Joule Thief 101

[sm0ky2]

Below is a scope shot that I produced last night. The battery was near fresh at 1.44v when I started. When I finally got to this point I let it set over night. At lunch today the batter showed a solid 1.46v.  But that could be due to many odd things. But it was nice to see.  Seems like a nice clean sine to me.

Here I had reverted back to my original transformer because the choke coil prewound with 2 windings wouldnt show a clean sine before it dies out. The only way I could come close to a sine with either transformer was to add the shorted winding. The original was better at it with these low inputs.

There isnt much range in which I can produce the clean sine before dying out. So I figure the best way around that is to up the input voltage. Will be doing that tonight. Got some various larger npn transistors to work with if the 3904 blows on me.

I need to get some pots with plastic casings and control shafts. The 100k Im using at the moment is the standard metal casing with split aluminum shaft that if I touch it it offsets tuning at these near 1mhz freq, even with a big rubber pipe end cap, if I touch the rubber the effect is still there.

Also, I did a check on the circuit without the led and it still operates. Not exactly the same freq. but still the same other than without the led there is no clamping of the spike.

Mags

you can "zoom in" on the scope, just before the lower peak, you'll notice a blip, this is a wavefront hitting slightly out of phase.
advanced a bit, probably due to an inductance somewhere not being exactly what it needs. you may or may not be able to get rid of that, but it would be interesting to play with things and figure out exactly where its coming from.

If your scope can show two signals at once, probe different spots looking for a signal that peaks just in front of the one your are scoping here.


Also, for more precision:
If you can OHM the VR near where you have it adjusted above, and try to find one in that range, that has a higher sensitivity.
for instance, if it reads around 72k with that transistor, to get a clean(er) signal, look for a VR that has a range of say:
60-90k (30k range, vs 70k range of a 30-100k trim pot)
this may give you a better adjustability.

[sm0ky2]

a side though, if your scope doesn't have multiple channels like that.

you can take like a 100 ohm resistor, or something small.
and make a jumper wire while you watch the signal.

B<-> (-)
or
B <-> E

and see if you can get that thing to increase or decrease. If it creates a 3rd peak, you know that's not it.
if it changes the secondary peak that is shown in your image above, you can adjust a few parameters there depending on which side of the circuit it is on.

thing to remember, the coil doesn't care where the impedance comes from, it can be simply a resistance, or another coil, or the reactance factor of another load...

at times, you can dig out a blip by simply adding a thin metal plate between the battery ground and the connecting wire.
(a battery holder already does this, provided the plate is the proper thickness needed in the circuit)

other times you may find the coil itself is interfering with itself, because of the ends of the start or ending loops,
or a spacing in the middle of the torroid may be bent

the important thing is to measure everything, and understand how they affect each other.

something like changing the LED, can increase voltage, and lower current.
while the freq. may not change, because the parameter that was altered was a function of the led, not the transistor.

it can get really deep, if you don't have a set circuit to examine.
one thing my brother played with for a while was finding LED's that had the same internal resistance,
but different capacitances.

This allowed him to take advantage of longer "on" times of the LED, without disrupting the waveform.
The same approach is used with dual-phase motors as the load.

-------------------------------------------------------------------------------------------------------------------------------------------

[not entirely ignoring the questions about me building a joule thief to display here]

apart from the two I linked from my youtube acct,

I do not have any to display at this time.
the last JT I built, was several years ago, after I lost my lab.
It flew.

2" microwave torroid, superbright LED in the center, in fashion of the Iron Man arc reactor.
a micromotor, with an oversized propeller.
following several crashes into the ceiling, my roommate at the time made me sign an addendum to our lease agreement, stating that I would not create dangerous technologies inside our apartment.

I haven't built one since.

I could probably indulge in experiments with the Armstrong oscillator for an endless series of lifetimes.

but I have not the free time, nor the proper equipment at this particular point in my life.
maybe after the little ones are grown, and I have amounted another 1/4 mil to sink into a proper lab.

but in all honesty, people like Bill or TK make anything I build look like a kid with crayons standing beneath Picasso.

[tinman]

 author=MileHigh link=topic=8341.msg474927#msg474927 date=1455958588]

Yes I watched the clip again and saw a different schematic in the clip as compared to the one you posted twice.  So you post a schematic in the thread twice that doesn't match what you are showing in the clip and you are giving me a hard time for not remembering your posted comment and the schematic shown in the clip.
In both the mismatched schematic and the schematic shown in the clip you can't be bothered to show the dot convention for the transformer and in the past I told you there was a major mistake in the schematic and after that I told you that you weren't showing the dot convention.  And of course in fact there are two major mistakes in the posted schematic.

