Joule Thief

[Rosphere]

I am finding something interesting about the jtc as I experiment with lasersaber's air batteries.

I have a couple of these that put out about 30mA
So, this should be enough to run a jtc with a secondary, and indeed it is.
BUT I cannot get anything like the high voltage spikes from this galvanic pile battery that I do from a chemical AAA.

So, I am wondering what this means about the joule thief and how it works...

I found a similar anomaly with my JTC MOSFET air-core variant, "Tubthumper."  (Because I use a MOSFET instead of of an NPN-BJT, I need to use higher voltage.  I use 9V.  Like a weak AA to a standard JTC, a weak 9V will still thump my circuit.)

Anyway, I used a Radio Shack Nickel-Metal Hydride 9V instead of my typical Ray-o-vac Alkaline 9V and I found that my wavelength was around two scope divisions longer, (lower frequency.)

EDIT: Scope Shot, 0.1uS/DIV

[jeanna]

indeed it is 'asking for'. the 'dead' AA battery is 'dead' in respect to its voltage 'capacity', 1v instead of 1.5-6v but still contains a large 'capacity' of amperage, relatively speaking of course. the jt uses this amperage to saturate the ferrite core which then effects the switch and it starts all over. your galvanic pile doesn't have this 'capacity' of amperage that the AA does... grossly simplified, sorry.

Hi wilby,

I do not think this addresses what is happening.
I have enough current for plenty of toroid saturation.
I am getting 100-333KHz on various jtc's using the air battery, so there is plenty of ferrite saturation followed by switching.

No, I think there is another answer.
I am looking for higher volts and probably a sharper spike.

I also think it is pretty interesting that the amp-metering methods we have cannot pick this up.

I appreciate your suggestion.
Thank you,

jeanna

edit
You too rosphere. I am seeing more of a curved sinewave and less of a crazy-tall spike. But I think the differences between a toroid based jtc and a mosfet based one are too many to be able to narrow this down.
But this is interesting.

[mscoffman]

Hi wilby,

I do not think this addresses what is happening.
I have enough current for plenty of toroid saturation.
I am getting 100-333KHz on various jtc's using the air battery, so there is plenty of ferrite saturation followed by switching.

No, I think there is another answer.
I am looking for higher volts and probably a sharper spike.

I also think it is pretty interesting that the amp-metering methods we have cannot pick this up.

I appreciate your suggestion.
Thank you,

jeanna

edit
You too rosphere. I am seeing more of a curved sinewave and less of a crazy-tall spike. But I think the differences between a toroid based jtc and a mosfet based one are too many to be able to narrow this down.
But this is interesting.

@jeanna

I think wilbyinhbriated's thinking is correct. In school they
taught us to always show a power source as a battery
with a resister in series with it. This is source's "internal
resistance" The air battery has a high internal resistance
due to lack of ion mobility. This assumes the 30ma is provided
when the air battery is "fully charged". The worn out AA carbon
zinc battery still has a low internal resistance and it could put
out many amps instantaneously when it was new. You may even
have to worry about them overheating. This is what the JT
pulses against.

One way to correct this situation (to some extent) is to use
an electrolytic capacitor (note+/-) in parrallel with the battery.
The capacitor
has low internal resistance so should act as an averager for
the JT to pull against. It will charge from the battery the whole
time but supply high current pulse only a short part of the total
cycle. Any DC source that is not a battery should likely have
this pulse filtering cap across it. Electrolytic caps value go from
1uf to thousands depending on the pulse rate. If you want a
really long lived electrolytic (they are funky devices, see
wikipedia) then use a tantalum type capacitor.

---

The way they build mosfets is responsible for the it's
problems as the control gate has high capacitance
to the other electrodes. That capacitance doesn't
matter at DC, turning on/off headlights in an auto
for example, but as one goes into audio frequencies
that capacitor starts wasting energy. By the time
it gets to RF, I'd say it's as difficult to design with
as bipolars are.

Don't worry, as I have said before the overunity field is one
heck of a difficult place to learn any electronics at all.

:S:MarkSCoffman

[jeanna]

Hi MarkS
Yes, with these batteries I have been using the joule thief circuits (with secondary) that employ a 1F memory cap across the battery leads. This is not a tantalum cap and it charges and discharges more slowly, but once the lights are on they stay on and frequently build in brightness.

I plan to make one of these air batteries that puts out 60-100mA just to see if conrad is right about needing twice the mA that the dmm says are being used.
I think his observation is right and pertinent here.

I doubt if I can make my scope show what his does so I have to trust his word, but this will answer a lot of questions that I have had since the time I was working with the 2 secondaries circuit and began to not trust the dmm across the 1 ohm resistor telling the real amps draw.

Thanks,

jeanna

[sm0ky2]

I am finding something interesting about the jtc as I experiment with lasersaber's air batteries.

I have a couple of these that put out about 30mA
So, this should be enough to run a jtc with a secondary, and indeed it is.
BUT I cannot get anything like the high voltage spikes from this galvanic pile battery that I do from a chemical AAA.

So, I am wondering what this means about the joule thief and how it works.

@ Jeanna

that is the exact line of thinking that made me put this down for a while....

do this::   play with some caps and find one you can charge and set off that transistor

charge it.
so you have a known value, uF/ voltage
JT it, and make sure you can charge the cap, fire the transistor, and pulse the LED.
once you get it set up, remove the LED and replace it with a 2nd cap of known value and a diode.

charge primary cap, and flip the switch.
then measure your voltage on the 2nd cap.
now you have two discrete energy values to compare.
in and out.
wilby is leading you in the right direction.
but you should see what happens for yourself.

theres no doubt this can be "useful". but it was not exactly what i expected. if ya know what i mean....
---------------------------------------------------------------

about the current, i think i posted this way back when on one of the threads, but it may have been passed over.. anyhow

i dont know what kind of scope you're using, but if you can set it to show you current amplitude, and set the divisions way high
it basically looks like a voltage spike. a triangle-wave

i believe it has to do with the time-delay from the current establishing the magnetic field through the core, and the subsequent field collapse.

the DMM is showing you the mean-current, which sits about where the green line is.