Author Topic: Joule Thief 101 (Read 944413 times)

tinman · « **Reply #1260 on:** March 28, 2016, 07:05:22 PM »

Quote from: picowatt on March 28, 2016, 05:38:51 PM

In these recent captures, there is no evidence that the base voltage is exceeding the Vbe breakdown voltage (I believe Tinman measured his 2N3055's as breaking down around 20 volts)

PW

To quote post 1219
The Breakdown voltage of the 2n3055s i use is 15.3 volt's.

Quote

Exceeding the base-emitter reverse breakdown voltage of Q1 does not turn on the collector-emitter junction.

Correct.
I am not sure why MH is talking about collector/base junction breakdown.

Quote

The spike seen in the recent "death spike" capture looks more like a Miller current spike than as being due to Vbe breakdown current.

That is also correct !some what! PW.
If we look at the scope shots below,even though the voltage is above the 15.3 volt breakdown voltage,current continues to flow below the 15.3 volt threshold. But i believe that the bulk of this! so called! current flow is happening around the breakdown voltage limit.

The first scope shot is across the base/emitter junction,and shows a voltage value peak of about 20 volt's.
The second scope shot is from across the 1k base resistor,and you can see that current is (apparently

) still flowing well below the 15.3 volt junction breakdown threshold.

!BUT! there is something odd with this !so called! current flow--but i am still looking into that.

Brad

picowatt · « **Reply #1261 on:** March 28, 2016, 07:09:49 PM »

Tinman,

As you change the scope ground connection point, different portions of the circuit become dynamic with respect to stray capacitance (mainly to your bench top or nearby equip). This can produce some "not really there" spikes or other anomalies as the strays are charge/discharged.

Consider using a short and neat circuit layout and placing the entire circuit on a piece of glass/plexiglass elevated above your bench by insulators.

PW

picowatt · « **Reply #1262 on:** March 28, 2016, 07:21:23 PM »

Quote from: tinman on March 28, 2016, 07:05:22 PM

If we look at the scope shots below,even though the voltage is above the 15.3 volt breakdown voltage,current continues to flow below the 15.3 volt threshold. But i believe that the bulk of this! so called! current flow is happening around the breakdown voltage limit.

The first scope shot is across the base/emitter junction,and shows a voltage value peak of about 20 volt's.
The second scope shot is from across the 1k base resistor,and you can see that current is (apparently ) still flowing well below the 15.3 volt junction breakdown threshold.

!BUT! there is something odd with this !so called! current flow--but i am still looking into that.

It is becoming difficult, if not impossible impossible, to keep track of what the scope captures are depicting. This has become more of a problem as the scope ground is also being relocated in different captures.

Consider posting a schematic of test probe locations with each of your captures as test points change.

A schematic of test point locations for this last set of captures would be handy.

PW

MileHigh · « **Reply #1263 on:** March 28, 2016, 07:23:49 PM »

Brad:

Quote

Whats not suprising, is that you dont know that voltage appears across an inductor,before current flows through it. If you did,you wouldnt be asking as to why the two time basses are different.

Why do you say such nonsensical trash talk idiocy like the comments about voltage leading current? Why do you lower yourself into the gutter and make such a fool of yourself like that? It just destroys your credibility and why should anybody trust you for anything when you play ridiculous mind games like that? It feels like a crank call from a 12-year-old sometimes.

I never said anything about the time bases being different. Brain ricochet! I asked you why the periods are different and you did not offer up any kind of real explanation.

I am getting exhausted with the stream-of-consciousness/churning spaghetti/brain-ricochet talk from you. It's like you need a bloody Google translator just for you. Throw in the ridiculous immature trash talk that is embarrassing and you are left with Dr. Brainfry on overload. Perhaps later I will try to deal with it but not now. I get a headache just thinking about it.

