Joule Ringer!

[PhysicsProf]

@PhysicsProf
PB has a new youtube channel Bill Mckraken - YouTube
We see on his notes that he has 104lm/W, but below we can read 124lm/W too
Lidmotor says here Simple --Super Joule Ringer 2.0---Boxed - YouTube that PB is modifying the caps on the light bulb itself. That's very interesting, and I suppose that's how he gets 124lm/W, but this video is private i suppose, and he sent the link only to friends for preview or something like that ... his right anyway
I think taking out the resistors in series with the LEDs (inside bub) improves brightness too, as SkyWatcher showed with his camping lamps ...

I noticed already your attention disorder dear Prof, please read care-fu-lly
love and light,
Jules

PS: Totoalas has the ultimate reason for all this research = bringing cheapest light to those in need of it We can't expect big companies or patent oriented researchers to do that

PB was kind to already answer my question, Jules -- and he said:

@ PhysicsProf and all to clarify here.
-My highest with a Lasersaber Super Joule ringer circuit; is the last shown (previous page). This is Highly viable and cheap.
*Using PhysicsProf preferred method (same calibration):
Shown setup was 76 Lu/W max @ 455 Lumens. Error of 5% though, as I don't have an 8% number to draw from.

*Using My preferred method (each bulb calibrated as well as number) We get a more realistic 59 Lu/W @ 354 Lumens. This is within 1/2-1% accurate.

If you wish to disagree with what PB himself has written in response to my question, you will need to explain your disagreement with PB.

PS -- I don't think I have an attention disorder; it was not kind of you to make a condescending and unsubstantiated remark about "your attention disorder dear Prof, please read care-fu-lly".

[Allen Burgess]

JR 2.0 & Hertz

I finished boxing my Jr 2.0 with the addition of a 12 volt bank of AA's, an on off switch and an 8 watt 120 volt LED bulb like Lidmotors.

I lit the 8 watt 120 volt LED bulb with my 12 to 120 volt 1500 hertz 3 amp, glopanel "E-Driver" inverter. The bulb drew 1.4 amps. The same bulb drew 100 milliamps from the 350 Kilohertz JR 2.0. 1/14 the draw at 233 times the frequency.

The bulb looks to be equal in brightness, I'll have to wait for my lux meter to be sure. It looks like for now that multipling the hertz by around 20 cuts the amp draw in half each time!

Someone with a signal generator ought to be able to confirm wether or not there's any real correlation between hertz and power pretty easily. This relationship appears occluded by routine oversight.

[Jules Tresor]

@ PhykicsProf : oops sorry prof, I was just kidding about the attention disorder, it's a forum, a friendly one, I thought we were happy guys. Sorry if you felt a different perception, I apologize !
Cheers, JT

PS: don't take life too seriously, it's just an illusion !
I mean, as you are Physics Professor you must have realized that it's all empty. Atoms are 99.999999999999 % empty, 100% in mathematical statistic.
Just a projection, nothing real, just animated emptiness.
And this comes from Physics School Books, not from me .... But it's another story, of course

[Lidmotor]

Led circuits

[QUOTE=b_rads;196958]I have several of the FEIT LED bulbs. Last night I purchased my first dimmable Utilitech 7.5 watt bulb. Noticed in very fine print that the bulb is distributed in the U.S. by FEIT Electric Corporation out of California just like the FEIT bulbs. Both are made in China. Is it possible that they use the same internal circuitry?

Brad

I don't know what circuit is in any other bulbs. The one Utilitech that PB took apart and showed is the only one that I have seen analysed . I have not taken any of my led bulbs apart yet to see what is inside.

@ Dr. Jones and PB

I have been doing some basic lumen / watt testing just to see how my Ringer build stacks up to other lights. The most surprising find was a 5 watt 12volt DC led bulb I bought on ebay that appears to pump out about 70 lumens / watt. That is with the bulb hard wired straight to a 12 volt battery. I used the 7.5 watt AC Utilitech as a gauge for the 450 lumens and compared other bulb's lux meter readings vs amp draw. I am wondering now if maybe we should divert from the AC bulb idea and try kilohertz "blinking" a DC bulb. It would call for a whole new circuit probably.

