LED driver-may have hit the jackpot.

Author Topic: LED driver-may have hit the jackpot. (Read 1148 times)

Topic Options

TinMan

Re: LED driver-may have hit the jackpot. « Reply #25 on: 2014-09-14, 13:35:31 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Darren

Please remember i was using the lowest voltage value when doing my calculation's.Looking at the scope shot across LED 2 below,we can clearly see that the average voltage is indeed 2.8 volt's(2.9v down to 2.7v over the cycl) Now i know it takes 2mA to drive those LED's at 2.8v-per the test i carried out in post 22.

2.8v x 60% x 2mA=3.36mW
Or-2.8v x 2mA x 60%=3.36mW

EDIT:-In fact Darren,i forgot all about the inductor,so i just went out and checked the current for the inductor and LED 1 combined.I have 3.88mA -->voltage across the two is the SG voltage in-->7.4 volt's x 40%.

LED driver-may have hit the jackpot.

Across LED 2 test 2-average voltage.JPG (32.45 kB, 624x304 - viewed 85 times.)

« Last Edit: 2014-09-14, 15:20:37 by TinMan »

---------------------------

Never let your schooling get in the way of your education.

TinMan	Re: LED driver-may have hit the jackpot. « Reply #26 on: 2014-09-14, 22:25:49 »
Group: Elite Experimentalist Hero Member Posts: 5677	I have ordered a few different size inductors,and also some 10mm super white LED's. I will have 2 sets of 2 parallel LED in the next test circuit,so as i can up the P/in P/out of the circuit.This will all be set out on a test board,so as i can messure P/out(dissipation) of each component individually. --------------------------- Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #27 on: 2014-09-15, 15:31:27 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

@ Darren

My time is limited now I'm back at work,but i managed the following test. I ask if you can take the time,and have a look at the scope shots,and crunch some numbers. All to often we look for the exotic and complicated to try and achieve OU,when there is as much chance at finding it within the simplest of circuit's-maybe?.

I also took the time to test my DMM's accuracy at these frequencies and duty cycle-all my DMM's are the same. Once again they are spot on.
DMM test 1-i placed a 22ohm 1 watt resistor between the circuit and SG(negative side of course) Here are the result's
DMM 1 reads 43.2mV across the 22 ohm resistor,scope reads 44 V/avg.
DMM 2 is reading the current-I/in. It reads 1.96mA.
43.2mV across a 22ohm resistor=1.96mA. This confirms my DMM's are accurate.

With regards to the scope shots below,my I/in is 1.96mA.
So my P/in is-->2.88v(the average voltage)x 1.96mA(the average current)=5.6448mW
With these new LED's(larger ones)It take 3.66mA to run them at 2.8v.This is 10.248mW. Not forgetting that my circuit runs a second LED at a 60% duty cycle@2.8 volts. This is another 3.2928mW.

There is also my light box circuit(as pictured a few post back.I get far more P/out from the light box using my circuit than i do from straight DC-even with a higher DC P/in.

LED driver-may have hit the jackpot.

ACROSS SG-CIRCUIT TEST 3.JPG (50.06 kB, 720x351 - viewed 79 times.)

LED driver-may have hit the jackpot.

ACROSS LED 1 TEST 3.JPG (43.35 kB, 720x351 - viewed 77 times.)

LED driver-may have hit the jackpot.

ACROSS LED 2 TEST 3.JPG (49.84 kB, 720x351 - viewed 77 times.)

« Last Edit: 2014-09-15, 23:15:17 by TinMan »

---------------------------

Never let your schooling get in the way of your education.

poynt99	Re: LED driver-may have hit the jackpot. « Reply #28 on: 2014-09-16, 03:25:43 »
Group: Administrator Hero Member Posts: 3538 It's not as complicated as it may seem...	OK Brad, I'll have a closer look in a day or two. --------------------------- "Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

TinMan	Re: LED driver-may have hit the jackpot. « Reply #29 on: 2014-09-16, 07:58:10 »
Group: Elite Experimentalist Hero Member Posts: 5677	Thanks Darren To the rest of the fredom energy guys-your input is more than welcome. @ Steve We may have to revisit your ou pulse circuit. --------------------------- Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #30 on: 2014-09-16, 14:41:18 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Chet has asked me to post the below result's and setup.

