Joule Thief

[broli]

 
  This could harm my scope, so I put it away, and I reduced the collector coil turns to 8 then 7 and tried to light things.
 

  Forgive my ignornca but what if you used a high voltage capacitor combined with a high voltage diode both from a microwave oven to measure these high voltages? If this works you could calibrate this by measuring a safer voltage with the scope and capacitor and compare numbers.

@all

http://www.youtube.com/watch?v=1uN-yNgtt4c

Its a must see from lidmotor again a new one.

  Tuning inductance in realtime like that is super handy.

[jeanna]

...
Let's see, Fuji...about 350 volts or so.....
Jeanna's coil...1439 volts.

Yes, I would say that you have exceeded your goal, ha ha.  Very nice work there Jeanna!  I will drink a toast to your success when I go off surveillance duty later this evening.

Thank you Bill,
Do you really think the fuji is 350v?
We need at least 450v to light a tube. I think there must be higher spikes than 350v in that camera circuit.

I am wondering about this 48"  tube of yours, though. What is making this light go on?

I have been wondering how to add the final cap as the fuji circuit has.
I am wondering if this is how you  can light the higher wattage tube??

===
I just tried the little MK1-Jeanna hybrid#2 which lit the string 'o leds the other day.
The voltage on that one is not as high with the TIP3055 as it is with the 2N3904.

But the part I am finding REALLY interesting is that the 'ring' amount of base resistance appears to be more specific to the transistor than the core or the winds on the core.

The TIP3055 ring base resistor amount is in the neighborhood of 50ohms.
So, no matter what core I am using,  the best results come from a base resistor at about 50 ohm.
It happens that this is not the best frequency for the little high frequency core, but it is the best this transistor can do.

@Broli... don't worry about this stuff for now.
I am sorry to hear about your only led, but it is actually good news, sort of.

Did you mean to use a pnp?
Most of us are using a npn which are cheaper and more generally available.
Radio shack sells a pack of rather decent assorted  colored leds for a few bucks.
Of course it is now the time you can get cheap led christmas light strings. I don't think they are better or easier to use, but depending on where you live, they may be more available.

jeanna

[resonanceman]

First...
Thank you xee2.
I have a hard time with it being a w because that is a high permeability, but mine was doing well at 2.27khz which was yesterday's video, and as your table tells, that IS high permeability...

It is strange this conflict in assertions of permeability between you and MK1. You both have good reasons, and you both have results, based on your reasons.

Maybe the permeability doesn't matter as much as the transistor. See below.

@MK1
Thank you for reposting that. It is a good method. Did you also post it on the 101 thread for beginners?

@All,
I made some tests today.
[I knew I had a strange 2N3055. It was reversed from the beginning. The higher number of turns should be on the collector coil but on this my first hand wound fluoro tube lighting circuit, it was reversed.

I must have very funny component karma. ]

So, the other day I bought a TIP3055 from radio shack.
It was this that I used yesterday to light the cfl for my first time and achieved 895v at 2.27khz frequency.

I re-tried the core with the red 240T from last spring but before even starting, I made the turns 2T,10T like yesterday's

It lit the cfl immediately with 240T, so I knew it was the transistor at fault for the struggle last spring.

Then I began to tune the 240T core. I reduced from 10 to 9,8,7, and at 6T the voltage on the red 240T was 1040volts !!!!!!!

So, then I went back to the gold one which has over 300T and again I reduced the collector turns from 10 to 9. The volts went to 1439v.
This could harm my scope, so I put it away, and I reduced the collector coil turns to 8 then 7 and tried to light things.
The cfl doesn't stop flickering. I may need to solder this, because at this kind of volts there is bound to be sparklers. I can feel the tingle when my hand touches a probe directly.
At 7T and  6T, I tried to light that big 48" tube. It didn't light, but  both ENDS lit up immediately and, it got to about 12 inches from the one side with the help of my hand and the magnet trick. Nothing got it to light. This is 36w and will probably light when I have 1900v-2000v
A 105 cap at the base seems to be the best frequency for the resistor resonance.