Cheers,

Lidmotor


PS---I ran a test last night with my boxed unit using plain old AA zinc carbon batteries. You can get 8 for $1 at the dollar stores. It ran the 7.5 w Utilitech for 4 hour and then the light turned off. The battery pack voltage had dropped from 12v to about 6v. I thought that it was done but for fun I installed the 2.5w Lights of America bulb and the device ran that little light ALL NIGHT and it was stlll on this morning with battery pack sitting at 6 volts. That is what I call a great bang for the buck. At this low voltage that little light is on part way but the amp draw is only about 15 mA. I'm running it again tonight to see if it will go until the morning. It makes a nice night light. All that fun for $1.

[PhysicsProf]

@ Dr. Jones and PB

I have been doing some basic lumen / watt testing just to see how my Ringer build stacks up to other lights. The most surprising find was a 5 watt 12volt DC led bulb I bought on ebay that appears to pump out about 70 lumens / watt. That is with the bulb hard wired straight to a 12 volt battery. I used the 7.5 watt AC Utilitech as a gauge for the 450 lumens and compared other bulb's lux meter readings vs amp draw. I am wondering now if maybe we should divert from the AC bulb idea and try kilohertz "blinking" a DC bulb. It would call for a whole new circuit probably.

Cheers,

Lidmotor

Good ideas, Lidmotor! I have purchased and tested several 12V DC LED bulbs -- and I agree, some are very efficient in the 60-70 Lumens/Watt range, running on 12 V. The bulb below runs at 0.2W at 12V, and puts out close to 70 Lm/W.

One has to be careful in testing bulbs with beam "directional" light output, since if that beam is pointed towards the lux-meter, one can get incorrect (high) results. But I took care with these bulbs to avoid the beam-pointing problem; and you're right, Lidmotor, the efficiency is quite good on some of these.
Running with a joule-ringer and a 12V bulb is a good idea... will try.

[Harvester]

Hey all,

I have replicated the above circuit, but have some issues with it.
Coil has 6 strands of 22AWG, 60m long.
Not using the input and output caps and only having a 1uF cap instead of the 10uF one.
Charge battery is an old 12V lead acid, currently at 8.4V
Input is from 12V power supply, max 2A.

Problem I got is that either the circuit draws next to no current, or so much that it fries my variable resistor.
The 5KOhm variable resistor does nothing and has to be on 0 Ohm.
With the 1 KOhm resistor in the higher range, the circuit uses 120mA and produces 12.3V at the charging battery and 94V when charging battery is disconnected.
When tuning the resistor in the lower range, the amp draw all of a sudden jumps to 730ma, charging battery voltage goes up to 17V and with disconnected battery, the voltage jumps to 210V.
Problem is that at this stage the 1KOhm resistor starts smoking.
Had a 680Ohm 1W resistor instead of 470+100+100, but started smoking so changed to the ones shown in circuit at 1W.
Have changed all resistors to 1W and all diodes to ultra fast ones, just in case.
Is there a bug in my replication or is the variable resistor just too weak in wattage?
The next jump of input amps is to 1.7A, 25V at charging battery and nearly 700V disconnected.

The changes are right at the beginning, when there is hardly any resistance and they are very close together, as in by turning the resistor maybe 10 deg.
The other coil I used out of CAT5e network cable with shielding removed, was easier in tuning, as the input current would rise in proportion to turning the resistor.
Now it is like using a switch - 0.12A, 0.73A, 1.7A
The coil frequencies also only jump when the amp draw changes - it is not a curve, but steps.

Has anyone got an idea what the problem might be?

[PhysicsProf]

me: "Running with a joule-ringer and a 12V bulb is a good idea... will try."

I used a 12V- DC bulb that drew 0.3 W, about 64 Lm/W in my light box.

