Joule Ringer!

[blackchisel97]

I'll try couple different transformers from various 110/whatever adapters tonight.
I found 100nF cap being "just right" with my setup. It is small, center tap trafo, maybe 5W at most. I'm using MJE3055. I had 21194 hooked up before but it was overkill for 2W LED.
I also incorporated two Lidmotor circuits into one sheet and redrew for clarity -
http://i29.photobucket.com/albums/c2...e2Lidmotor.jpg

Thanks
Vtech

[Peanutbutter29]

Question for PhysicsProf --- Light box

I had a quick question or two if okay. I got my light meter today.
And have finished construction of the box AND finished calibration testing; however this is one question.

First, Meter is 100,000 Lux, 4% accurate, 2% repeatability, model LX-1010BS. I noticed in the manual that there's a spectrum "area" for this meter to work. It does not include into the blues' much and seems, daylights need a separate Factor from warm white.

I'll do a quick video or pics to show, but quickies and questions.
Calibration:
Inc:
60w warm white 850 lumens, 1-6 bulbs (13 point check)
100w warm white 1600 lumens, 1-4 bulbs (7 point check)

Compact:
13w Warm white 900 lumens 1-5 bulbs (8 point check)
13w Warm white 835 lumens 1-4 bulbs (5 point check)

LED
7.5w Daylight dimmable 490 lumens 1-6 bulbs (5 point check)

So, first, my line was a bit low, so power averaged 93-98% between bulbs, only LED showed full power. I adjusted each test for each watt % lower for accurate lumens of bulb; then figured Calibration ratio.

I must say I'm very impressed with the setup, design and everything and it is more accurate than I'd hoped.

Just with Inc and Compact (warm white) 29 points varied 5% max in ratio. Toss out one "odd" high Inc bulb and lowest "placement figures"; A 22 point final reference ends up with 3.7% variation across all points!!! Very accurate indeed.
The mean average then for conversion ratio of Warm whites, ends up at .201 for my box. I did notice, since larger; you get MORE accurate numbers with more bulbs, or having 1-2 closer to meter (set so cannot be direct); so that will work well for me.

So, Dead on +/- 3.7% Inc and Compacts are .201

But, enter LED daylight. they average .156 ratio with a variation of ONLY 2.5% across the 5 points!? This is less variation (including placement) of ALL other bulbs.

But the average for LED then is a full 4.5% off my Inc. and compact calcs. (.201 avg warm white, .156 avg daylight)

I did design the box well to not get direct light and I'm happy with it. Also having most accurate with top bulbs is nice. And I cannot complain about a 3% accuracy.

**However, I cannot in good conscious use the same factors from warm white, OR even include them as they are so far off from other 30 points. E.G, with a 3% accuracy, there MUST be a different ratio for these LED's in my box.

So after all that, what do you suggest I do? I'm going back in (20 bulbs later, ugh) to get 6 daylight compacts to see here. Otherwise, I feel .201 and .156 are BOTH right.

Just want to make sure calibrations are right before trying a test run.

Thanks PB

[PhysicsProf]

I had a quick question or two if okay. I got my light meter today.
And have finished construction of the box AND finished calibration testing; however this is one question.

First, Meter is 100,000 Lux, 4% accurate, 2% repeatability, model LX-1010BS. I noticed in the manual that there's a spectrum "area" for this meter to work. It does not include into the blues' much and seems, daylights need a separate Factor from warm white.

I'll do a quick video or pics to show, but quickies and questions.
Calibration:
Inc:
60w warm white 850 lumens, 1-6 bulbs (13 point check)
100w warm white 1600 lumens, 1-4 bulbs (7 point check)

Compact:
13w Warm white 900 lumens 1-5 bulbs (8 point check)
13w Warm white 835 lumens 1-4 bulbs (5 point check)

LED
7.5w Daylight dimmable 490 lumens 1-6 bulbs (5 point check)

So, first, my line was a bit low, so power averaged 93-98% between bulbs, only LED showed full power. I adjusted each test for each watt % lower for accurate lumens of bulb; then figured Calibration ratio.

I must say I'm very impressed with the setup, design and everything and it is more accurate than I'd hoped.

