp start t(s)   20.400      
   Temp end t(f)   21.800      
   Water/SH   1.000      
   m = mass/gm   500.000      
   dt = t(final) ? t(start) 'C   1.400      
   W/Hr   4.434   15973.405   Joules
   Run Time (minutes)   60.000   1.00   Hours
        Vs = supply voltage   24.000      
   Is = supply current   0.0694   1.666   Watts
   W/Hr   1.666   5995.680   Joules
   COP   2.662      

[mrl]

@ALL

Gentlemen!

Try not to get all worked up again. Here are some thing that will help your confusion.

Power In: It is damn hard to measure (period). On my SEC devices you can filter th hell out of the input and still see artifacts >50MHz. I have been working with RF for 40+ years and never have had a problem similar to making these simple circuits quiet. That is the thing that is most interesting and (maybe) why no one has seen this effect before. Bad, noisy circuit, filter, bypass or throw it away. Conventional EE does not want noise, spikes, jitter, bleed, whatever you want to call it. EE wants stable calm circuits that can show no noise or artifacts on the  power rails.

Yes you can quiet these circuits, but that is not the point in measuring input. The lab that (I think) will test this circuit uses a major piece of equipment that has a band width of 20-30khz for True Power. Guess what, SEC will never show OU this way. You need a very expensive digital integrating oscope, something with a sample rate >10MHz, then and only then will the facts appear.

I do not want to make anyone mad or say they are wrong, but someone please show me one of my circuits (working) with a clean set of wires feeding back to the power supply??? Please, I am willing to learn even this far into the game.

Now for output. I have to admit it has to be Heat and this is far from easy or simple, even though the concept is simple. Building a device that gives correct results is damn hard.

You lock away your circuit in a container, hoping the tuning does not change, calibrate the whole instrument with a resistor and DC known current and voltage, determine the 'Offset' for the device, let it site for many 10's of hours to again reach equilibrium, power up your circuit and hope for the best. No adjustment when its in the box, just cross your fingers.

In my lab the device that (hopefully) will be lab tested showed an OU of 2.158 after an 8 hour run. This may sound great, but in the first hour of testing showed an OU of 2.663, it should not have gone down, it must go up or stabilize. So I have fingers, arms and everything crossed. My lab cell (crude) but following all requirements for calorimetry may be in some way hosed and I do not know it. Yet power out in a SEC, because of the high impedance on the output is best left to a calorimeter.

I think I am okay, but maybe the next few weeks will really tell. I do not think I have that much wiggle room.

My device is shown below, 1.5' foam around circuit and water container, cotton batting and two layers of bubble wrap.

Get yourself some large caps (maybe some super caps).  Charge them up then let them run the circuit and take note of the voltage drop within a measured period of time.  Then do an RC time calculation.  That should give you an idea of how much energy is being drawn.  If done right, it could yield some reasonably accurate readings.  It would be somewhat the equivalent of a dual slope AD converter.  You apply a known load to the cap for a certain time, measure it, and get an error reading.  Then charge it up again and measure the load with the charged cap discharges rate and apply your error reading to your math.  You can use digital timers.  A PIC micro would do this nicely for you.  You would have a very accurate instrument that, as far as I can see, would not lie if you did it right.  The error value would compensate for any resistance in your relays or FET switching circuits.  Also, the switching doesn't have to be fast.  Maybe 20 times per second or even less on larger caps.




This is just off the top of my head.