Connected to the SJR 2.0, I decided to start at 8 V DC input to see what would happen. It drew 0.76A @ 8V = 9W but only put out 24 lumens, so roughly 3 Lm/W. Also, the bulb would no longer function at 12V DC input. (Don't worry -- I purchased several of these before; not expensive.)

So -- at least for this 12V-DC LED bulb -- running on the SJR 2.0 did not increase the output yield per watt... decreased it significantly.

Worth a try!

[SkyWatcher]

Hi folks, Hi harvester, I have been using that circuit to charge batteries very well, though I do not use the 20 kohm resistor from base to collector and also do not use the 10 kohm resistor from base to ground.
Also, the 1n4148 diode from base to ground for additional transistors is an option.
Also, I use a non-polarized 100uf-100volt capacitor in parallel with base resistors and works well.
I have a 24 gauge, 5 strand air core coil that works very well also, though a little thicker gauge would be needed for higher current draw, heat buildup.
I would say with those mentioned parts removed, it is the best self oscillating charger I have used.
Hope that helps.
peace love light
tyson

[Lidmotor]

New circuit needed

me: "Running with a joule-ringer and a 12V bulb is a good idea... will try."

I used a 12V- DC bulb that drew 0.3 W, about 64 Lm/W in my light box.

Connected to the SJR 2.0, I decided to start at 8 V DC input to see what would happen. It drew 0.76A @ 8V = 9W but only put out 24 lumens, so roughly 3 Lm/W. Also, the bulb would no longer function at 12V DC input. (Don't worry -- I purchased several of these before; not expensive.)

So -- at least for this 12V-DC LED bulb -- running on the SJR 2.0 did not increase the output yield per watt... decreased it significantly.

Worth a try!

Dr. Jones,
I tried that also and it didn't work. Luckily I didn't destroy the bulb. There is probably a circuit already out there that will do what I am thinking about. A CMOS 555 timer circuit would do it. The trick (as always) is not losing energy in the process and ending up back at square one. A long time ago we worked on a CFL Joule Thief project here where we blinked the whole JT circuit with a seperate timer circuit and it worked. That project end up in a boxed version that I called the "Halo light" (because it used a round bulb) and that is the light I have used on my boat for several years now. Now that we have better led bulbs to work with I might try that whole circuit again.

Lidmotor



PS---The cheap dollar store AAs powered my new "Ringer" light again through the night and it was still on this morning. The loaded voltage was around .6 volts. The question now is ----when will those cheap batteries start leaking. Gotta watch out for that.

[Peanutbutter29]

@ Lidmotor. Ya, I mentioned in one of the first posts, I believe frequency starts to help above 250Khz or so (well can). You can also use PWM to lower duty and have frequency as well. This was what I looked into for Aquarium lighting when they first came out; still need 100w of LED on a tank, so not economical yet. The problems with HF can be many, but transformer loss and transmission may be most common; these "could" offset advantage from Frequency.

Seems like a 12v PWM motor control, tweaked for higher frequency; may be a good start here.


@ PhysicsProf and about my light post......did you feel I posted these incorrectly? Here's some prior to catch up:

You first noted when testing these (168-252-80 Lu/W) Chinese bulbs. Not sure which is the Lu/W you consider as all were used; I note too.....these are 220v bulbs...did you calibrate on 220 then? If not that could easily show incorrect Lu/W
Anyway you first stated:

....so far, I've found that the Warm light and the "pure white" bulbs give very close to the same lux values with my little meter from Amazon as they are rated, within 3-5%. If we can get +/- 8% values on these tests, I'll be happy
--Steven J

Later, in trying to convey the error and that it could be eliminated I stated.

Adjusting for each bulb, CAN cut down to 1% error rather than 5% you state and is shown above. This is why I showed my "proven" variation in bulbs can change 10-15 Lu/W of readings at the end.