I will try out the open office spreadsheet and make a table of this for you.
I was in a happy hurry, but I will return.

jeanna

Jeanna

Iam glad to see that the TIP3055  helped



About lighting  the 48
WIth that kind of  voltage  I can't imagine that   you need more .
I think   you need more current  to light  the tube.

You mentioned  several other  JT  coils ....... try  connecting  the  secondarys   of  a couple of those  coils   to the   output   of  the  JT  you  are using .
The  coils  should  be in series ......  connect the secondarys   just  as if  they were  the power lines   going to your load .

If  your secodarys are good sized   they  should  choke  your  output  voltage  some ..........but  give you more current .
When I do this with my LED array  the  voltage  goes down but the LEDs get brighter .
If  you want to limit the voltage  more .......connect  the  primarys of  those  coils for feedback .
CHances  are  good that if you  connect both of the directly for  feedback    you will kill  the output .....you may need to  add a cap in  in series  with one of the  feedback loops. 

check  the output  when  you connect  the  feedback ........once in a whille it jumps  way up .


gary

[gadgetmall]

First...
Thank you xee2.
I have a hard time with it being a w because that is a high permeability, but mine was doing well at 2.27khz which was yesterday's video, and as your table tells, that IS high permeability...

It is strange this conflict in assertions of permeability between you and MK1. You both have good reasons, and you both have results, based on your reasons.

Maybe the permeability doesn't matter as much as the transistor. See below.

@MK1
Thank you for reposting that. It is a good method. Did you also post it on the 101 thread for beginners?

@All,
I made some tests today.
[I knew I had a strange 2N3055. It was reversed from the beginning. The higher number of turns should be on the collector coil but on this my first hand wound fluoro tube lighting circuit, it was reversed.

I must have very funny component karma. ]

So, the other day I bought a TIP3055 from radio shack.
It was this that I used yesterday to light the cfl for my first time and achieved 895v at 2.27khz frequency.

I re-tried the core with the red 240T from last spring but before even starting, I made the turns 2T,10T like yesterday's

It lit the cfl immediately with 240T, so I knew it was the transistor at fault for the struggle last spring.

Then I began to tune the 240T core. I reduced from 10 to 9,8,7, and at 6T the voltage on the red 240T was 1040volts !!!!!!!

So, then I went back to the gold one which has over 300T and again I reduced the collector turns from 10 to 9. The volts went to 1439v.
This could harm my scope, so I put it away, and I reduced the collector coil turns to 8 then 7 and tried to light things.
The cfl doesn't stop flickering. I may need to solder this, because at this kind of volts there is bound to be sparklers. I can feel the tingle when my hand touches a probe directly.
At 7T and  6T, I tried to light that big 48" tube. It didn't light, but  both ENDS lit up immediately and, it got to about 12 inches from the one side with the help of my hand and the magnet trick. Nothing got it to light. This is 36w and will probably light when I have 1900v-2000v
A 105 cap at the base seems to be the best frequency for the resistor resonance.

I will try out the open office spreadsheet and make a table of this for you.
I was in a happy hurry, but I will return.

jeanna

Sweet  Jeannas ! oww. I must know exactly how this one works . I have all night to myself tonight and i want to Wind / three" first . I'll go with what you Pm'ed me and reduce to 6?
Albert

[jeanna]

Jeanna

I am glad to see that the TIP3055  helped

About lighting  the 48
WIth that kind of  voltage  I can't imagine that   you need more .

Right, but it is twice the wattage (36w) I have been lighting (20w). So, It might be reasonable, but I think it needs a pop like the camera circuits, and it doesn't need it to continue.
It IS going on for the end 12 inches of 1 side and the 2 inches on the other side. - until I let go 

These large tubes need a ballast to start up when used in the mains.
[anyway, that was my thinking, --- something like a ballast.]

You mentioned  several other  JT  coils ....... try  connecting  the  secondarys   of  a couple of those  coils   to the   output   of  the  JT  you  are using .
The  coils  should  be in series ......  connect the secondarys   just  as if  they were  the power lines   going to your load .

I will try this. It makes sense to me.
The feedback has never made sense so I will skip it for now.

Thanks for those suggestions. I will try them and report back.

jeanna