Here is the video with the waveform and all and I am really impressed by this little bugger.
www.youtube.com/watch?v=FbwzAmO0998
I was very busy today with my car but I did manage to squeeze in about an hour to reverse engineer the circuit. Rather odd at one point as it makes no sense to me so I will run it up in the simulator and see what I get in a day or so.
Your scope is showing a frequency of about 1.7 MHz.
I've taken apart a few garden lights myself. There are two main variants that I've found. One uses a photoresistor CdS as the ambient light sensor, to switch from charging in daylight to lighting the LED at night. The other simply uses the photovoltaic cell itself, both for sensing and charging.
The first type uses a tiny chip marked ZE002 and the second type uses ANA608. These are both four-terminal devices. Here are the schematics of the two types I've analyzed:
This is neither of those and has no IC on it (probably why it is better). I already can tell it has a darkness sensor via the solar panel since I reverse engineered it I just haven't laid it out yet. Works extremely well and I will take my DFC-1000 out to be 100% sure what the frequency is but I can't remember if it is a high impedance load or not. I would hope it would be.
Oh, here is the parts count: 11...1 Diode, 3 Resistors, 1 cap, 2 transistors, 2 inductors, 1 Solar Panel, and 1 white led that I have no idea what the specs are for it.
That's a lot of parts. What are the transistors? I would like to see the schematic when you have a chance. The DFC-1000 counter has a 1Meg input impedance for low frequency signals (Below 100 MHz), the JT won't even know it's there probably.
ETA: It appears to be the same as the Elenco F-1000:
http://electronicsandbooks.com/eab1/manual/Hardware/D/Dynatek/DFC1000%20c20060821%20%5B9%5D.pdf
Yes, it had 3 models that rebadged it and I forgot what the other one was.
S9013 and if I were to design this circuit for my needs I would use the S8055 over the S9013 since they fight for price supremacy a bunch.
Quote from: TinselKoala on October 15, 2013, 09:47:45 AM
Your scope is showing a frequency of about 1.7 MHz.
1712.5Khz using the 0.01s gate (only 1 digit past the decimal point doing this).
btw, I looked in the box and my original receipt from JDR was in it. 1/28/1991 (my 26th birthday) and it cost me $199.95 from them. So, my 50 to 80 dollars was way off (must have been something else I purchased but never got back from the flood/fire of 2006).
Data sheets:
We crossed posted it seems.
btw, I looked in the box and my original receipt from JDR was in it. 1/28/1991 (my 26th birthday) and it cost me $199.95 from them. So, my 50 to 80 dollars was way off (must have been something else I purchased but never got back from the flood/fire of 2006).
The S8055 has a higher amp rating and from what I saw, when I looked yesterday, it had better gain too.
Quote from: Legalizeshemp420 on October 15, 2013, 10:46:33 AM
1712.5Khz using the 0.01s gate (only 1 digit past the decimal point doing this).
Yep. I counted just about 6 full cycles in 3.5 microseconds (7 full divisions). 6/0.0000035 = 1714285.714285... and respecting sig digs that's 1.7 MHz. Pretty good for eyeballing a YT screenshot of an analog scope, eh?
Quote from: TinselKoala on October 15, 2013, 10:56:13 AM
Yep. I counted just about 6 full cycles in 3.5 microseconds (7 full divisions). 6/0.0000035 = 1714285.714285... and respecting sig digs that's 1.7 MHz. Pretty good for eyeballing a YT screenshot of an analog scope, eh?
Yes, I am amazed to be honest.
You are EXACTLY the type of viewer I love having on my channel. They can see and set forth because frankly I try to explain as best as I can but I have always only given most of the story to something and, just like a teacher, expect you to do just a little leg work. That is in real life though and on my YT vids I sometimes forget to give all of the answers and besides I can't know what some guy in Zimbabwe is wanting out of the video, hehe, though I do try and give as much info as I know.
I love all of my viewers and just wish I had more to be honest but I don't know how to get them as I have never been a salesman type.