Well your bulb? It is a daylight.....sooo, it's got a lower ratio than your INC and Compact conversion. By applying the Same you have introduced an error of 5%+ by NOT re-calibrating for that bulb as per mfg.

You then replied with;

Originally posted by PhysicsProf (post 1313)
Now, about the lumens -- I have repeatedly said that my light box calculations have an error of about 7 or 8%, so I have NO PROBLEM with your statement that I have introduced an error of 5%+:
QUOTE:
Well your bulb? It is a daylight.....sooo, it's got a lower ratio than your INC and Compact conversion. By applying the Same you have introduced an error of 5%+ by NOT re-calibrating for that bulb as per mfg.

The error is at least 5% as I have said, probably more like 8% +/- . I still think that this measurement is very helpful.

Okay, here, in all honesty; my mind cannot comprehend how a man of "science" says an 8% error is fine. What confounds me more, is that you want to compare your Error with others Accurate results. This is NOT NOT NOT scientific in any way!! The intention of making a "scientific instrument", is to make it as accurate as possible. I'm still utterly mind boggled here as if CERN applied this; we already would have the "god particle"! Heh, well within an 8% error. If anyone in science has an option of 1% error or 8%; The 8% get's tossed and the 1% is used.

Moving on.....

@ PhysicsProf and all to clarify here.
-My highest with a Lasersaber Super Joule ringer circuit; is the last shown (previous page). This is Highly viable and cheap.
*Using PhysicsProf preferred method (same calibration):
Shown setup was 76 Lu/W max @ 455 Lumens. Error of 5% though, as I don't have an 8% number to draw from.

*Using My preferred method (each bulb calibrated as well as number) We get a more realistic 59 Lu/W @ 354 Lumens. This is within 1/2-1% accurate. So, efficiency of whole circuit is ABOVE any inverter.

**If we applied PhysicsProf preffered method to the High power mode; we would assume 583 Lumens (100 over rated) and 62.3 Lu/W. Since factory is 65 Lu/W here, this would be well into overunity and NOT the case.

My post above, PhysicsProf below:

PB was kind to already answer my question, Jules -- and he said: QUOTE:
@ PhysicsProf and all to clarify here.
-My highest with a Lasersaber Super Joule ringer circuit; is the last shown (previous page). This is Highly viable and cheap.
*Using PhysicsProf preferred method (same calibration):
Shown setup was 76 Lu/W max @ 455 Lumens. Error of 5% though, as I don't have an 8% number to draw from.

*Using My preferred method (each bulb calibrated as well as number) We get a more realistic 59 Lu/W @ 354 Lumens. This is within 1/2-1% accurate.

If you wish to disagree with what PB himself has written in response to my question, you will need to explain your disagreement with PB.

I've pointed out above, as I assume you still want to compare my 1% number with your 5%...... Again, I am only posting both, since your "fine" with 5%. So, if you, in your house tested that EXACT same system; you would have reported 76 Lu/W.... I don't understand if you have an issue here.

A note too. I got a nice laugh out of the psychology behind changing bold. I'm sure Freud would've enjoyed it too! At least your consistent.

Next,

One has to be careful in testing bulbs with beam "directional" light output, since if that beam is pointed towards the lux-meter, one can get incorrect (high) results. But I took care with these bulbs to avoid the beam-pointing problem;

Okay, uhm did you make a new box? Mine is 12" ABOVE and perpendicular to light....... I would like to draw attention to Exhibit A: Ecore10V45LpW.MPG - YouTube
(watch from 2:30 about)
Here we see your box and meter with your (80-168-252 Lu/W) bulb.
-The sensor is mounted low in the box AT bulb level
-The sensor is mounted in "partial" parallel to light
-The sensor is DIRECTLY pointing at a minimum of 9 LEDS
-We see a 1/2 Cu. Ft. box with little to no reflection!!

Uhhh, how can your statement above about design be true?

