Joule Ringer!

[Slider2732]

Agreed Totoalas, especially with the fact that many poorer countries will have busted up radios etc from which to draw parts from.
I'd like to see landfill site volunteers, actively engaged in salvage. Trucks of such items could then be sent out and volunteers at the other end would help with first assemblies.

Progress here...not that I wish this to be a blog. Various caps later, a .1uF across the transformer input has proved to be most effective. Less bright at high pot resistances, but when cranked up reaches a better brightness than without it. Best light to draw ratio is 0.33A at 3/4 travel of the 50K pot. Less resistance and up to 0.56A draw, only increases brightness by a 'smidge'.
A lightbox is now being built (to remove the word 'smidge' from subsequent posts).
Anyone tried a LDR and an analog VU meter ? just an idea

Here's my circuit so far, with the 2W LOA LED bulb:



And an update.
I built another, to try another transformer (from old hi-fi and marked TKP0108-A124). Just a pot and 82ohm resistor to Base of a D4793.
It's 3x better
Slightly brighter illumination and yet with only 110mA at 1 particular spot on the 50K pot. Less resistance shows up to an Amp and more, but around that little sweet spot at 12V, the draw is drastically reduced in comparison.
Now, on to caps and such with this one
I should mention that the Positive goes to the center lead of the 3. One transformer input goes to the collector, one to the pot and on to the base., It's a simple tranny switch at the moment with variable Base resistance.

Here it is, same 2W LOA LED bulb:

[totoalas]

Super Joule Ringer Lux Test 220512

Thanks to All and to Lasersaber , Peanutbutter , and Xee2 et al

This team effort of ours will save lives and properties
In the Philippines where used candle is the no 1 killer and cause of fires with or without electricity
hard earned money for food is diverted to electricity bills which they say rank 10th highest in the world
back to the led lighting
Using a Lux meter ( x 100 lux) 220 v ac 5 watt led read more than 20 the SJR 12 v dc 270mA with 8 5 watts led lamp read 6 to 8 in the scale
With 50 lux ( barely lit) and 60 mA the circuit is charging back the source by 1.14 v dc ( test in progress) after 4 hours the MJE 121194 is stone cold same as the transformer resistors @ 270 mA
Super Joule Ringer Lux Test 220512 - YouTube

totoalas

[Slider2732]

long as usual

Who cares ?! it's great information

I now see what you mean about the bounce condition and of the double effects within Darlingtons. Also, better understanding of how a Base resistor with these higher spec transistors really does bottom out, yet too high and there's no strike at all...all being the familiar act of balancing. The change last night was to an NPN, regular common or garden flyback transistor with 1500V ratings (removed from a CRT monitor). I usually use something around 100W capability (though 70W with this one: https://docs.google.com/viewer?a=v&q...WxRthu08Rfy4ew ). I mean, just like your 2N2222A experiment, it's not really a cool idea to blast signal trannies is it...unless the wish is to emulate a Mythbusters 'blow stuff up with dynamite' situation for the heck of it lol

Am back engineering at the moment. A previous circuit, using a C5929, with diode and cap to Base, 330ohm from pot to transformer winding. It uses the center tap connection for the Positive supply and the aim was always to reduce current used with that type...though they are similar to these present circuits. Point being, now that I have a couple of LED bulbs, the output is 'really good' as tested just now on that. So, being as I know that transformer is good (marked as WH35-14B4C), am going to probably alter that one to your circuit, then to Lidmotor's version

[Peanutbutter29]

Hello all! I suppose I can try and answer some if possible. I'll try to go in order.

@ Totoalas, very nice results!! I'm so glad that it helped and you have a good "real world" use for it. Running 16 hours and still having another 3aH left is nice!
As for dropping current with the caps, yes. That's one of the things that "can" let you know the coil is tunable versus L2 dominate.
**Not sure if anyone has done it with my version shown; but with the 9 bulbs I had running. Removing the L2 cap, raises (yes raises) input 200mA or 2.4watts with almost No light change, this is a key point! Conversely, with the v1.0 coil removing the L2 cap lowered current, BUT light dropped a large amount, so highest efficiency was still with both connected. Again, though v1.0 is in a "Somewhat" separate class, that I have not replicated yet.

@ Jules first question, The radio shack is plain ole' laminated Iron core. No specials. This is partly why its' not "as" efficient as other designs; but our operation frequency below 2khz should keep this somewhat minimal.

