Joule Thief 101

[shylo]

Is it because the backspikes are adding to the stored energy?
artv

[tinman]

Is it because the backspikes are adding to the stored energy?
artv

No


Brad

[MileHigh]

Dear MH
What kind of a load is an LED?

Oop's.

Brad

There is no "oops" at all.  The fact remains that using a resistor will result in half the energy being lost for two equal sized capacitors, period.  I don't think you acknowledged that switching to a motor is false and meaningless and can't be compared to a resistor.  Your rebuttal to my statement about using a resistor was nonsense.

Big deal you used a LED.  You are doing energy measurements now.  That calls for an attempt at precision.  What is the tolerance on your capacitors?  Since you are doing energy measurements don't you think it is incumbent upon you to actually do serious measurements on the size of each capacitor first, including your own error tolerance?  Perhaps try two or three different ways of measuring them and see if they all agree?  Because they are supercapacitors, perhaps there are some special considerations when trying to measure their value?  Once you have established their actual value, then you can make much more accurate energy measurements.

[MileHigh]

So the next thing to do ,was to find out why we lost less energy by applying a load to the little motor.

I am going to guess that the rotational inertia of the propeller allows the motor to sometimes act as a generator.

[tinman]

    Your rebuttal to my statement about using a resistor was nonsense.

   

There is no "oops" at all.  The fact remains that using a resistor will result in half the energy being lost for two equal sized capacitors, period.

An LED is a resistive load,and i guess you missed my post to Poynt.
When the LED is in play,then the energy lost is less than half-->and remember,there is a 5k resistor across that LED,so as C1 keeps draining after the LED stops conducting.
But dont get your knickers in a twist yet.
If i use just a resistor alone(without the LED),then yes,your are correct,and exactly half the energy is lost. As the LED is also a resistive load,and should not make a difference to the test,i can only conclude that the LED is putting some energy back into the system,by way of the ambient light--as we know,they act like a small solar panel as well.

So i do have the ability to correct my own mistakes MH,and yes,this time i spoke to soon.
This hinges on the fact that the LED is able to return energy to the system by way of acting as a solar panel.

I don't think you acknowledged that switching to a motor is false and meaningless and can't be compared to a resistor.

Of course i do,as a brushed DC motor is not just a resistive load,but also with a small inductive content as well.

What is the tolerance on your capacitors?

We dont need to know what the tolerance of the caps are(but they are +/-2%),as long as there the same cap's. We eliminate the tolerance issue by swapping the cap's around,and carrying out the same test. As long as they are the same cap's,we should get the same results--and we did.

Since you are doing energy measurements don't you think it is incumbent upon you to actually do serious measurements on the size of each capacitor first, including your own error tolerance?

Yes,and that is why we swap the caps around,and carry out the same test. If the caps are equal in there energy storage capacity,then the results should also be equal when the caps are swapped around,so as cap A becomes cap B,and cap B become cap A.

Perhaps try two or three different ways of measuring them and see if they all agree?  Because they are supercapacitors,

OK. To clear this up,the caps used in the LED tests were 10 000uF high current caps,and the ones used in the motor tests are the 55F super caps.

perhaps there are some special considerations when trying to measure their value?  Once you have established their actual value, then you can make much more accurate energy measurements.

As i said,we swap the caps around,and carry out the same tests. If the results are different,then we look for the issue. If the results end up being the same,then no need to look for issue's regarding the caps actual value.

So the conclusion is-in the LED tests,using the 10 000uF cap's,we can conclude that you are correct,based on the fact (maybe) that the LED it self was adding energy into the system,by way of acting as a solar panel.


Brad