**Also, please professor you try this and anyone else with these LED's. If you get to 150 Lumens.....the camera will not show exact LED outline as in your video. Also, if you compare a "line run" 150 Lumen bulb to those dim LEDs; it is OBVIOUSLY no where near 150 Lumens. You are in fact closer to 100 there. Again I am the one with the "scientific box", to be blunt.

Finally then, have you gotten a normal transformer to run in SJR design yet? I know the first failed, but since most seem to be using that; I wondered if you've tried again?

With the DOME light, I'm not sure that an SJR is best for 12v bulbs? Maybe with a 1-2v input....might be interesting.


Last then from OU forum, since it was asked

Quote from Lynxsteam OU post #85
Here is my latest test with the Utilitech LED warm 7.5 watt bulbs. The watt reading off the house grid was done through a watt meter and it is calibrated. Please try this yourself and see that these bulbs consume more power than stated on the package, note that my grid voltage is 120 v and that may cause a higher amp draw.

First, I'm pretty sure normal line is 120v as that's what the bulbs are rated for; so I'm not sure "higher amp draw" will apply. I have 13 of these Utilitech bulbs that range from a couple weeks old to 8 months. I double checked them all to be sure with my Kill-A-Watt (tm) meter. They all show rated 90ma of draw and 7.5w.
1 Bulb- 7w
2 Bulb -15w
3-Bulb -22w
4-Bulb -30w
5-Bulb- 37w
6-Bulb- 45w
7-Bulb- 52w
8-Bulb- 60w
9-Bulb- 67w
10-Bulb-75w
11-Bulb-82w
12-Bulb-90w
13-Bub-98W

Note, 2 are WW and 11 are daylight. So, I'm not sure how you get 16w for 1 and 35w for 2???? Maybe the meter is bad? Possibly the bulbs are bad, but from knowing the circuit; that SSL chip would have to be fried at the very least for this to happen. Testing 13 and all being the same would not suggest a large variance in rated.


Finally @ ALL. I personally am not trying to attack ever, BUT if you (I / we) are representing Science; we are bound to respect THAT above our person. If we are trying to make scientific advancement, we ALL must adhere to common practice as much as possible. This should not be about numbers as much as viability and what good it can do. If we do not go about this accurately and correctly; then we may not accomplish anything. The only thing in all this that gets damaged with errors, is science and "opportunity" for less fortunate.

E.G-
"If I'm going to tell a guy to trade this set of batteries for his family noodle pack to light his house; I'm going to be DAMN sure it does!!!!" I feel VERY strongly about this

Sorry to rant again!
Thanks
PB

PS @ lidmotor, just saw your last post.
Yes 12v prolly have a CC circuit, see above for 1-2v option?
Nice run times with your light box too!!
Also, batteries can leak if they are pulled down too low; maybe try a Zener across battery for Low Voltage shut off or something?

[Allen Burgess]

Hertz

Peanutbutter29,

Running the Hertz up from 250k on a bulb linked to a lux meter, showing increased brightness with no increase in input power would make a landmark video!

[totoalas]

Hey all,

I have replicated the above circuit, but have some issues with it.
Coil has 6 strands of 22AWG, 60m long.
Not using the input and output caps and only having a 1uF cap instead of the 10uF one.
Charge battery is an old 12V lead acid, currently at 8.4V
Input is from 12V power supply, max 2A.

Problem I got is that either the circuit draws next to no current, or so much that it fries my variable resistor.
The 5KOhm variable resistor does nothing and has to be on 0 Ohm.
With the 1 KOhm resistor in the higher range, the circuit uses 120mA and produces 12.3V at the charging battery and 94V when charging battery is disconnected.
When tuning the resistor in the lower range, the amp draw all of a sudden jumps to 730ma, charging battery voltage goes up to 17V and with disconnected battery, the voltage jumps to 210V.
Problem is that at this stage the 1KOhm resistor starts smoking.
Had a 680Ohm 1W resistor instead of 470+100+100, but started smoking so changed to the ones shown in circuit at 1W.
Have changed all resistors to 1W and all diodes to ultra fast ones, just in case.
Is there a bug in my replication or is the variable resistor just too weak in wattage?
The next jump of input amps is to 1.7A, 25V at charging battery and nearly 700V disconnected.