Well, I just finished with this thing in Multisim and it does not have 2 components I need to simulate this 100%
Ffirst it doesn't have any solar cell models of any kind. So, I just replaced it with a SPST switch to simulate what the Solar Panel is doing in the circuit. POOF, I did that and it worked when you close it (this is exactly what a dark solar cell/panel would act like in this circuit).
Now, not having a model for the S9013 (I suck at model making because I change what is needed to the tee and it will be wrong no matter how many times I go over it) and using the BC337 in its place works but I know from the real world that will change brightness, frequency, etc... and it did. I simulate this at 1845 Khz and my LED is not lighting up because it is simulating 2.5v from 1.2.
So, I know it works in the simulator at least but I am not done with this yet.
Schematic?
I posted these photos on the forum back in 2010 and I never could figure out what type of JT circuit my garden lights were using. 2 resistors on one side of a very small circuit board and a gray blob on the reverse side. I was told by someone here that the blob covered the chip. Someone else said it covered the transistor but, only 2 wires run to this blob and the blob is just about the diameter of a pencil erraser and maybe .070" high.
I had about 10 of this type and have used the circuit for making many other things but had no idea how it worked as a JT with no transistor, coil or chip. (as far as I could see anyway) This circuit would drive a nice 5mm ultrabright LED very well. (The lights come with a 3mm led)
Bill
PS I even made a device using this circuit and a cd driving 2 red leds spaced apart like eyes to scare away predators from my friend's chicken coop. Hangs from a tree, charges by day runs all night.
I see 1 coil, 1 resistor and 1 IC (under the blob) in your pictures.
The black blob covers the chip.
There are more than two connections to the chip, I believe, two traces visible on the foil side connecting components and two solder blobs connecting wires.
The green resistor-like thing on the component side is an inductor.
I think this is essentially the same circuit as the one posted above with the ANA608 chip, which is tiny, barely room for the four leads. A smd version would easily fit under the blob.
Quote from: TinselKoala on October 15, 2013, 03:49:38 PM
Schematic?
I got sick yesterday so finishing up now but still not 100%. I am sad but DALM circuit is no more as I reimaged my drive and that is what I lost. :( All I have left is the video on YT but should be easy to reproduce if I need to.
Here we go:
D1 - IN5819
R1 - 20K 5%
R2 - 100K 5%
R3 - 5.6K 5%
C1 50pf 10%
Q1/Q2 S9013
L1/L2 150uH 10%
1n5819(g) is a common Schottky rectifier diode.
And... here's the data sheet for it. Up there, I mean.
Quote from: TinselKoala on October 16, 2013, 03:58:20 PM
And... here's the data sheet for it. Up there, I mean.
Yep, I knew it was Schottky barrier diode but all parts listed in the parts list are what is on this board (Q1/Q2 were changed in the simulator because S9013 is not available).
Quote from: TinselKoala on October 16, 2013, 10:26:11 AM
The black blob covers the chip.
There are more than two connections to the chip, I believe, two traces visible on the foil side connecting components and two solder blobs connecting wires.
The green resistor-like thing on the component side is an inductor.
I think this is essentially the same circuit as the one posted above with the ANA608 chip, which is tiny, barely room for the four leads. A smd version would easily fit under the blob.
Thanks TK. I had no idea there was an IC that would run on 1/2 volt, which is about what these run down to before the lights go out. If this chip will do this I should make a JT with one of my own coils. I have stayed away from using chips because most of them need like 5 volts to run. (555's, etc) Why did they enclose the chip in that blob?
Bill
***EDIT***
LH Thank you for pointing out the same thing. I had no idea that green resistor looking thing was an inductor. I have several of those in my resistor collection box from electronics that I have salvaged parts from. I guess I need to place them in another bin.
Over the last 30 years whenever I opened up a China, or Japanese, cheap electronics item I almost always found that blob. Back in the day (circa early to mid 80's) if it had a piezo speaker it always had that blob. That blob always ended my exploration because there is nothing we can do with it. :(
What is ironic is that these chips can be ran on such a low voltage. Look at my video where I show a square wave that makes the solar powered figures dance. That is no battery and at less than half a volt. Only discrete part on it is a 10v 470uf 105c Electrolytic capacitor. The black blob has 4 connections. 2 for the solar panel (.5v) that has the cap in parallel with it and two going to the pancake coil. It makes a wonderful square wave pulse that the time, and amplitude, is dependent upon the voltage it receives but it is always a clean square wave.