Well as you need to be spoon feed MH,o hope the schematic below passes inspection-regardless of the fact that you should have associated the schematic in the video with the post that went with the video. If it dose not meet your requirements-->too bad,as thats all your getting. If you new anything about the SSG circuit's,then you would not have needed a dot schematic,as the conections give the answer as to what way the bifilar coil is wound.

"But even when you do this" my ass.  Look in the mirror because your self-documentation level is abysmal.  If you tried to pull that kind of nonsense off on a serious electronics forum they would probably be merciless and slice you to pieces.

Thankfully everyone here(other than your self) new exactly what i was showing,and have no problem with the schematic.

Looking at the correct schematic, since you have no base resistor at all, it is pretty likely the case that you are passing way too much current through the transistor to switch it on.  So that could represent a significant loss in energy.  On the schematic you posted in the thread you state 50-60 turns and that length of wire and the associated resistance represents more lost energy during the transistor ON time.  I am assuming that that didn't even occur to you.

No,it did not MH,and this is where bench time once again win's hands down. 50 turn's of .4mm wire oscillating at that frequency--wonder how much the resistance rises due to the skin effect at that frequency ?. The current flowing through L2 would have to be about 8 time's less than the current flowing through L1,due to there being an 8:1 winding ratio--would it not?. But the current would be raised i guess,as the voltage is dropped. But what is that voltage drop to that of the supply voltage?. Oh wait,the supply voltage was less than the voltage at the base of the transistor-about 3 times less . There is also the fact that the transistor stayed at room temperature,which would indicate very little power dissipated as heat.
Are you sure you are full bottles on EE MH ?.
I think i am going to put a low value resistor on the base in that circuit,and we will see just how little current is flowing to the base--would you like me to do that MH ?-->just so as we can see if your comment has any merit at all.

From looking at the scope traces and by doing a bit more simple testing you could have determined just how much energy is lost to keep the transistor ON during a single cycle.  Since you are making better efficiency claims about this setup you would think that you would want to try to at least make some measurements along those lines and see how that compares to a standard Joule Thief, but you don't.  The setup is right there on your bench and you could simply configure it as a conventional Joule Thief and then as a "zero resistance" Joule Thief and make some measurements, but you don't.  Welcome to Brad's world of stream-of-consciousness electronics.

Mmm-you mean like using a super cap charged to a certain voltage,then doing a rundown time test to say .4 volt's on the cap,while measuring the light output with my lux meter ?.Hell MH,i would never had thought about doing that before making my claim
Of course you will want video proof--yeah?
Let me know,and i will waste a little more of my time on you.

I also notice there is still a decaying voltage/current waveform during the coil discharge cycle into the LED.  So the idea I floated about possible coil discharge efficiency issues due to that factor is also happening for this setup.  But just to play your game, I guess you are not reading the thread.

MH
Could you draw up the current loop during the flyback pulse of each circuit,and post them here please?.

So all in all, I am not convinced there is anything better in your "zero resistance" Joule Thief setup and I strongly suspect that it would be less efficient than a regular Joule Thief because of excessive current used to switch on the transistor.  Beyond that, we have only anecdotal evidence for your claim - "You said so."

So lets do it MH-->your beloved circuit V my circuit. You V me-->a joule thief challenge
If there is disbelief in the results given by both of us,we could post our circuits to an agreed third party for testing,and let them post the results them self.
What do you say MH ?


Brad

[sm0ky2]

@ Brad

I can't help but point out here, that he skillfully avoided my question as to whether or not he owned one....

[MileHigh]

@ Brad

I can't help but point out here, that he skillfully avoided my question as to whether or not he owned one....

I am trembling in my tracks but I will answer your question:  No I don't "own" a Joule Thief and never have and never will.

Think about all of the questions and issues that I raised with you that you "skillfully avoided."

finding LED's that had the same internal resistance

Ha ha ha - busted.

Brad:  No to the competition, but you already knew the answer to that question, it's just theater on your part.  I understand Bedini motors inside-out, never built one, just more theater on your part.

Same thing for you, think about all of the questions and issues that I raised with you that you "skillfully avoided."

MileHigh