MileHigh

SeaMonkey · « **Reply #1264 on:** March 28, 2016, 10:11:47 PM »

Quote from: Miles Higher

I am getting exhausted with the stream-of-consciousness/churning spaghetti/brain-ricochet talk from you. It's like you need a bloody Google translator just for you. Throw in the ridiculous immature trash talk that is embarrassing and you are left with Dr. Brainfry on overload. Perhaps later I will try to deal with it but not now. I get a headache just thinking about it.

Gosh Miles!

That sort of a response is troubling indeed!

Let us hope that you're not going into a "breakdown"
mode and that perhaps a little rest might restore your
vitality and sense of reason?

The Tin Man makes perfect sense to me.

tinman · « **Reply #1265 on:** March 29, 2016, 12:52:39 AM »

Quote from: picowatt on March 28, 2016, 07:09:49 PM

Tinman,

As you change the scope ground connection point, different portions of the circuit become dynamic with respect to stray capacitance (mainly to your bench top or nearby equip). This can produce some "not really there" spikes or other anomalies as the strays are charge/discharged.

Consider using a short and neat circuit layout and placing the entire circuit on a piece of glass/plexiglass elevated above your bench by insulators.

PW

I have the circuit pretty much as small as i can get it,but i do believe that the voltage seen across the base resistor is not showing a reverse current flow,but is more some sort of artifact that you mention above.

As i said in my previous post,i am looking into it a little further,and will post my findings here when they become apparent.

I have been trying to keep up with schematics along with every scope shot,but it was late last night,and it gets hard to keep up,so i post a quick description instead of a schematic with every scope shot.

Brad

Magluvin · « **Reply #1266 on:** March 29, 2016, 12:57:45 AM »

Just found my JT with the 11t by 11t transformer running low and turning on and off, like 10sec on 30 sec off. Running the 33 ohm resistor, 3904, the batt is at .44v when it dies out then the battery recoups and back on. Ill try to eliminate and the resistor and see if I can get it to stay on below .44v.

Anyway, it seems to be a battery cycling thing if the jt is blinking, so far.

Just ditched the resistor and goes down to .43v, cuts out, and back on at .45v. seems odd. I might expect the need to go higher before coming back on.

Mags

tinman · « **Reply #1267 on:** March 29, 2016, 01:29:13 AM »

author=MileHigh link=topic=8341.msg478791#msg478791 date=1459185829]

MileHigh

Quote

Why do you say such nonsensical trash talk idiocy like the comments about voltage leading current? Why do you lower yourself into the gutter and make such a fool of yourself like that? It just destroys your credibility and why should anybody trust you for anything when you play ridiculous mind games like that? It feels like a crank call from a 12-year-old sometimes.
I never said anything about the time bases being different. Brain ricochet! I asked you why the periods are different and you did not offer up any kind of real explanation.

Quote from you MH : The cycle time is 75 microseconds for one trace (spike) and 60 microseconds for the second trace (collector).

First off,you have that ass about,where the cycle time for the spike(blue trace as marked with attached schematic) is about 45uS,and as i said,this is from a voltage source,and the scope is being triggered from that voltage source.
The yellow trace is across the 1 ohm CVR,and there for is reading the voltage drop across that 1 ohm CVR,and so that is from a current source.
The two dips in the traces in the center of the scope screen,are showing you the OFF period. The blue trace shows a positive voltage(take note of where zero volts is for blue trace)-transistor switching on,before the yellow trace shows a voltage rise,as that trace is measuring the voltage drop across the CVR,and as current lag's voltage,the yellow trace will show a longer off time period than the blue trace.

As i clearly stated in my reply,the difference is because voltage lead's current when it comes to inductors. Quoting my post where you said i left no explanation
Quote: The spike trace--blue base trace is shorter than than the collector current trace-the yellow trace.
Can you not follow a very simple schematic and scope placements?--do you not know that voltage leads current in an inductor,and thus the reason for the blue trace(base trace) being slightly shorter than the yellow trace-collector current trace in time period-->voltage leads current MH-you doofus.

Quote

I am getting exhausted with the stream-of-consciousness/churning spaghetti/brain-ricochet talk from you.

