Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search
« previous next »
Pages: 1 ... 133 134 135 136 137 [138] 139 140 141 142 143 ... 182   Go Down

Author Topic: Joule Thief 101  (Read 947674 times)

MileHigh

  • Hero Member
  • *****
  • Posts: 7600
Re: Joule Thief 101
« Reply #2055 on: April 18, 2016, 02:04:15 AM »
Quote from: Pirate88179 on April 18, 2016, 12:42:01 AM
This is exactly why most of my JT circuits, including the high voltage units, all have a VR on the base.  I tried to explain this earlier in the topic but, it really does work well for what you want to get out of the circuit.

Lidmotor even used this method on his replication of the Jeanna Light if I am not mistaken.  I actually used 2 VRs in that project...I think he used a rheostat if I remember correctly.

Bill

Well if you are following along with what I am saying, then it means that you start your Joule Thief with the base resistor at the higher value to keep the power drain to a minimum, and then some time later when the battery voltage has decreased to a certain value you lower the value of the base resistor by perhaps 10%-25% to ensure reliable switching to a lower battery voltage.

That's it, there is no brightness control, no twirling of pots to look for a bright spot, just a controlled decrease in the base resistor value to ensure that the switching can operate properly at a lower battery voltage.  There is no "sweet spot" at all.

However, this leads to a question, is it even worth it to go to this kind of trouble?  How much power can you really save compared to your nominal running power using a fixed resistor and how much proportionally longer a run time can you get when you do this?  What if changing the value of the resistor in two steps only saves you 2% on your nominal running power and gives you 2% more run time?  You have to think about it and crunch some numbers for that.

Supposing you crunch the numbers and a variable resistor solution only gives you a 3% overall improvement?  Most people would then say forget it, it's not worth it.  It's not worth the extra complexity or the trouble.  The better solution is to do your experimenting and analysis and determine the optimal resistor value.  The next step is to decide what standard resistor value or combination of standard resistor values to settle on for the final design.

That's the real deal - the real, serious experimentation and analysis and evaluation of the various trade-offs and design choices to arrive at a final Joule Thief design.  From what I can tell, you have never seen anything like that done on this forum.
Logged

Pirate88179

  • elite_member
  • Hero Member
  • ******
  • Posts: 8366
Re: Joule Thief 101
« Reply #2056 on: April 18, 2016, 02:38:29 AM »
Quote from: MileHigh on April 18, 2016, 02:04:15 AM
Well if you are following along with what I am saying, then it means that you start your Joule Thief with the base resistor at the higher value to keep the power drain to a minimum, and then some time later when the battery voltage has decreased to a certain value you lower the value of the base resistor by perhaps 10%-25% to ensure reliable switching to a lower battery voltage.

That's it, there is no brightness control, no twirling of pots to look for a bright spot, just a controlled decrease in the base resistor value to ensure that the switching can operate properly at a lower battery voltage.  There is no "sweet spot" at all.

However, this leads to a question, is it even worth it to go to this kind of trouble?  How much power can you really save compared to your nominal running power using a fixed resistor and how much proportionally longer a run time can you get when you do this?  What if changing the value of the resistor in two steps only saves you 2% on your nominal running power and gives you 2% more run time?  You have to think about it and crunch some numbers for that.

Supposing you crunch the numbers and a variable resistor solution only gives you a 3% overall improvement?  Most people would then say forget it, it's not worth it.  It's not worth the extra complexity or the trouble.  The better solution is to do your experimenting and analysis and determine the optimal resistor value.  The next step is to decide what standard resistor value or combination of standard resistor values to settle on for the final design.

That's the real deal - the real, serious experimentation and analysis and evaluation of the various trade-offs and design choices to arrive at a final Joule Thief design.  From what I can tell, you have never seen anything like that done on this forum.

I follow along with what you are saying, I just totally disagree with it.  I have easily seen 3 times the running time using a vr on the base as without.  I have JT lights here that will run for months using this method.

All of this was discussed years ago in the main JT topic area.  You can unsolder and replace as many resistors as you want during the runtime and I will just turn the vr a little bit.  This method takes less than 1 second and allows you to milk every drop of energy from the battery which was one of the purposes for this circuit to begin with.

No compromises needed...just an adjustment now and then.