While i am aware that a pulsed LED using less P/in than that driven by a DC can look brighter to the human eye,a solar pannel is not fooled in the same way. The output from a solar pannel is relient on the true amount of light the pannel is exposed to.

Below i have reposted the light box setup. This is a sealed box that houses the two LED's and a small solar pannel from a garden light. the P/out from the solar pannel is hooked in parallel with a 200uf cap and a 4.7k resistor. From this we can measure the light output power from the LED's by measureing the voltage across the resistor/capacitor/solar pannel.

The result's are as follow's.

DC P/in to LED=2.8v @ 3.76mA. This is 10.528mW.
Output from light box(across pannel/resistor/cap) is 251mV/4.7k

SG P/in to circuit. 7.6vpp at 40% duty cycle. This is an average voltage of 3.04v.
I/in average is 2.23mA. This is 6.7792mW.
Output from light box is-325mV/4.7k

So we use 3.7488mW less,and we get 29.48% more P/out from the light box.

LED driver-may have hit the jackpot.

LIGHT BOX SETUP 1.JPG (17.91 kB, 713x376 - viewed 75 times.)

---------------------------

Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #31 on: 2014-09-16, 14:56:33 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

The scope shot below is over a 220 ohm 5 watt resistor with an exact value of 216.1 ohm's.
This is as clear as i could get the wave form over a resistor to get our I/in. If some one could tell me the current value from this shot,then i can compair that to what my meter says-which is 2.59mA

Cheers
Brad

LED driver-may have hit the jackpot.

216.1 OHMS.JPG (47.31 kB, 720x351 - viewed 92 times.)

---------------------------

Never let your schooling get in the way of your education.

PhysicsProf

Re: LED driver-may have hit the jackpot. « Reply #32 on: 2014-09-16, 15:57:38 »

Group: Professor
Hero Member

Posts: 3249

This is very interesting, Brad. Especially your numbers are looking good

Quote

With regards to the scope shots below,my I/in is 1.96mA.
So my P/in is-->2.88v(the average voltage)x 1.96mA(the average current)=5.6448mW

With these new LED's(larger ones)It take 3.66mA to run them at 2.8v.This is 10.248mW. Not forgetting that my circuit runs a second LED at a 60% duty cycle@2.8 volts. This is another 3.2928mW.

There is also my light box circuit(as pictured a few post back.I get far more P/out from the light box using my circuit than i do from straight DC-even with a higher DC P/in.

I've played a lot with LED's as you know, and enjoy all that. I've got tons of LED's now, different colors etc..

Not nearly as fun, but easier to get a solid value on Pout perhaps --
would be to replace the LED's with a diode and a resistor, then place these in water (in an insulated vessel) and measure temp-rise of the water. Simple calorimeter. Clear results.
If it still "works" with the diode+resistor in lieu of LED's.

I'm also interested in your HHO work, Brad.
--Steve

TinMan

Re: LED driver-may have hit the jackpot. « Reply #33 on: 2014-09-16, 22:29:14 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: PhysicsProf on 2014-09-16, 15:57:38

This is very interesting, Brad. Especially your numbers are looking good

I've played a lot with LED's as you know, and enjoy all that. I've got tons of LED's now, different colors etc..

Not nearly as fun, but easier to get a solid value on Pout perhaps --
would be to replace the LED's with a diode and a resistor, then place these in water (in an insulated vessel) and measure temp-rise of the water. Simple calorimeter. Clear results.
If it still "works" with the diode+resistor in lieu of LED's.

I'm also interested in your HHO work, Brad.
--Steve

I was going to try replacing the LED's with a resistor and diode,but first i must be absolutely sure my DMM's are reading the correct average I/in. Every test i have done so far says they are correct,and if they are,then this system is running at 140%+. This circuit is based around a duel figure 8 current loop,and as soon as i get some feed back on the scope shot's(so as i know I'm reading them correctly). Tonight after work,i will post some scope shot's showing exactly what is happening within the circuit,along with an explanation.