From me ??>
Things are clearly explained to you,but you say they are not.
Then you say that i agreed with you and PW on the collector/base breakdown thing--but i did not do anything of a kind.
You first agree on emitter/base breakdown voltage being the cause of the reverse current flow(that i still think is not there),and then out of no where you come up with plan B,and switch to some !unicorn! collector/base breakdown voltage claim.

Quote

It's like you need a bloody Google translator just for you.

For me?.
How did we go from emitter/base breakdown voltage, to collector/base breakdown voltage??
Where did that translation go,and how did it come about?
I never agreed to anything about collector/base breakdown voltage's,as the V/cb for the 2n3055 is over 60 volt's.

Quote

Throw in the ridiculous immature trash talk that is embarrassing and you are left with Dr. Brainfry on overload. Perhaps later I will try to deal with it but not now. I get a headache just thinking about it.

You are in need of a break MH.
One minute we are talking about the breakdown voltage of the emitter/base junction--which i was in agreeance with--along with the junction capacitance being the first order of events for the reverse base current flow,and then out of the blue,you switch to collector/base junction breakdown.

You have people running all over the place here MH,and you are getting hard to follow,with these leap's from one theory ,to the next.

I tried to show you that there was no reverse current flow spike at the collector,and you just continued to insult me.
I then explained as to why the two time periods are different,and still you kept insulting me.
I then ask you to go back and read what PW said,and still you kept insulting me-as can be seen with nearly every reply you make to me,and so i give them right back to you.

Why have you switch to collector/base breakdown voltage?. Is this another of your attempts to catch me out?--another loaded question?.

You continue to insult me every chance you get,but as you can see,it is you that is wrong every time.
So please stop your insults toward me--at least until !you! get it right.

Brad

TinselKoala · « **Reply #1268 on:** March 29, 2016, 05:06:18 AM »

Quote from: tinman on March 29, 2016, 12:52:39 AM

I have the circuit pretty much as small as i can get it,but i do believe that the voltage seen across the base resistor is not showing a reverse current flow,but is more some sort of artifact that you mention above.

As i said in my previous post,i am looking into it a little further,and will post my findings here when they become apparent.

I have been trying to keep up with schematics along with every scope shot,but it was late last night,and it gets hard to keep up,so i post a quick description instead of a schematic with every scope shot.

Brad

Here's a scopeshot you might find interesting.

MileHigh · « **Reply #1269 on:** March 29, 2016, 05:49:49 AM »

Brad:

This is quickly approaching a dead end.

You said this:

The first scope shot is of your death spike.
The second scope shot is showing the trace over a 1 ohm CVR at the collector.

I responded with this:

The cycle time is 75 microseconds for one trace (spike) and 60 microseconds for the second trace (collector).

Your response was this:

Quote

First off,you have that ass about,where the cycle time for the spike(blue trace as marked with attached schematic) is about 45uS,and as i said,this is from a voltage source,and the scope is being triggered from that voltage source.
The yellow trace is across the 1 ohm CVR,and there for is reading the voltage drop across that 1 ohm CVR,and so that is from a current source.
The two dips in the traces in the center of the scope screen,are showing you the OFF period. The blue trace shows a positive voltage(take note of where zero volts is for blue trace)-transistor switching on,before the yellow trace shows a voltage rise,as that trace is measuring the voltage drop across the CVR,and as current lag's voltage,the yellow trace will show a longer off time period than the blue trace.

As i clearly stated in my reply,the difference is because voltage lead's current when it comes to inductors. Quoting my post where you said i left no explanation
Quote: The spike trace--blue base trace is shorter than than the collector current trace-the yellow trace.
Can you not follow a very simple schematic and scope placements?--do you not know that voltage leads current in an inductor,and thus the reason for the blue trace(base trace) being slightly shorter than the yellow trace-collector current trace in time period-->voltage leads current MH-you doofus.