Great! Very happy you are jumping into these measurements.

Just with Inc and Compact (warm white) 29 points varied 5% max in ratio. Toss out one "odd" high Inc bulb and lowest "placement figures"; A 22 point final reference ends up with 3.7% variation across all points!!! Very accurate indeed.
The mean average then for conversion ratio of Warm whites, ends up at .201 for my box. I did notice, since larger; you get MORE accurate numbers with more bulbs, or having 1-2 closer to meter (set so cannot be direct); so that will work well for me.

So, Dead on +/- 3.7% Inc and Compacts are .201

But, enter LED daylight. they average .156 ratio with a variation of ONLY 2.5% across the 5 points!? This is less variation (including placement) of ALL other bulbs.

But the average for LED then is a full 4.5% off my Inc. and compact calcs. (.201 avg warm white, .156 avg daylight)

I did design the box well to not get direct light and I'm happy with it. Also having most accurate with top bulbs is nice. And I cannot complain about a 3% accuracy.

**However, I cannot in good conscious use the same factors from warm white, OR even include them as they are so far off from other 30 points. E.G, with a 3% accuracy, there MUST be a different ratio for these LED's in my box.

So after all that, what do you suggest I do? I'm going back in (20 bulbs later, ugh) to get 6 daylight compacts to see here. Otherwise, I feel .201 and .156 are BOTH right.

Just want to make sure calibrations are right before trying a test run.

Thanks PB

OK -- as I mentioned before, I adjusted the INPUT WATTS when running bulbs to the WATTS specified for each bulb on its packaging, that is, by adjusting the input voltage (one can use a Variac or dimmer-if Vin is too high) so that the input WATTS on the watt-meter reads just what is specified on the bulb's package. We're making use of the calibration that the manufacturer did, you see.

Do you have a watt-meter?

Using a watt-meter and making the small correction (if you haven't done so already) may help. The meter also may be off a little; but important I think to do the small corrections that we can find.

Beyond that, I would agree that calibrating with LED BULBS then using this LED-bulb calibration factor -- is the right approach, and then run the blocking oscillator with the SAME LED bulbs, and make your comparisons in Lumens/Watt. I would say go with this, and start making your measurements.

This should be quite interesting to see. If your light output is more than say 30% higher than the controls (controls = same LED bulbs running on the mains, in Lm/W), then you clearly have something important. One can always go back and get an even better measuring device... Basically what we're doing is a "light calorimeter", and its a good metric IMO for us to use.

Right now, I'd say we're trying to find out roughly (within about 8%) the measure of Lumens/Watt with these devices, AND then try to maximize Lm/W. I achieved 53 Lm/W as I mentioned above, with an air-core xformer.

Sounds like you are ready to make such measurements. I look forward to hearing how it turns out!!!

PS -- I suspect my older 2n3055 has been giving me some problems. I already burned one out completely. So I'm looking for a few more of these... holiday weekend might slow me down a bit...

[Peanutbutter29]

I'll go in order I suppose. I don't know how I missed Limotor's comments, heh. Was focused I suppose.

Ya that Cree I saw a while ago, they announced another already that's better than that one; but it's a couple years before that 150 hits the market sadly.

@ that guy, with the scopes. I did comment in the short YT space; he was really pokin' huh? You now now that you say "worlds smallest lab" I think I may have watched some of his stuff. I agree good guy.

I DID say this is barely tunable. Again I've tuned all of these without scopes. I don't know why he said no one shows scope's as I have a video up. Also, since I got the 1.0 out again, I checked the scope.
YEP, same wave (close) to the 1.1. Those two I said were definitely tunable. I also did some looking at the stock Radio tranny and I think I can help "Show" more there. I'll do a vid here sometime with it, soon.

For tuning have I not talked "adamantly" that "BOTH" 2 caps are required to "Tune". I cannot understand, as with this scope guy "I don't see any change". Well NO, TURN ON THE CIRCUIT then maybe you will. Notice he tried to show a funky form "he" approached from L2 only, lol. If you can't see it, then it ain't real. When he turned the circuit up to full power and the BLOCKING started properly. He would've had the normal square, but he didn't show that?