From Lidmotor stating on YT that the open circuit is drawing 3ma I did a little rough estimation of core loss. Don't hate on me as this is rough and an educated guess at best. However, I'd been doing some calcs on dissipation of the tuned circuit and I believe I can get an estimation here.
I calculated then, the open circuit "most likely" is dissipating about 2mW or a bit less (Hfe 10, Vce .2v, Ice 3ma, Vbe 2v, Ibe .3ma)in the transistor and then 34mW to core loss. Again, this would be a higher frequency than operational; which will increase loss. However, this is also at very low power, so loss is lower than operation from this perspective.

At any rate, I'm going to toss out that this "tuned" (12v 1.3A) is loosing appx. 350-700mW in core loss. The dissipation from the transistor (a more accurate number than core loss) should be around 100-125mW (Hfe 25, Vce .1v, Ice 1.3-1.5A, Vbe 1.5-2.0v, Ibe 50-60ma)

*note Base current limit resistor value (220) and max Ibe .

Next,
@Slider2732 about the Strobe effect. Yes, I've seen it in few cases; so possibly I can help. But also (since tranny is unknown) we need to know if it's "tunable" or L2 dominate (too high ratio) or poor winding for this type of oscillation (not likely to ruin operation). No worries here as we should get bulbs working, then you can find tuning.

Three situations that you can get a strobe; this is much more common for compact than LED.
First, if the L2 inductance is too low; then the circuit cannot effectively deliver power or deliver power at the right time for full operation. In this case you will note the bulb Off more and "popping" into full lit mode less often.
Second, if L2 is too large, lol. I saw this with a big MOT of 8 henries and was noted particularly with 6 bulbs (LED here). If we have a HUGE L2 and come "close" to the internal bulb circuit; there will be a more direct "attempted" resonating. Here, rather differently you get a more realistic strobe of "full off" to "full on". This is related to an actual cancellation or beat of L2 with load.
Third, as in the case with Lidmotor and the CFL with this setup. (Again, it should light fine as I show IF you start with resistor at 220 ohm) This a case were L2 is "just" enough to run the PC. Also, PC's circuit's "Buck" more and why they cannot be connected in parallel. However, the case here is a changing of the "base bias voltage" from this bucking. I talked before about the negative bias and positive bias mode. Basically, if it does not start properly (e.g. too high base res. or fast V rise) you are seeing a "bouncing" from negative to positive. Here, similar to case 1, you have a low to high light state. However, the bulb in low state is almost not visible if PC. The oscillation is going on, but we have not "popped" into stable operation and the base is still "floating".

As for suggestions. First thing, please try an NPN rather than a PNP. I believe, they usually have a lower Hfe and a higher junction capacitance (not listed with bipolars, like FETS), but not a great thing here.
Also, all transistors introduce a phase variation relative to the base and collector currents. Darlington type transistors are in fact 2 in one case. This allows for higher Hfe, which is good yes; but the doubling of the phase variation is NOT. I'm sure there may be a situation where these will work well, but I've not been happy at all with Darlington's and these SJR's.

Please FIRST try a normal NPN bipolar of 80w or more as this may be all you need to change to start. Audio output transistors are better quality than General OP's like the 2n3055. Many will work though, if robust and a nice low junction capacitance. I prefer the NTE36 best, so far, (100w, Hfe 100, Ft 15mHz); followed by 2n3773 and the 2n3055 last. Mainly last on the 3055, cause it does not deal well when tuning, lol.


@ Penno Jules and all 220v.
I have been chewing on this some, so here's where I'm at.

First, not the best choice, but most reliable. Order bulbs and tranny from US so you can use 120v. 120v trannies may be over there, so possibly just bulbs. Mainly since we are running of 12v DC; we can use any 12v supply and are not restricted.

Second, yes it is possible a 220-24v may work; but I have not tested these or 220 bulbs to know. Here we are trying to double step up AND doubling input inductance; while outputting to a load of similar frequency. I cannot even theorize without some trial; but I'm inclined towards a no.

Third, theoretically we could A. use the center tap on a similar transformer (so, 6-120v); but this would definitely make it L2 dominate and non-tunable. Or B., the v1.1 I show. I tuned that one using essentially a 120-26v ratio of winds (since I bridged two 13's). This leads me to believe that the same toroidal at 12-120 would work for 220, possibly; but still introduces L2 dominance being likely.

Fourth, I can try and get a 220 tranny or two and a dimmable LED 220 if I can. I won't be able to use 6 of these for tuning, but I may get close if tunable. If I can run some tests on something in this range I might get known aspects, then.