Essentially everything in that response is wasted energy, a bunch of mush. The reason it is mush is because you did not make the connection between "death spike" and "spike" and "CVR at the collector" and "collector" and you are talking about something completely different.

The gross difference in periods between the two captures suggested to me that your circuit had changed or the scope probe was changing the circuit frequency though loading. Incidentally, putting the scope ground on the transistor collector node is a somewhat bizarre thing to do. But of course you didn't even mention the different periods or different voltages when you posted the captures and only after the fact did you mention that the battery voltage was significantly different between the two captures which explained the different frequencies. It's mass confusion.

It's exhausting and it never ends.

tinman · « **Reply #1270 on:** March 29, 2016, 06:14:48 AM »

Quote from: MileHigh on March 29, 2016, 05:49:49 AM

Brad:

This is quickly approaching a dead end.

You said this:

The first scope shot is of your death spike.
The second scope shot is showing the trace over a 1 ohm CVR at the collector.

I responded with this:

The cycle time is 75 microseconds for one trace (spike) and 60 microseconds for the second trace (collector).

Your response was this:

Essentially everything in that response is wasted energy, a bunch of mush. The reason it is mush is because you did not make the connection between "death spike" and "spike" and "CVR at the collector" and "collector" and you are talking about something completely different.

The gross difference in periods between the two captures suggested to me that your circuit had changed or the scope probe was changing the circuit frequency though loading. Incidentally, putting the scope ground on the transistor collector node is a somewhat bizarre thing to do. But of course you didn't even mention the different periods or different voltages when you posted the captures and only after the fact did you mention that the battery voltage was significantly different between the two captures which explained the different frequencies. It's mass confusion.

It's exhausting and it never ends.

MH
As I stated, the spike shot was across the 1k resistor on the base. This makes it a current source. This voltage will come after voltage seen across the inductor.
The second scope shot is as per schematic attached to that scope shot.

The point of the test was to show you that there was no current flow through the collector during the off time of the transistor.
I am also confused as to why or how you went from emitter/base junction breakdown, to collector/base junction breakdown. Just out of the blue, you make this switch, and I was doing my best to check your claim, and try to work out where this switch came from.
I then spent more of my time tracing back through the thread, trying to find some posts I must have missed by PW.
I still have no answer from you for the reason for the switch from emitter/base breakdown, tocollector/base breakdown.

You asked Chet to complete the current path due to this collector/base junction breakdown, and you abused me for saying it is non existent.

So please tell me what I am suppose to see at the collector during the off period of the transistor.?

Brad

tinman · « **Reply #1271 on:** March 29, 2016, 06:23:38 AM »

I am at work ATM, and will not know the below to be true until I get home, and carry out further test.

After thinking about it for some time, I believe the reverse voltage spike seen across the base resistor---is a phantom voltage, and no actual current is flowing through the resistor as the scope may suggest.
It is either due to the miller effect between the base/emitter or base/collector junction, that make the actual value look much larger than it is.
The second possible explanation is as PW said--there is some sort of capacitive conection with the scope and leads.

All this can be checked using my FG to simulate the pulses across the 1k resistor.

Brad

sm0ky2 · « **Reply #1272 on:** March 29, 2016, 07:20:16 AM »

Quote from: Pirate88179 on March 27, 2016, 06:28:57 PM

https://www.youtube.com/watch?v=K53beWYdIpc

How to make an authentic Joule Thief. By BigClive

Clive is the guy that coined the term Joule Thief way back when and here he shows how to make a "real" one.

Bill

PS He also gives a detailed explanation on how the circuit works.

Thanks Bill

I don't know how much you remember of this, but we were playing with this circuit, prior to the 1999 magazine article.
What was sent to the magazine, and later became the "Joule Thief",
is a variant of BruceTPU's oscillator circuit.