Brad

---------------------------

Never let your schooling get in the way of your education.

Itsu	Re: LED driver-may have hit the jackpot. « Reply #34 on: 2014-09-17, 08:18:27 »
Group: Elite Experimentalist Hero Member Posts: 4848	Brad, i tried to replicate, but the "ceramic inductor" has a huge impact on the signal. Do you have any specifications on this inductor, like dc resistance / inductance? (guess the inductance is rather high, as in mH range). Regards Kees « Last Edit: 2014-09-17, 10:00:44 by Itsu »

TinMan

Re: LED driver-may have hit the jackpot. « Reply #35 on: 2014-09-17, 12:51:11 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: Itsu on 2014-09-17, 08:18:27

Brad,

i tried to replicate, but the "ceramic inductor" has a huge impact on the signal.
Do you have any specifications on this inductor, like dc resistance / inductance? (guess the inductance is rather high, as in mH range).

Regards Kees

Kees

Im not sure how you replicated the circuit,as i havnt posted it yet. The circuit in my first post was just a sample so as i could ask the question on how to use a resistor and scope to accuratly get the I/in for such a circuit. Befor i post the circuit,i want to make sure that my P/in measurements asre correct-other wise im just wasting your time and mine.But i believe that my DMM's are reading correct,and this will be shown in the next few post,along with what the circuit is doing.

Brad

---------------------------

Never let your schooling get in the way of your education.

Itsu

Re: LED driver-may have hit the jackpot. « Reply #36 on: 2014-09-17, 13:03:20 »

Group: Elite Experimentalist
Hero Member

Posts: 4848

Quote from: TinMan on 2014-09-17, 12:51:11

Kees

Im not sure how you replicated the circuit,as i havnt posted it yet. The circuit in my first post was just a sample so as i could ask the question on how to use a resistor and scope to accuratly get the I/in for such a circuit. Befor i post the circuit,i want to make sure that my P/in measurements asre correct-other wise im just wasting your time and mine.But i believe that my DMM's are reading correct,and this will be shown in the next few post,along with what the circuit is doing.

Brad

Well, i just toke a resistor (in fact 2, one of 3.3 Ohm, and one of 220 Ohm), a led and an inductor and put the SG as you mentioned (square wave, 4.1V all positive offset at 127KHz).

The signals are similar as you showed, only when using a high inductance coil (5.5mH).
All other coils with lower values shown more square wave like signals across the resistors, not the peak like you have.

I do have severe peaks on the SG square wave signal transitions which skew the scopes measurement, so they are NOT in agreement with my Fluke DMM current (rms) value.

Anyway, i will await further posts of you or others, thanks.

Regards Kees

TinMan

Re: LED driver-may have hit the jackpot. « Reply #37 on: 2014-09-17, 14:25:40 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

just made a mod to the circuit. I have now reduced the P/in to only 3.42mW to achieve more light power than that of 10.52mW DC. This is 307.6% efficiency to that of DC. All my new 10mm LED's and inductors of all different value's arrived today. So this weekend(starting Friday night) i will be putting together a test bed,so as we can go through each component and get the P/in and P/out of all components.

Below i have a scope shot showing the trace across LED 1 and LED 2. I have dropped the duty cycle to 39% on the SG ,so as you can see on the blue trace(LED 1)where the P/in starts and finishes.This also lines up with the bottom half of the yellow trace(our 39% on time) LED 1 and the inductor receive there power from the SG for 39% of 1 cycle,and the inductor supplies both LED 1 and LED 2 with power the other 60% of the cycle. At a 40% duty cycle,LED 1 is on 100% of the time(as shown in other scope shots here on this thread).

As soon as I'm sure that i have what i think i have,i will post the circuit. I'm happy and wanting others to replicate it and test for them self's,but only with a real device-no sim's results to be taken notice of. I'm sorry it has to be that way,but if we are looking for the unknown,then we wont find it in a simulator based around the known.

LED driver-may have hit the jackpot.

LED 1&2 TRACE 1.JPG (53.55 kB, 720x351 - viewed 26 times.)