Last here, if someone takes a simple bit of time to look at real power and light (even with solar tube) you will see what's more efficient.

There IS however a KEY with LED's themselves and noting it I found out why. If you look at specs for high power LED and "Overlay"; "Forward current vs. voltage" with "Forward current vs luminous flux" you will see they are inverse in arc.
So, from one Perspective there literally IS a slight higher efficiency near start-up.

Sorry to blather .

@ PhysicsProf, uh lets see if I get it all. first, Glad you said run with LED numbers as that seemed more logical. I DID get more bulbs and they are interesting; so I'll add those.
-First, Ya I use a Kill-a-watt, heh...first model made, so no decimal.
-Second I fully expected them to "overrate" wattage; particularly inc.'s, but none did. My line is only like 117, think the pole pig is old. But, ya. All numbers were LOWER than rated, so I had to de-rate. Now if the variation was less than 1% I wouldn't but overall was 4% variation from full at times.
I figured when Conversion ratio was figured; I would know if this de-rating was correct. Like I said, 3% variation; I believe it was the right method.

So, extra bulbs tested. Makes an interesting note that daylights may in fact be different in rating; as with LED's......compared to Compacts or Incs.

So I got 2 Utilitech LED warm white dimmable 7.5w.....to see if LED were lower, or brand, or color.
----I tested 1-2 (6 points) Average here .184 +/- 1%

I also bought 13w Daylight (5000K) compacts to for verification about daylight specifically.
-- I tested 1-4 (7 points) Average here. .177 +/- 3%

Seems, both were somewhat verified, daylights are different with my meter at least, AND LED's seem too. A note my LOWEST Compact or Inc. number of 29 points was .194, though most were +.2xx

Last, I re-checked LED's I use and got same numbers +//-1%, so I'm going to use .156 for these +/- X% of course. Thanks for the confirmation there.

I'm anxious to get final real readings on the V1.0 as it was by far the best. 1.1 was good, but after I was only looking to get something purchasable easily.

Was up all last night working on another circuit, lol. Offers interesting things possibly and may be a solution for 220v. As I got 2 LED's to light is series....yup both lit. Though not bright as they "bucked" each other before that point. E.G. though a 220 LED there would run fine. I'm analyzing efficiency before putting up though.

Try to get something up showing the light box too! I assume LED's will just live in there now, lol. I'll just have a box and circuit, hehe.

Thanks PB

[PhysicsProf]

PB --

I'm anxious to get final real readings on the V1.0 as it was by far the best. 1.1 was good, but after I was only looking to get something purchasable easily....
[snip]
Try to get something up showing the light box too! I assume LED's will just live in there now, lol. I'll just have a box and circuit, hehe.

Thanks PB

You got my attention, PB! I look forward to your readings (and others who have taken the step to do a light box!) My LED's pretty much live in the light box now, though I will put one in and take one out occasionally. Make sure they're screwed in tight when making a change...

Note that the power strips sometimes have a light on the switch that puts out several lux typically -- I'm using in the box only strips WITHOUT the light on the switch to avoid that (small) systematic error.

PB -- sorry to be lazy, but could you refresh my memory on your v. 1.0?

Did you wind on the spool (and E-core) recommended by Lasersaber?
What size wire did you use for the primary -- and how many turns?
What size wire did you use for the secondary -- and how many turns?
What "hints" do you have for the tuning?

Thanks so much, great progress happening! exciting times.

[PhysicsProf]

PB -- I just saw your latest vid with your new "Cadillac-model" light box -- very well done! Lots of good ideas, as usual from you, such as mounting the sensor the way you did and using ties to affix the power strips down in the box.

PhysicsProf Light Calibration Box finally made! - YouTube

Looking forward to your results in the box, especially version 1.0 !
--Steve

[Peanutbutter29]

SJRC 1.0 Accurate readings / Sincere apology

I finally got a video uploading with accurate readings from the 1.0. I am very disappointed and totally surprised at the error margin.
I want to state here as on the video I sincerely apologize for false numbers as I should not have such confidence in the old meter and method both. I do not ever intend to mislead or mis-state and I have done that here in a big way. I have removed light % estimates from titles of the 2 videos that had it too.