I think you were there for that. '97 maybe?

sm0ky2 · « **Reply #1273 on:** March 29, 2016, 07:25:53 AM »

Quote from: MileHigh on March 28, 2016, 07:23:49 PM

Brad:

Why do you say such nonsensical trash talk idiocy like the comments about voltage leading current? Why do you lower yourself into the gutter and make such a fool of yourself like that? It just destroys your credibility and why should anybody trust you for anything when you play ridiculous mind games like that? It feels like a crank call from a 12-year-old sometimes.

I never said anything about the time bases being different. Brain ricochet! I asked you why the periods are different and you did not offer up any kind of real explanation.

I am getting exhausted with the stream-of-consciousness/churning spaghetti/brain-ricochet talk from you. It's like you need a bloody Google translator just for you. Throw in the ridiculous immature trash talk that is embarrassing and you are left with Dr. Brainfry on overload. Perhaps later I will try to deal with it but not now. I get a headache just thinking about it.

MileHigh

this is the kind of nonsensical argumentation that I am talking about....

You could have just as easily stated that it is the current that leads the voltage,
then explained why.

I think that would be the more productive approach.....

wattsup · « **Reply #1274 on:** March 29, 2016, 02:58:12 PM »

@tinman

The scopes of base/collector junction show 20 something volts so one side of the scope probe (ground or probe) had to receive 0 volts (or near 0) while the other side had to receive the higher volts to produce that scope shot. So where did that 0 come from while the transistor is on (closed).

As per my last post, can you please as per your test "explain 2.JPG" where you have the cvr placed between the collector and L1, now try it with the cvr between the emitter and ground. That's the one I want to see and maybe even you too. Then put the cvr between the LED ground and ground and do it again. If you can do that, it will help explain this circuit.

So in essence the present location of the LED is the main bummer in this circuitry always bleeding and not helping to find the real (stabilized) circuit effect because it is placed parallel to the collector/emitter.

Here is a curiosity? Try repositioning the LED between L1 and the collector. Since the LED would require a ground from the collector, what will happen? Will it light or not? I can get into more detail showing the actual circuit flow with dots, if these tests can be done. The "effect pattern" because it is not really a "flow pattern", because there is not really any flow, should marry perfectly to the scopes wave pattern.

Then, for more confirmation try with the LED between the emitter and ground and this will show if positive will now pass the transistor when it is on. The LED is good because it can delimit up to where the negative can reach before it is consumed with the positive from the other side.

The final test is across the L1 to provide the full effect during those pulses. With these I can draw out the effect patterns to match the wave forms and it will amaze you to know what is really happening in that circuit. No EE Kool-Aid but fully logical and fully provable.

@MH

EE Kool-Aid is fine for our day to day toys but for OU we need something better. Better tasting, more nutritious and no additives. The Kool-Aid is stunting our growth where growth is required to pass to the next level. Otherwise, maybe you can find a better way to advance but downing jugs of Kool-Aid all day will not do it.

Also, I think your free attacks on @tinman are highly unwarranted and will soon approach some level of indecency. Best you stop it. We are all in the same damn boat. All a bunch of ignorant assholes, all of us trying to cope with both observation and logic. Your logic is not God given, it is man derived and thus highly fallible. The day you accept it is the day you will advance as well. If you don't think @tinman has a shred of credibility here, then why waste your time? This one little furty circuit could help learn mountains of information but only if you ask the right questions.

wattsup

News:

User Menu

Author Topic: Joule Thief 101 (Read 944413 times)

tinman

Re: Joule Thief 101

picowatt

Re: Joule Thief 101

picowatt

Re: Joule Thief 101

MileHigh

Re: Joule Thief 101

SeaMonkey

Re: Joule Thief 101

tinman

Re: Joule Thief 101

Magluvin

Re: Joule Thief 101

tinman

Re: Joule Thief 101

TinselKoala

Re: Joule Thief 101

MileHigh

Re: Joule Thief 101

tinman

Re: Joule Thief 101

tinman

Re: Joule Thief 101

sm0ky2

Re: Joule Thief 101

sm0ky2

Re: Joule Thief 101

wattsup

Re: Joule Thief 101