Finally then @ Lidmotor. Cool! glad to hear explanations are okay. I fully know that some of what I have / will talk about is odd from what most may "jump" to; but again I don't feel there's any violation of common laws here.

I've spent a lot of time over the years I suppose with general philosophy. I do feel that a large part of reality is based on perception (no, not nuts like I'm a banana); you'll find then I tend to "move" my perspective around when explaining things. I do try to point out, from which perspective I'm looking though.

Finally, I DO want to finish / better explain what I do and how I'm looking at this. However, as I said before; its' easier to show you how to fish (say do this) than it is to teach you how to fish (so you understand). I'm also posting a lot already and don't want to flood things; but I may have to show first to help with the 220's.

I much prefer to relay the understanding first, as I perceive; knowing the same as knowledge, and understanding the same as wisdom. I can apply knowledge if a situation arises that applies to something I know; however, I can apply wisdom to all things.

Hope this helps. eh, long as usual...aye!
Thanks
PB

EDIT:
@Bill glad the tutorial helped! I tried to explain everything and those 2n3055's don't all show pin-outs on the pack which sucks.

@PhysicsProf nice job with the bulbs! I'm always overjoyed at how "good quality" everything is now, lol. I think better brands will give better results, if we can get such a thing. Also, heat with these will prolly be what kills them in the end (on line volts) that may vary with mfg. too. If they have electrolytics in them like PC's then the Hour rating on those, will be when they fry.

[Slider2732]

@Limotor - Thanks, it's a funny thing that oscilloscope measurements or in fact anything further than simple point to point continuity checks usually only happen with a fully running circuit
I changed the lot, tried a C4924 with another, larger, transformer. The result was again a flash and yet the ability to see the darn LED brighten with the pot. However, I realised the flash is brighter on this transformer with increased current.
Time to RTDM with what you said and build PB's version first.

@PhysicsProf - Looks like one of those customer in the store situations with the bulbs. 3 on the shelf, 2 that say 400 Lumen and the other says 600, all the same size and price. Any regular customer is unlikely to check. Still, plenty there for tests and valuable info for others with those bulbs
People have however had a century to check the incandescent rating.

[PhysicsProf]

Today I received three LED bulbs purchased from Tmart (Chinese-made), and measured them in the light-box system. Compared with the bulb from Costco and an incandescent 40W bulb (the latter two bulbs were used in calibrating the box).

I would say the results were disappointing in that the Tmart bulbs did not appear to put out the advertised Lumens/Watt -- as advertised in the Tmart ads, although there may be a correction given the fact that these bulbs put out light mostly UPwards. However, I moved the light-detector upwards (still on the side of the box), and the lux-observed did not change much.

See attached. The bulb on the left (smallest) was said to put out 90 lumens for a 1 watt bulb. In fact, it ran at 3.6 W on my watt-meter (not 1 W at all) and I measured 233 lumens.
Next bulb (moving right) was supposed to put out 300 Lumens for a 3 W warm-white bulb. Running at 3.6 W, it only put out about 175 Lumens.

The largest bulb was supposed to put out 900 Lumens for a 10 W bulb. I found about 600 Lumens. This is however almost the same as the next bulb to the right -- 600 Lumens from a 9.5W bulb purchased at Costco. So that is in agreement, but disappointing in that Tmart advertised 900 Lumens rather than 600 Lm.

The 40W incandescent bulb put out 500 Lumens, as its packaging advertised.

[Billxx]

PeanutButter's Tutorial...

@Peanutbutter

Really appreciate the tutorial video you posted on assembling a SJR 2.0 circuit. It's really easy to follow...

Tutorial for wiring DIY SJR 2.0 "Tuned" and DIY SJRC v1.2 - YouTube

@Lidmotor

Your latest video with your changes to the circuit is really helpful. Really like your current LED to the transistor base...

Super Joule Ringer 2.0----The Amazing Peanutbutter model - YouTube

[Lidmotor]

Blinky Blinky

A question here too -

@Lidmotor, great idea to use Dr. Stiffler's method in the circuit.
I've had to use another transformer when replicating and it may be where my trouble is. The circuit shows the current changing, via the LED, but a 13W CFL or a 2.5W LOA, or a 2W LOA only flash on power connection.
I've tried a bunch of different caps and it seems to be a lack of Base continuity...the transistor fires great, the flashes (lighting of bulbs) being very bright at 12V. But, the bulb goes straight out again. As I say, the LED can stay on and show the current, if the power is left connected, as per your video. Transistor is a TIP117 PNP Darlington, with all appropriate connections reversed. It would seem that there is a ton of something there, which is why there's confusion as to the source of the problem.
Just wondering if you experienced such a thing in your testing ?