---------------------------

Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #38 on: 2014-09-17, 14:50:37 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Below is the modified circuit. I have added 2 diodes( D1 and D3 ).Because of the voltage loss,you will have to lift the V/in to around 9vpp. Adding these two diodes now eliminates the SG as being the current path during the off period. This halved the P/in needed to achieve the same output result's-for some reason???.

For those who wish to replicate and post there result's,could you set up a simple light box(as i posted),and check the light output power from the circuit,and then that of the LED running from straight DC current.
Thanks.

EDIT:
D1 is around the wrong way in the circuit below.

LED driver-may have hit the jackpot.

LED DRIVER CIRCUIT V2B.JPG (41.14 kB, 713x376 - viewed 29 times.)

« Last Edit: 2014-09-17, 23:21:54 by TinMan »

---------------------------

Never let your schooling get in the way of your education.

Slider2732

Re: LED driver-may have hit the jackpot. « Reply #39 on: 2014-09-17, 18:09:05 »

Group: Elite Experimentalist
Hero Member

Posts: 1602
... .-.. .. -.. . .-.

Looks like we're all jumping in with this !

Chet rang and said to check it out.
Also, sorry for my recent absence, got some sciatica and hip troubles.

"we can obtain a higher voltage across the light box circuit using the pulsed circuit than we can using a DC current"
That is the key quote that i've taken from the writings so far.

FYI, but related to this thread perhaps, a Cool Joule has been running since the middle of June. It's got 2x 1.2V 200mAh Ni-Cd AAA's on it and 'dancing flower' 500ohm coils. The circuit began running when Lasersaber first showed his fancy coil/low energy system and Brad mentioned the Cool Joule...its run 24/7 ever since, just needing a tap on the Emitter+Collector to fire up and run for a week at a time if it stops. No battery swap overs since the beginning of August.

Own tests usually show a 50% duty cycle to be the optimum, so 38% is intriguing. The inductor is another area to explore, maybe a 'dancing flower' coil with ferrite either side...or does the inductor have to be low uH for response timings ?
I'm going to try a variety of LED's.
See, greens are great at generating all on their own in bright light, in common with other exposed silicon....so it's own output may affect P/in. For example, 4x 3mm regular frosted greens will power a cheapo digital watch in sunlight.
Reds for their lower voltage needs, ultrabrights for light, try I/R types too.
A string of them relates well to your fathers device Brad.
In my opinion, we can use a combination of this circuit, Lasersabers efficient LED driver and the effects passed from one LED to another.
Efficient collection and harvesting would then point to practical OU or only a neat curio.

ETA: for the light box, I will be using an amorphous panel, they work best with non natural light.
ETA2: An opto-coupler would seem a good alternative to a light box ?
And, have different waveshapes other than square been tried ? no point ? Just wondering about a ramp type, with a square top to drive the inductor differently.

« Last Edit: 2014-09-17, 20:26:56 by Slider2732 »

---------------------------

ʎɐqǝ from pɹɐoqʎǝʞ a ʎnq ɹǝʌǝu

PhysicsProf

Re: LED driver-may have hit the jackpot. « Reply #40 on: 2014-09-17, 22:03:16 »

Group: Professor
Hero Member

Posts: 3249

Great stuff, Brad! I'm excited to get going on this.
Thanks for INVITING replications!

Did you state the approx frequency needed from the SG? (sorry if I missed it)

It would be nice to get a simpler SG than my big Cenco SG box... Is some sort of simple circuit to "replace" the SG possible?

PS - I bought an SG from china a while back; will look for it. Its a little solid-state device...

« Last Edit: 2014-09-17, 23:28:39 by PhysicsProf »

TinMan	Re: LED driver-may have hit the jackpot. « Reply #41 on: 2014-09-17, 22:20:37 »
Group: Elite Experimentalist Hero Member Posts: 5677	Sorry guy's,i got D1 ass about in the first circuit post. Very tired. LED driver-may have hit the jackpot. LED DRIVER CIRCUIT V2B1.JPG (41.03 kB, 713x376 - viewed 38 times.) --------------------------- Never let your schooling get in the way of your education.