Bill
Logged

poynt99

  • TPU-Elite
  • Hero Member
  • *******
  • Posts: 3582
Re: Joule Thief 101
« Reply #2057 on: April 18, 2016, 04:15:48 AM »
Quote from: tinman on April 18, 2016, 12:02:56 AM
The phase shift video.


https://www.youtube.com/watch?v=EhpP7Bmcwhs



Brad

If you block the magnet (but don't change anything else) and run through a wide frequency sweep in the same area, what do you see in terms of the phase relationship?
Logged

MileHigh

  • Hero Member
  • *****
  • Posts: 7600
Re: Joule Thief 101
« Reply #2058 on: April 18, 2016, 04:55:54 AM »
Quote from: Pirate88179 on April 18, 2016, 02:38:29 AM
I follow along with what you are saying, I just totally disagree with it.  I have easily seen 3 times the running time using a vr on the base as without.  I have JT lights here that will run for months using this method.

All of this was discussed years ago in the main JT topic area.  You can unsolder and replace as many resistors as you want during the runtime and I will just turn the vr a little bit.  This method takes less than 1 second and allows you to milk every drop of energy from the battery which was one of the purposes for this circuit to begin with.

No compromises needed...just an adjustment now and then.

Bill

What you are saying is not incompatible with what I am saying.  I am just talking about a Joule Thief in normal operating mode.  I don't know how much energy is left in a battery when the voltage falls below the minimum voltage for normal operating mode and for all I know you can still get a lot of light out of the device in that range.  It just requires more measurements and investigation.
Logged

verpies

  • Hero Member
  • *****
  • Posts: 3473
Re: Joule Thief 101
« Reply #2059 on: April 18, 2016, 05:13:39 AM »
Quote from: tinman on April 18, 2016, 12:02:56 AM
The phase shift video.
https://www.youtube.com/watch?v=EhpP7Bmcwhs
I just got a PM from one of my "fans" directing me here because you are vibrating some ferromagnetic materials near coils, which is a well known pet peeve of mine ;)
I never looked in here before because I expected this thread to contain some boring Joule Thief circuits - not an off-topic device like that.

I am too lazy to read the entire thread, so could you summarize for me what is your investigation about and what are you trying to determine?
Logged

Magluvin

  • Hero Member
  • *****
  • Posts: 5884
Re: Joule Thief 101
« Reply #2060 on: April 18, 2016, 05:19:37 AM »
Quote from: tinman on April 18, 2016, 12:02:56 AM
The phase shift video.


https://www.youtube.com/watch?v=EhpP7Bmcwhs



Brad

Hey Brad

Just thinking. Would the L2 trace lead if its peak is offset to the left of L1 trace peak?  Maybe Im wrong. Just thinking that leading is ahead of others in time. If the trace is left to right, then the L2 peaks before L1.  Either way though, you have shown the shift due to the magnet interaction. ;)

Mags
Logged

tinman

  • Hero Member
  • *****
  • Posts: 5365
Re: Joule Thief 101
« Reply #2061 on: April 18, 2016, 06:51:51 AM »
Quote from: verpies on April 18, 2016, 05:13:39 AM
I just got a PM from one of my "fans" directing me here because you are vibrating some ferromagnetic materials near coils, which is a well known pet peeve of mine ;)
I never looked in here before because I expected this thread to contain some boring Joule Thief circuits - not an off-topic device like that.

I am too lazy to read the entire thread, so could you summarize for me what is your investigation about and what are you trying to determine?

Just have an oscillating  magnet in front of an air core transformer,where i can shift the current phase relationship of the primary and secondary to a state where the secondaries current can lead that of the primaries current. Once at the magnets resonant oscillating frequency,the primaries P/in drops while the secondaries P/out increases by over 2000%

Just looking into the effect ATM
See the mechanical resonance thread for the un/cluttered version of the topic

Brad.
Logged

tinman

  • Hero Member
  • *****
  • Posts: 5365
Re: Joule Thief 101
« Reply #2062 on: April 18, 2016, 06:57:31 AM »
Quote from: poynt99 on April 18, 2016, 04:15:48 AM
If you block the magnet (but don't change anything else) and run through a wide frequency sweep in the same area, what do you see in terms of the phase relationship?

I have only tried raising  the frequency with the magnet away from the coil,never with the magnet in position -but held still. Without the magnet there ,i see little to no change in phse relationship up to 10khz,but do see amplitude changes.

I will try with the magnet in place when i get home from work tonight.


Brad
Logged

tinman

  • Hero Member
  • *****
  • Posts: 5365
Re: Joule Thief 101
« Reply #2063 on: April 18, 2016, 07:04:56 AM »
Quote from: Magluvin on April 18, 2016, 05:19:37 AM
Hey Brad

Just thinking. Would the L2 trace lead if its peak is offset to the left of L1 trace peak?  Maybe Im wrong. Just thinking that leading is ahead of others in time. If the trace is left to right, then the L2 peaks before L1.  Either way though, you have shown the shift due to the magnet interaction. ;)

Mags

The leading trace/phase is to the right-is it not? as the trace starts from the left of the screen,and travels to the right of the screen. So the one furtherest to the right is the one that started to the left first.
---OR i have been looking at scope traces wrong all this time lol.