I don't want anyone else to get discouraged or stop looking into these as a result of my mistake. I think there IS merit here and; I suppose all these currently may well suit a persons needs. Lidmotor said this best " It works for me"

I think also, this re-enforced PhysicsProf light box as a required item to get any grasp on real light output. Light to the eye is just too deceiving and other methods not accurate.
I'd asked everyone that I have seen since then at my place (3 lol) if it looked more than a bulb on A/C and it does. But this is just not the case.

I encourage to watch the video for all details. I also, used a new calibration (placed bulbs as in video). I state, but can show if needed; it is within 10 lumens final result. So, again I'm not hiding or cheating.

From what I show, we end up at:
12v @ 1.31A -2230 lux x .221 (see video) = 493 Lu or 31.36Lu/W

I show at the end completely un-tuned and get
12v @1.09A - 1530 Lux x .221 = 338 Lu or 25.8 Lu/w
Note:I had to lower resistor a bit, but another test was similar
12v @1.20A - 1570 Lux x .221 = 347 Lu or 24 Lu/w
(this is where I'd expected to be on video, but oh well)

So, tuning here added 5% efficiency, 155 Lumens and the charging (440 mw). I show tuning aspects in the video and to compare "if it helps". Also, I show scope shots, tuned and not running.

***I wonder if a bunch of those 1w LED you tried PhysicsProf, would double my numbers? then I'd be around 60 lu/W

It is interesting that my stated lumens were close originally, but my error was multiplying by bulbs. So I would (for now) de-rate v1.1 to whatever lumens I came up with (for the one bulb, don't multiply) by the watts in. i'll re-test them all, as what I thought most efficient "may" not be the case.

I tested a bunch and learned a few things I didn't show on video.
It seems,
1 bulb is more efficient than multiples in terms of Lu/W, BUT you still get more total lumens with more bulbs. Also, ability to have multiple placed. With one bulb I show same efficiency tuned or not (again I tuned for 7) of 30 lu/W. However I still get 5ma charge tuned with 1 bulb. So, from un-tuned you loose 4.2 % efficiency with multiple bulbs. I guess the tuning just more than "countered" this, ultimately and added the charge.

Also, I tested High resistance Vs. Low, tuned or not. In all cases (with this trans) efficiency dropped! Tuned and Highest resistance we get to around 24 Lu/w and 5ma charge. Un-tuned we get 20 Lu/w

Now, at Highest resistance Un-tuned and 1 bulb. I could NOT get much lower on variable with out the transistor being "over-biased" . However, it ran 12v 60ma, yup less than one watt (starting running and lowering). The bulb was lit too....efficiency. 15Lu/W.

Also, I didn't get to show, but in high resistance mode, I get a more "normal" square wave like you see with these. So, I believe what I considered "L2 dominate" means (from scoping) maybe that the square wave cannot be changed?

Last here, about the box. I even tried having one bulb IN the box and all others out to see if something was up there. There IS a discrepancy here. With 1 in and 3 out. the total (if multiplied by 4) "should have been 27 lux higher. Also with 1 IN and 6 out, it was a 52 lux discrepancy. (one bulb, verticle; so using the .156 ratio here) That works out to just over 8 lumens; so it is "possible" 8 lumens are being lost that don't hit the meter here. Won't change efficiencies much but a change nonetheless.

So, finally as I proposed on YT. IT's up to you all to decide whether I should "bow out" here, as the mistake was large. I fully understand anything bad someone might have to say and welcome what is due there.

Again I am deeply sorry for mis-statements of light.
Sincerely,
PB

I'll post the video link when finished

[PhysicsProf]

Actually, my dear colleague Peanutbutter, this is REAL progress and I congratulate your honesty in sharing your results even when disappointing.

It is SO easy for us to think we are getting one result JUST USING EYEBALLS and that's why it is so important to use a calibrated "IMPARTIAL" device.
You did well -- we're NOT accusing you of deception before, we realized a quantitative measure would be needed -- and now you HAVE DONE IT and honestly presented us with the NUMBERS.

