# Easter engine problem

Discussion in 'General Electronics Discussion' started by wingnut, Aug 30, 2012.

1. ### wingnut

242
9
Aug 9, 2012
I would greatly appreciate some suggestions as to why the following circuit of an Easter engine, does not seem to be working.

http://www.instructables.com/id/The-Easter-Solar-Engine/

I constructed the circuit which is supposed to be latching. In a LTSpice simulation it shows that the current through the motor should almost instantly go from nothing to saturation.

What I found instead was that I could not get the motor to run. Substituting a diode with a resistor in place of the motor, the diode seemed to gradually come on, and fairly immediately go off if the solar panel was shaded. I would have expected the diode to snap on. A voltmeter across the 80 ohm diode resistor also seemed to gradually rise instead of suddenly rise. I thought transistor QP was supposed to store the charge and then suddenly release it.

I used a 4400u capacitor which was barely able to keep the LED glowing for 2 seconds, so I dont see how it could ever drive a motor.

Any ideas as to where I go from here to build a solar engine?

2. ### john monks

693
1
Mar 9, 2012
Yes. You have no choice but to use a bigger capacitor, C1, or a smaller motor.
If the capacitor charges to 2.3 volts the energy is 0.0116 joules. Almost nothing.
You might try two AA nicad cells in series in place of C1 and see what happens.

Last edited: Aug 31, 2012
3. ### wingnut

242
9
Aug 9, 2012
Thanks for the input John. I am not sure how you calculated the energy to be 0.0116J, but I can see via another calculation that the circuit may be working, but I could never tell (except as you say to get a bigger cap).

I calculate that the maximum voltage drop I can get with my about 4.5V PV solar panel is 2V.
If capacitance C = Q/V then 4400x10^-6/2 gives a total charge of
8800x10^-6 Coulombs of usable charge.

And if I = Q/t and my PV solar panel delivers 0.15A then
t=8800x10^-6/0.15 = 0.06s

If my maths is correct, this means that the capacitor is charging (and discharging) in 0.06s which is too fast for the eye to see, which is why the LED is on all the time.

Maybe I can skip the whole circuit and attach a rechargeable battery directly in series with the PV solar panel.

4. ### CDRIVEHauling 10' pipe on a Trek Shift3

4,960
651
May 8, 2012
That article was written by an Emeritus Professor of Mathematics. Sadly, there's a heap load of crap science out there, all spawned by the "Magic Energy" crowd. Yes, a 4,400uF cap can barely store enough charge to light a LED for two seconds. Note that the professor gave no specs for the cap or the motor.

5. ### john monks

693
1
Mar 9, 2012
Your close to being on the mark.
This appears to be about right because if you divide the junctions of Q1, D1, D2, & D3 you 0.575 volts which seems to be right for them to just start turning on.
When QP turns ON then the voltage will drop until it gets to about 1.2 volts.
At that point Q2 and Q1 with turn OFF.
So you don't get to use all the charge in the capacitor.
If you put in 2 or more nicad batteries it might be best that they not be charged so you don't cause damage.

As for how I derived the energy in the capacitor please try to follow the logic.

The average voltage in a capacitor is (0v+V)/2 = 0.5V V/2

So the total energy in a capacitor is charge X (0.5 V) E=0.5CV

And charge = capacitance X voltage. Q=CV

So energy in the capacitor = 1/2 capacitance X voltage squared E=0.5CV^2

So E = 0.5(.0044F X 2.4^2) = 0.0127joules in the capacitor at 2.4 volts.

6. ### wingnut

242
9
Aug 9, 2012
Thanks for explaining the energy in detail John.

And as CDDRIVE pointed out, they gave no specs for the motor which would run on so little power.

I found where the circuit originated and it seems they used a pager motor or a very efficient 1.5V clock motor, of the sort they admitted is hard to find. I was using a 3V motor from a toy car which cost \$1, so there was no way the energy from that capacitor (or that solar panel) was ever going to drive that motor.

I tried a crystal earpiece in place of the motor to hear if the circuit was latching and unlatching, but could hear nothing.

I have come to realize that not only are there motors and motors, but solar panels and solar panels. I have seen tiny Lego solar panels about the size of 3 Lego blocks drive a toy car - which is pretty impressive when one thinks of how little energy there is in 3 Lego blocks worth of solar radiation.

7. ### john monks

693
1
Mar 9, 2012
A crystal headset needs a sharp voltage transition to work.
I suggest you put about 100 ohms in parallel with it.

Maybe you should find a very small motor out of a cell phone or something.
One thing is obvious is that this circuit is designed only to nudge the motor forward maybe
Only 1/2 a revolution.

If your LED is visibly coming ON then QP is not coming on or the motor is not conducting.
By the time the LED comes ON Q1 turns ON turning on Q2 and QP and the motor should turn slightly. Then the capacitor should start recharging.

Last edited: Sep 1, 2012
8. ### CDRIVEHauling 10' pipe on a Trek Shift3

4,960
651
May 8, 2012
I would not believe everything seen in a YouTube video, especially if it's brought to you by the magic energy crowd. More so if they're selling the plans to their magic!

