Joule Thief 101

[MileHigh]

Brad:

I am probably jumping off the bus pretty soon.  This whole ordeal started when Smoky2 said he was going to coach you and Magluvin to make a "resonant Joule Thief."  He claimed that something really special was going to happen with the "tuning" and his brother had some "resonant Joule Thieves" that ran "permanently."  I said the whole thing was complete and utter crap, remember?  Then that started the discussion about resonance and the whole nine yards.

You need a foil like me, because without that you and the boys (no gurus participating) create this "enabling" environment were "everything is awesome" and you deceive each other by never having the guts to correct each other.  That's why you learned very little over the past six years and with me being your foil you kept on being stung in the ass and it was good for you, no matter how upset and angry you got.  Without a foil like me to upset the awesome party, ultimately that's why you and many of the other boys couldn't solve for a simple circuit consisting of a power supply and one single solitary component.  Or you had all been discussing resonance for years but couldn't properly define it or answer two simple questions about it even through you were given a month and a half to answer the questions.  That's comfortably awesome numbness.

Ultimately, I don't really care.  You guys do whatever the hell it is that you want to do.  It's just such a pain to see all of this sleepwalking going on sometimes.

For example, when I look at most of your recently posted Joule Thief scope captures, the current waveform is a saw-tooth or triangle waveform.  To me it appears that that's not normal Joule Thief operation.  It's still lighting the LED of course but it looks to me like the Joule Thief operation has broken down and resembles a generic oscillator more than a Joule Thief.  I have a feeling that there is a very good chance that that didn't even occur to you.

I am perfectly content to just sit and watch what you and TK produce.

MileHigh

[tinman]

 author=MileHigh link=topic=8341.msg488361#msg488361 date=1468349727]

I am perfectly content to just sit and watch what you and TK produce.

MileHigh

I am probably jumping off the bus pretty soon.  This whole ordeal started when Smoky2 said he was going to coach you and Magluvin to make a "resonant Joule Thief."  He claimed that something really special was going to happen with the "tuning" and his brother had some "resonant Joule Thieves" that ran "permanently."  I said the whole thing was complete and utter crap, remember?  Then that started the discussion about resonance and the whole nine yards.

I would have to agree that the whole resonant thing went south,but i think Smoky2 was hoping that we could find or design that resonant JT,and see what the outcome was.

For example, when I look at most of your recently posted Joule Thief scope captures, the current waveform is a saw-tooth or triangle waveform.  To me it appears that that's not normal Joule Thief operation.  It's still lighting the LED of course but it looks to me like the Joule Thief operation has broken down and resembles a generic oscillator more than a Joule Thief.  I have a feeling that there is a very good chance that that didn't even occur to you.

There is a reason for the saw-tooth wave form MH,and perhaps you may have noticed when the transition took place from that of a current wave form that would be seen on a normal JT.
The circuit being used is your JT circuit (circuit number 1). But in saying that,the current trace is not really that different to what my other JTs show--see pic below.
The only change i notice,is that the waveform in the latest scope shot's is more !crisp!,where the trace rises and falls in a much more linear way.

I will do some comparison tests tonight--as soon as the light box is completed,and post the relevant scope shot's,and efficiency data.

Brad

[MileHigh]

Well you can see in your latest scope capture that the current is rising and falling in an exponential-type waveform which is what you expect to see for a Joule Thief.  You can also see a little tell-tale signature of a switching event in that current waveform, suggesting that the Joule Thief is switching normally.  You don't normally expect to see a saw-tooth or triangle waveform for a normally operating Joule Thief.

<<<  I am going to qualify the above paragraph for when it comes to watching what happens when a fresh battery transitions to a depleted battery when powering a Joule Thief.  I don't really have a handle on how the current waveform will change as the output impedance of the battery starts to slowly creep up.  >>>

When looking at the latest capture, the unexplained vertical spike noise on the older current waveform captures stands out even more and requires investigation.

You are back to your bad habits again and you are not showing where the test points are for that latest scope capture.  Weren't you embarrassed when you posted a scope capture a few weeks back and Picowatt told you that he couldn't make head or tail of it and asked you for the corresponding schematic?

If you don't at least sample what the waveform looks like at the base resistor input (bottom of L2) every time you fiddle with a new Joule Thief configuration then you are making a mistake.  Same thing applies for the transistor collector waveform to see if you are getting clean switching or not.

[tinman]

Well you can see in your latest scope capture that the current is rising and falling in an exponential-type waveform which is what you expect to see for a Joule Thief.  You can also see a little tell-tale signature of a switching event in that current waveform, suggesting that the Joule Thief is switching normally.  You don't normally expect to see a saw-tooth or triangle waveform for a normally operating Joule Thief.

<<<  I am going to qualify the above paragraph for when it comes to watching what happens when a fresh battery transitions to a depleted battery when powering a Joule Thief.  I don't really have a handle on how the current waveform will change as the output impedance of the battery starts to slowly creep up.  >>>

When looking at the latest capture, the unexplained vertical spike noise on the older current waveform captures stands out even more and requires investigation.

You are back to your bad habits again and you are not showing where the test points are for that latest scope capture.  Weren't you embarrassed when you posted a scope capture a few weeks back and Picowatt told you that he couldn't make head or tail of it and asked you for the corresponding schematic?

If you don't at least sample what the waveform looks like at the base resistor input (bottom of L2) every time you fiddle with a new Joule Thief configuration then you are making a mistake.  Same thing applies for the transistor collector waveform to see if you are getting clean switching or not.

MH
The test points for the latest scope capture are clearly shown in the attached schematic,so i dont know what you are talking about.
CH 1 is obviously inverted,when you take note of scope probe polarity on the attached schematic.

We will sort out the two differing wave forms for the current trace soon enough,and the same go's  with the base trace. You seem to be back to that bad habbit of wanting everything all at once,but i can only do so much each working night-friday and Saturday nights  i get a bit more time.


Brad

[MileHigh]

No Brad, the test points are ambiguous.  I don't see a definitive ground connection.  I don't see any probe colours matching the scope channel colours like I am used to seeing before.

Here is the bloody point:  Every time I look at a scope capture with a provided schematic from you I don't want to have to play a guessing game and spend 90 seconds figuring out what is what.  I don't give a damn if you say it is "obvious."  It is pretty straightforward in this case but I still don't give a damn.  It's a pain in the ass.

I literally just saw the text "CVR 1 trace" in the upper right quadrant of the image after staring at the image about 20 times.

You took the trouble to stitch together your scope capture with the schematic using an image editing program.  You couldn't be bothered to spend just one minute while you were in the image editing program to label where the scope channel was connected in a clear and unambiguous manner.

Look at the attached image.  I updated your graphic.  I can look at that graphic and in 1/2 second I see exactly what is going on, no muss, no fuss, no bullshit looking at "options."