Edit. On inadvertently touching the Positive supply to the transistor case and with Emitter not connected, the circuit behaved in the same way, flashing the light ! Trying the same thing again, the LED to show current could be dimmed or brightened in the same way too. I've no idea how that could happen.

When I tried to run CFLs on the 12v transformer circuit all I got was the "Blinky" light situation also. Gutted CFLs didn't do anything at all. I looked at the output voltage this morning and my circuit is only putting out 80 to 90 volts and that explains it. All of my 110v LED bulbs WILL work on this low voltage though. You might be having problems because of the transformer you are using or the transistor. I have a 9v (wall wart) transformer that I will try to see if I can increase the output voltage. I might sacrifice a 2N2222 just to see if it works and at what power input it blows up at.

@Peanutbutter
Keep the explanations coming. The more we all understand what is going on the better the builds will get. I still have the dream of getting a circuit to run frosty cold. They say that Tesla was able to regulate the temperature in his lab with electrical devices somehow.

@PhysicProf
A scientific analysis of what we are doing is sort of a double edged sword. It is great to know exactly what the numbers are and the efficiency but it can be discouraging also. I like the way that you have an open mind about these projects and can apply the science as a (+) tool rather than a (-) one.
To many people just want to "run the numbers", look at the "laws or rules", and never build anything. One of the guys who commented on my last Youtube video tried to build this device on a simulator and could not get it to work. He asked me what was wrong and all I could think of was----his whole approach was wrong. He should just build the darn thing, watch it work, and then analyse it. I know that this sounds like a backwards scientific way but it works for me. Each project "build" adds to the overall knowledge base and the next project gets easier. Sometimes I discover something just by dumb accident. That is when I am having the most fun. You do need a basic understanding of the "laws and rules" to do this. Unfortunately I learned those things in a very boring classroom long long ago and many of them are fogotten. I needed a Dr. Jones.



@ The guys with 220 mains
Peanutbutter should have a simple solution for that.

Lidmotor

[Jules Tresor]

low voltage regulated power supply for tiny SJR

[Jules Tresor]

120 vs 240V

You are right Penno !! It's a main concern for us ... may be 240/24
I wanted to ask PeanutButter if he had an idea of the values of the capacitors for a 240V transformer We ask him so much

Mains electricity by country - Wikipedia, the free encyclopedia I counted 180 countries using 200-240V, 35 countries using 100-120V, and 10 using both.
100-120V is mainly on American continent + 2 in Asia, Japan and Taiwan + very few in Africa.
Yep, the majority of users are on 240V. But I think the gain might be larger with higher voltage, isn't it ?

[penno64]

Hi

10 to 1 ratio for you guys on 110/120 mains

Us poor buggers on 240 - 20 to 1 maybe ?

Penno

[Jules Tresor]

3 kinds of transformer for SJR

[Slider2732]

A question here too -

@Lidmotor, great idea to use Dr. Stiffler's method in the circuit.
I've had to use another transformer when replicating and it may be where my trouble is. The circuit shows the current changing, via the LED, but a 13W CFL or a 2.5W LOA, or a 2W LOA only flash on power connection.
I've tried a bunch of different caps and it seems to be a lack of Base continuity...the transistor fires great, the flashes (lighting of bulbs) being very bright at 12V. But, the bulb goes straight out again. As I say, the LED can stay on and show the current, if the power is left connected, as per your video. Transistor is a TIP117 PNP Darlington, with all appropriate connections reversed. It would seem that there is a ton of something there, which is why there's confusion as to the source of the problem.
Just wondering if you experienced such a thing in your testing ?


Edit. On inadvertently touching the Positive supply to the transistor case and with Emitter not connected, the circuit behaved in the same way, flashing the light ! Trying the same thing again, the LED to show current could be dimmed or brightened in the same way too. I've no idea how that could happen.

[Jules Tresor]

transformer core material ?

Hi guys, can I ask what the cores of the Radioshack transformers are made of ? Laminated steel or Ferrite ? (couldn't find the answer on their website) Thanks, Jules

[totoalas]

a DIY Lightmeter THANK YOU

Your caps did reduced my set up current from 360 mA to a 260 Ma which is needed on Solar Night lighting using 5 watts solar panel and an sla 12 v 7aH battery from 6 pm to 1- am steady brightness 13 v start and end 12.13 v for 8 pcs 5 watts led SIMPLY AMAZING

totoalas