BTW, your numbers of around 30 Lm/W are in the same ballpark as what I've seen. I did get up to about 50 Lm/W as I reported above after working through about eight different bulbs and modifying Lynxsteam's AIR CORE approach.
That 1w LED bulb I found to be "best" is indeed something you might try, also the air-core approach. (Kudos to Lynxsteam for that work.)

Again, Peanutbutter being honest and forthright here! :

I want to state here as on the video I sincerely apologize for false numbers as I should not have such confidence in the old meter and method both. I do not ever intend to mislead or mis-state and I have done that here in a big way. I have removed light % estimates from titles of the 2 videos that had it too.

I don't want anyone else to get discouraged or stop looking into these as a result of my mistake. I think there IS merit here and; I suppose all these currently may well suit a persons needs. Lidmotor said this best " It works for me"

I think also, this re-enforced PhysicsProf light box as a required item to get any grasp on real light output. Light to the eye is just too deceiving and other methods not accurate.

Right! we are learning and progressing and MEASURING devices make this progress accelerate IMO.
Kudos to Peanutbutter! and let us press onward.

[blackchisel97]

Peanutbutter29 Thank you for the effort and great job in taking those measurements

-"and let us press onward". I second that.

I did some tests on other small transformers, mainly looking for availability. I found two in discarded power backup supply for PC. It has two secondaries but I used only one atm. and got it running. I was busy today gathering parts for pyrolysis unit and looking for large transformer to try Matt's scalar charger but I hope to get some more work done with ringer this week.

Thanks

Vtech

[Jules Tresor]

Thanks PB for your honesty and courage to announce your mistake
Thanks for showing that a light-box is a must to find any super efficient circuit ... eyes aren't enough
Now you have a very good ground to start and research ! Thanks PB for sharing
Jules

[Peanutbutter29]

TYVM for the good comments. I didn't know what to expect as a response. Very kind words so far. Someone has to give me crap, heh. Guess you can always count on YT for that :P

Video link finally,
SJRC 1.0 ACCURATE lighting results. Dissapointing....Apologies & Does tuning help - YouTube
lot of neat stuff other than terrible report, sigh.

PB

[totoalas]

TYVM for the good comments. I didn't know what to expect as a response. Very kind words so far. Someone has to give me crap, heh. Guess you can always count on YT for that :P

Video link finally,
SJRC 1.0 ACCURATE lighting results. Dissapointing....Apologies & Does tuning help - YouTube
lot of neat stuff other than terrible report, sigh.

PB

DIY construction has given a wide range of selection off the shelf parts AS LONG AS IT WORKS PEOPLE ARE USED TO CANDLE POWER SO A LITTLE
EYE COMFORT MAKES GOOD SENSE

totoalas

[Billxx]

TYVM for the good comments. I didn't know what to expect as a response. Very kind words so far. Someone has to give me crap, heh. Guess you can always count on YT for that :P

Video link finally,
SJRC 1.0 ACCURATE lighting results. Dissapointing....Apologies & Does tuning help - YouTube
lot of neat stuff other than terrible report, sigh.

PB

@Peanutbutter,

Interesting video, just not sure about the need for an apology. I guess what you're saying is that the total luminance of the 6 bulbs is a percentage greater rather than a multiplier of 1 bulb? This will greatly affect the numbers, no doubt, but, you didn't include in your video what the total luminance is when your strip is plugged directly into a wall outlet (1 bulb and multi-bulbs) to show the difference collectively with multi-bulbs.

How does the efficiency compare to using a standard battery/inverter arrangement?

I'm only a student in this, not a teacher with any expertise, and I'm not criticizing anything or looking to give ability to explain away any mistaken calcs, we all make mistakes, but, are the comparisons in the right place?

You're adding a considerable amount of interesting progress to this project, why would you entertain stopping now?

Please continue...

Thanks

[Peanutbutter29]

Ty again for comments, I'm encouraged and surprised!

@ billxx to try and answer

I could probably do an hour video or longer just on findings from yesterday and ran through all kinds of variations. Both ensuring accuracy and seeing what's more efficient.