poynt99	Re: LED driver-may have hit the jackpot. « Reply #42 on: 2014-09-18, 00:42:09 »
Group: Administrator Hero Member Posts: 3538 It's not as complicated as it may seem...	Brad do you have a part number for those super brights? --------------------------- "Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

TinMan	Re: LED driver-may have hit the jackpot. « Reply #43 on: 2014-09-18, 01:47:53 »
Group: Elite Experimentalist Hero Member Posts: 5677	Quote from: poynt99 on 2014-09-18, 00:42:09 Brad do you have a part number for those super brights? No sorry Darren, but I do for the 10mm ones that turned up yesterday-that im yet to try. I will post that part number when I get home tonight. --------------------------- Never let your schooling get in the way of your education.

poynt99	Re: LED driver-may have hit the jackpot. « Reply #44 on: 2014-09-18, 02:17:32 »
Group: Administrator Hero Member Posts: 3538 It's not as complicated as it may seem...	What is the resistance of your coil? You pretty certain of the 1500uH spec? I'm simming this...or trying to. --------------------------- "Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

poynt99

Re: LED driver-may have hit the jackpot. « Reply #45 on: 2014-09-18, 02:24:33 »

Group: Administrator
Hero Member

Posts: 3538
It's not as complicated as it may seem...

You've not shown your CVR resistor on the input (or gnd side). What value should I use? And where in the circuit is it installed?

btw Brad, try the circuit without D1, I don't think it is required.

---------------------------

"Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

TinMan

Re: LED driver-may have hit the jackpot. « Reply #46 on: 2014-09-18, 03:10:09 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: poynt99 on 2014-09-18, 02:24:33

You've not shown your CVR resistor on the input (or gnd side). What value should I use? And where in the circuit is it installed?

btw Brad, try the circuit without D1, I don't think it is required.

Im using my DMM as my CVR, but I would place it on the ground side of the circuit/SG.The value I used at these low P/in is 220ohms, but of course this reflects on the circuits operation.

D1 is just there to drop the voltage so as D2will be the primary current path during the off period

---------------------------

Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #47 on: 2014-09-18, 13:23:50 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Looking at the scope shot below that shows voltage and current,we can see the P/in. From this i can also see that my DMM amp meter is reading correctly. The average voltage is quite easy to work out from the trace(blue) which is across a 216.1 ohm 5 watt resistor,as it is almost a straight incline.It seems to me that the average voltage across that resistor is 800mV,and the average voltage across the circuit/SG is 7.7 volts.

So 800mV over 216.1 ohms is 3.70mA x 40%= an average current of 1.48mA. My DMM in this test read 1.51mA-so very close.
7.7v x 40% is 3.08 volt's.
P/in is 4.558mW going by the scope,and 4.65mW going by my DMM. A difference of 0.092mW's.
I have run this test at 5 different voltage levels and frequencies,and the DMM is always within 2% of what ohms law says it should be.

LED driver-may have hit the jackpot.

P-in scope shot..JPG (49.27 kB, 720x351 - viewed 24 times.)

---------------------------

Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #48 on: 2014-09-18, 13:57:43 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: poynt99 on 2014-09-18, 02:17:32

What is the resistance of your coil?

You pretty certain of the 1500uH spec?

I'm simming this...or trying to.

Looking a bit closer,it may be a violet band-not a brown one(hard to tell)
So that being the case,then the inductor would be 7.5 Millihenrys. The resistance is 9.2 ohm's.

Schematic update below.

LED driver-may have hit the jackpot.

LED DRIVER CIRCUIT V mod.JPG (42.42 kB, 713x376 - viewed 27 times.)

---------------------------

Never let your schooling get in the way of your education.

TinMan

Re: LED driver-may have hit the jackpot. « Reply #49 on: 2014-09-18, 14:12:13 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: poynt99 on 2014-09-18, 00:42:09

Brad do you have a part number for those super brights?

These are the ones i will be using from now on,hopefully there will be no change to the way the circuit performs. The one i have been using are from one of those dynamo torches.

Viewing Angle: 25 degrees
Brightness: 5500mcd
Voltage Typical: 3.5V
10mm water clear
Part number-Z0908

---------------------------

Never let your schooling get in the way of your education.