Try replacing the motor with a 470Ω resistor in series with a LED. For that matter if you have a variable power supply I'd use it in place of the solar cell for testing this circuit.

OPINION: I think using a Cap in this application amounts to a useless novelty. If the Solar Cell was charging a battery, instead of a Cap, it would be a far more viable and practical circuit.

Chris

9. ### john monks

693
1
Mar 9, 2012
Careful cdrive. You have four PN junctions in series. Q1, D1, D2, and D3. A little to much voltage and something is going to blow.

From the looks of this circuit, it appears that this is only a way of powering a small motor from a small solar cell by storing a little charge in a capacitor and discussing it into a motor nudging it forward a little bit.

10. ### wingnut

242
9
Aug 9, 2012
Here is my modified circuit.

I substituted a battery pack for the solar panel. The LED on QP came on as I increased the voltage to 2V. All LED's are yellow. I expected the LED on QP to pulse visibly as it latched and unlatched. I forgot to include a 1K resistor in series with the voltage source to slow the charging of the 4400u capacitor.

This is also a test of whether I can upload a diagram of my circuit.

File size:
16.6 KB
Views:
250
11. ### john monks

693
1
Mar 9, 2012
R1 and R5 may need to be a little larger. Otherwise a small solar panel will not get the circuit to work. I'm anxious to see what happens.

12. ### CDRIVEHauling 10' pipe on a Trek Shift3

4,960
651
May 8, 2012
Thanks for catching that John. A low val resistor (200Ω to 470Ω) in series with the supply should make it safe .

Chris

13. ### wingnut

242
9
Aug 9, 2012
Thank you all for the help. The circuit is working and latching as expected, but it would only do so when I reduced the diodes from 3 to 2. It works fine with a diode and resistor as the output, with the solar panel (which delivers a surprisingly respectable 5.5V in full sunlight).

But even with a 6600uF cap it would still not drive a 1.5V, \$1.5 electric toothbrush motor which I bought this morning. Seems like there is not enough "magic" in the energy this circuit provides.

John, I did try replacing the 470 ohm resistors with 600 ohm ones, and both work equally well. I need to figure out how to get more diodes in series to work, to raise the voltage difference between charging and discharging of the cap.

And I will take into account what you mentioned about protecting the 4 pn junctions. To tell the truth, I did not realize that current flows out of the base of a pnp transistor. But checking a Java simulation of a pnp transistor, I see that you are totally correct.

14. ### john monks

693
1
Mar 9, 2012
Mayb you can just add a other diode in series with the rest.
What is the open circuit voltage of the solar panel?
I believe you need a much larger capacitor than 6600uF.

15. ### CDRIVEHauling 10' pipe on a Trek Shift3

4,960
651
May 8, 2012
Those are high torque little motors. You need a little micropower motor. They may not be able to due much work but it'll probably rotate for a bit.

Chris

16. ### wingnut

242
9
Aug 9, 2012
5.5 V seems to be the max (I am tempted to try a Fresnel lens - can one fry solar panels?)

With the fourth diode it was one diode too far. One could see all the LED's begin to glow, but there was not enough to activate the output.

Even with an external battery box over 6V it would not latch.
Strangely, probing the circuit with a voltmeter (infinite resistance) got the fourth diode to work, but a large resistance of 100k would not.

17. ### john monks

693
1
Mar 9, 2012
It seems r1 should be increased some more and r3 should be decreased.
I'm not sure what 100k you are referring to.

From my experience with solar panel the voltage seems to reach some voltage without a load and doesn't seem to increase much with more light. So I don't think a fresnel lens will help much.
I think the whole idea was to be able to run a motor to some extend without a whole lot of light.

Last edited: Sep 3, 2012
18. ### CDRIVEHauling 10' pipe on a Trek Shift3

4,960
651
May 8, 2012
The LED was not intended to light. Look at this plot (D1(A)) and you'll see why I say that. There isn't enough current flowing through it to light it. Also, I've heard the term "Latch" used repeatedly, while these plots clearly indicate that this circuit operates more akin to a "Relaxation Oscillator".

The motor specs used in this spice sim is based on the motor found at the URL on the schematic.

BTW, after looking at these results it's become clear to me why it's called a Solar "Engine". Actually, it's clever!! I also think that it was primarily intended to be used with a Super Cap. That's a cap >= 1 Farad.

Chris

File size:
66.2 KB
Views:
181

4,960
651
May 8, 2012
20. ### wingnut

242
9
Aug 9, 2012
Thanks a lot Chris.

Checking out the circuit you posted, I noticed that in place of the two 470k resistors I had mistakenly used 470 ohm resistors. Now with the corrected circuit, the voltage across the cap slowly climbs to 6.5V then abruptly drops to 2.3V and then repeats the cycle (using a battery).

My 1.5V motor reads 70 ohms resistance which is probably less than the one you mentioned - so only .3V is generated across it at each spike (or my multimeter does not have time to register higher).

So the circuit is now working perfectly with 3 diodes in series thanks to everyone's help.

I will see if there is a similar motor locally which runs on anything near 110mA.