For the first question. Yes it's a win / lose (in general and un-tuned) with multiple bulbs. now I haven't verified with others, so this is one transformer and may not always be the case.
But yes. 1 bulb un-tuned was around 30 lu/w and 7 dropped it to 25 Lu/W (or less). So this is bad as efficiency dropped.
However, lumens with one was 338 total (untuned) and with 7 it was 493 total (tuned). So, from this perspective more is better for max light. Since I was checking with that old meter, this makes sense and matches (at least increase light I was seeing).

*I suppose what is considered "reading" light level or office light are both 500 lumens. From this perspective, 7 bulbs would then give a "reading" level whereas it isn't possible with one.

If your going for "what works" I suppose 7 still permits more locations off one circuit.

Next, total light numbers. For shown bulb placement, you can see a "mains" reference on the paper. It was 13,300 Lux with 6 bulbs & 47w. The conversion (2940 /13300) was the .221 I used.
A note here, I DID and can show running these vertically on 2 spike bars. Here I apply the .156 from the top of the page (has more reference points). I mention that the variation in lumens (using 2 calibrations and placements) is very close; within 10 lumens.

Efficiency to a standard inverter....well here's a kicker. I don't have one, so let's take an "estimation" from Lidmotor and Lasersaber shown inverters. If we assume 300ma of inverter draw alone that's 3.6w. If we run the 7.5w on that @ 490 lumens, we have a total cost of around 11.1watts.

This means, to exceed an inverter (assuming 300ma) we have to exceed (Actual) 490 lumens for less than 11.1 watts.
My Shown and tuned version then hits the mark for lumens, but for 4.62w more. Still a ways to go here, but from tuning improving values I have hope!

**I stated tuning improvement poorly on video as I said % instead of Lu/w.
So, net improvement (for this coil and 7) is 5 Lu/W, 155 lumens and the charging. This works out to 20% INcrease in efficiency over un-tuned. Pretty decent I suppose.

Again I can show all kinds of stuff. I've not been able to get full "tuning" on one bulb, so that worries me, since efficiency is higher there. However, I DID exceed 1 bulb efficiency by 1%, so I'll have to see. Also, I suppose, if I'm trying to hit reading 500 lumen range, that may be tough with one bulb for efficiency too.

I ran all modes of current with all bulbs, from 1 to 9. I ran currents from 760mw up to 20w. In all cases, tuned or not; efficiency was always HIGHER with MORE current (less base resistance). This matches what I assumed, so I imagine this is will apply. Lowest of 760mw was 15 lu/w and 9 bulbs max was around 20 lu/w.

Also, since I bought 2 of the WW 7.5 dimmables that are the same brand; I compared those. These were run vertical and I used each of their conversions (.156 daylight, .184 warm white LED). Interestingly; Waveform same, current same, sound same, power same......Same bulb company and wattage, but what about efficiency? Well, the daylights are 4% more efficient. This is about 1/2 of the rated difference from the package of 8% (450 Lu vs 490 Lu).

Again, all kinds of info that opens up new perspectives on tuning and circuits. I'm MORE anxious now about things I was working on a couple days ago; as this new circuit may work well; dunno.

Hope this helps clarify some and I'll try to get some sort of informative on this up maybe.

Thanks again,
PB

[SkyWatcher]

Hi folks, Hi PB, well my inverter draws 2.2 amps-12 volt input, while my boxed standard transformer super joule ringer 2.0 runs the 15 watt cfl at 1.2 amps-12 volt input.
Is the brightness the same, by looking at the bulb and how it lights the room, it looks equivalent to me.
That is what matters to me, does the device provide the same usable light in a room or not.
From my readings, at least these cfl bulbs, not sure about the led bulbs, this 15 cfl is drawing 26.4 watts or so because of the power factor in relation to the AC voltage wave, I assume if i hooked a killiwatt meter between it and my AC wall power, it would also draw around 27 watts or so.
Though these joule ringer devices are almost doubling that efficiency, either by better power factor, or the fact we are using the collapsing field also.
So it works for me, the SJR2.0 will get around 6 hours of run time with my 7AH SLA battery powering the 15 watt cfl, where as with the inverter, it will get around 3 hours run time.
And to be honest peanut butter, I forgive everyone no matter what they did and even if it was intentional or not.
We are all living in a dimension where we tend to forget who and what we really are, pure unconditional loving beings.
peace love light
tyson