The changes are right at the beginning, when there is hardly any resistance and they are very close together, as in by turning the resistor maybe 10 deg.
The other coil I used out of CAT5e network cable with shielding removed, was easier in tuning, as the input current would rise in proportion to turning the resistor.
Now it is like using a switch - 0.12A, 0.73A, 1.7A
The coil frequencies also only jump when the amp draw changes - it is not a curve, but steps.

Has anyone got an idea what the problem might be?

HI
using same circuit and sjr load the lamps are brighter @150 mA as compared w/ sjr only with less brightness@ 250 ma freqncy and caps as they say makes it brighter

totoalas

PS added to a noodle pack are the children studying at night

[Harvester]

Thanks all for your comments!

I have been using that circuit to charge batteries very well, though I do not use the 20 kohm resistor from base to collector and also do not use the 10 kohm resistor from base to ground.

Have tried that as well, removing the B-C resistor and changing the E-B resistor with a diode, like the other ones.
Same behaviour.

Also, the 1n4148 diode from base to ground for additional transistors is an option.

So you are leaving those out..... that is something I have not tried yet.

Also, I use a non-polarized 100uf-100volt capacitor in parallel with base resistors and works well.

Are you talking about exchanging the 10uF one, or each 100 Ohm resistor to each base?

The largest non-polarized cap I have is the 1 uF one.
Have tried a trick I read somewhere, where I took 2 polarized 80 uF 330V capacitors,
soldered the + together and attached the pack to the circuit.
The frequency was sooooo low, that when going over the air coil with a magnet,
I could feel the frequency that the magnet was drawn towards the coil.
It feels like 4Hz or so and the pulls are quiet strong.
Using the 1uF capacitor, no pull is noticeable.
Maybe I should play around with smaller values?

I would say with those mentioned parts removed, it is the best self oscillating charger I have used.

Will give that a go, thanks.
Getting the close to 700V is scary as well, I recon...

the "trick" w/ the cap is really just using it to tune for freq, pulse width, and voltage to the base. So experiment with several size caps so you can get the feel for it. the small cap you have + the impedance of your trigger coil could be the cause for the small window you are seeing while tuning.

Will give that a go, thanks.
The CAT5e coil had the same dimension, but only weight about 300g, whereas the one I am using now is easily >1Kg

Furthermore pay real close attention to how PB is tuning his ckts.
PB's video "Explainaintion for operation of Super Joule Ringer, Bias resistor and SJRC" was the best light shedding explanation I've ever seen/heard.

Unfortunately the video does not exist any more - account closed.
@PB
Could you please send a link to the video?
If sent via PM, then I will not pass on.

using same circuit and sjr load the lamps are brighter @150 mA as compared w/ sjr only with less brightness@ 250 ma freqncy and caps as they say makes it brighter

Interesting.....

Thanks again to all.

[Allen Burgess]

120 volt LED bulb circuit

I disassembled a damaged 120 volt 2 watt "Lights of America" LED bulb. There is indeed a circuit inside the base. I spotted what looks to be a rectifier, and several tiny transformers in metal caseings.

I tested my boxed Jr 2.0 on the sailboat too. The 8 watt bulb lasted one hour running on a 9 volt AA bunch. I'm conducting an endurance test now on a 2 watt bulb on 12 volt AA.

[SkyWatcher]

Hi folks, Hi harvester, yes, I am not using the 1n4148 diodes and I am just exchanging the 10uf capacitor for a 100uf.
I am using 6 resistors in parallel with that 100uf capacitor, 3-1kohm resistors in parallel and in series with another 3-1kohm in parallel, think they are 1/2 or 1 watt, so about 666 ohms total going into the main base diode, then I have 300 ohm resistors going to each transistor base.
Though all of those parts will be adjusted to your setup.
peace love light
tyson