Brad
Logged

Magluvin

  • Hero Member
  • *****
  • Posts: 5884
Re: Joule Thief 101
« Reply #2064 on: April 18, 2016, 07:11:23 AM »
Quote from: tinman on April 18, 2016, 07:04:56 AM
The leading trace/phase is to the right-is it not? as the trace starts from the left of the screen,and travels to the right of the screen. So the one furtherest to the right is the one that started to the left first.
---OR i have been looking at scope traces wrong all this time lol.


Brad

Lol. I havnt dealt with phase of 2 signals for 25yrs. Hadnt the need to.  But was hoping I was wrong and you could let me know. I might have mistaken you in the vid.

Mags
Logged

Magluvin

  • Hero Member
  • *****
  • Posts: 5884
Re: Joule Thief 101
« Reply #2065 on: April 18, 2016, 07:15:00 AM »
Like I said, either way you still show the phase shift.

Also, is it possible the difference in the 2 scope trace freq numbers, maybe the pendulum is trying to hold its resonant freq while the func gen is over or under that freq?

Mags
Logged

picowatt

  • Hero Member
  • *****
  • Posts: 2039
Re: Joule Thief 101
« Reply #2066 on: April 18, 2016, 07:54:53 AM »
Quote from: tinman on April 18, 2016, 07:04:56 AM
The leading trace/phase is to the right-is it not? as the trace starts from the left of the screen,and travels to the right of the screen. So the one furtherest to the right is the one that started to the left first.
---OR i have been looking at scope traces wrong all this time lol.


Brad

It is not...  The left most trace is the "leading" trace. 

For example, at around 8:46 in your last video, the scope "appears" to be telling you that the yellow trace is lagging the blue trace.

I say "appears" because there are a few issues.  The first is verifying how the transformers are connected phase wise (dot convention) and second, I would want to use the second channel of the FG to generate a marker pulse, modified half cycle, etc, to confirm who is leading who...

I believe using a marker was touched on during discussion of your flywheel version of this setup, which was never instrumented as suggested to observe acceleration and deceleration of the flywheel.  Although possible, it would be a bit more difficult to do so with this new "waving wand" version.

PW   
Logged

MileHigh

  • Hero Member
  • *****
  • Posts: 7600
Re: Joule Thief 101
« Reply #2067 on: April 18, 2016, 08:28:08 AM »
Measure with a micrometer.
Mark with chalk.
Cut with an axe.
Logged

tinman

  • Hero Member
  • *****
  • Posts: 5365
Re: Joule Thief 101
« Reply #2068 on: April 18, 2016, 12:01:59 PM »
Quote from: MileHigh on April 18, 2016, 08:28:08 AM
Measure with a micrometer.
Mark with chalk.
Cut with an axe.

There is no doubt about ya MH--->you are the king clown,and a liar.

Quote
And ultimately I am trying to help Brad.

Brad
Logged

tinman

  • Hero Member
  • *****
  • Posts: 5365
Re: Joule Thief 101
« Reply #2069 on: April 18, 2016, 12:10:06 PM »
Quote from: picowatt on April 18, 2016, 07:54:53 AM
It is not...  The left most trace is the "leading" trace. 

For example, at around 8:46 in your last video, the scope "appears" to be telling you that the yellow trace is lagging the blue trace.

I say "appears" because there are a few issues.  The first is verifying how the transformers are connected phase wise (dot convention) and second, I would want to use the second channel of the FG to generate a marker pulse, modified half cycle, etc, to confirm who is leading who...

I believe using a marker was touched on during discussion of your flywheel version of this setup, which was never instrumented as suggested to observe acceleration and deceleration of the flywheel.  Although possible, it would be a bit more difficult to do so with this new "waving wand" version.

PW   

I will be winding a new 1:1 coil tonight,as requested by Poynt,and will be able to supply the dot convention.
I cannot determine the winding direction of the primary winding of the transformer,as it is taped and lacquered,and the wire very fine. Guess i could use a compass and DC current to work it out.

Also remember the secondary winding is on top of the primary winding--not along side it as with a normal E core type transformer. So in this case,at the 8.46 minute mark,the yellow trace is actually leading the blue trace,as it was lagging at the start without the magnet in play,and thus show's the 90* phase shift as expected.


Brad
Logged
Pages: 1 ... 133 134 135 136 137 [138] 139 140 141 142 143 ... 182   Go Up
« previous next »