Yep, the chip is a little low-power square-wave oscillator.
Which gives me an idea. A friend of mine gave me a bunch of these:
http://www.surplus-electronics-sales.com/index.php?main_page=product_info&cPath=38&products_id=589 (http://www.surplus-electronics-sales.com/index.php?main_page=product_info&cPath=38&products_id=589)
They need a 5v supply but make a very nice accurate and clean square wave output.
Quote from: TinselKoala on October 16, 2013, 11:04:58 PM
Yep, the chip is a little low-power square-wave oscillator.
Which gives me an idea. A friend of mine gave me a bunch of these:
http://www.surplus-electronics-sales.com/index.php?main_page=product_info&cPath=38&products_id=589
They need a 5v supply but make a very nice accurate and clean square wave output.
Isn't that a sine wave oscillator?
Quote from: Legalizeshemp420 on October 16, 2013, 11:12:03 PM
Isn't that a sine wave oscillator?
That one may be, sorry. I have the MTO version which is definitely a square wave. The distortions you see are caused by my casual hookup.
Hmmm, that is kinda nice even with the distortion. Maybe use it in a pulse generator?
Where did you find the documentation for your version at?
Here's the closest thing to a "data sheet" I could find.
btw, I just did a freq measurement on that circuit I showed for the sq. wave osc and it had a period of 180.162 ms. That comes out to be about 5.5hz using my flashlight. In direct sun I have no idea what this would be at.
A crystal oscillator like TK is showing was the standard workhorse for clocking computer boards and other boards for years. Now some of those parts literally look like specks of pepper.
Anybody hear of Fox?
http://www.foxonline.com/
Just name your frequency and your package.
Quote from: MileHigh on October 17, 2013, 01:23:59 AM
A crystal oscillator like TK is showing was the standard workhorse for clocking computer boards and other boards for years. Now some of those parts literally look like specks of pepper.
Anybody hear of Fox?
http://www.foxonline.com/
Just name your frequency and your package.
First thing I hit was Oscillators - Through Hole and most of them have now been discontinued. :(
Check Ebay, these things are all over the surplus market, in many many frequency choices.
I also have about 20 of these little amplifiers:
MC1590G
Damn, an all in one small package amplifier and AGC in one. What is the noise on that?
I dunno, the most I've done is to breadboard one up and confirm that I can put a signal in and get a signal out. I can't even get near their frequency limit though, especially without putting it on a good PCB. At present I have no real use for them. They are brand new, still on their little plastic carrier things to keep the pins straight. The friend who gave them to me was a technician at a local company that made MRI machines but went out of business years ago, and these are new old stock from there.
Are solar charging lights really a Joule thief? Or are they just good efficient solar battery chargers and LED circuit? I guess you could call them a JT if you want but I always thought of a JT as having a bifilar wound toroid or at least a bifilar coil and I doubt these have bifilar coils in them. Just a regular inductor. But it's nice that they are getting so cheap thanks to mass production (and of course China).
Quote from: e2matrix on October 17, 2013, 07:08:45 PM
Are solar charging lights really a Joule thief? Or are they just good efficient solar battery chargers and LED circuit? I guess you could call them a JT if you want but I always thought of a JT as having a bifilar wound toroid or at least a bifilar coil and I doubt these have bifilar coils in them. Just a regular inductor. But it's nice that they are getting so cheap thanks to mass production (and of course China).
A Joule Thief is just a cutesy name for a blocking oscillator circuit anyway and if you look up some schematics for blocking oscillators most hardly use a bifilar coil if at all. JT (Blocking Oscillators) have been around for longer than you are old but only in 1999 did some fool call one a Joule Thief. I hate the term but it is what the modern people know it as so stuck using it.
Quote from: Legalizeshemp420 on October 17, 2013, 07:31:51 PM
A Joule Thief is just a cutesy name for a blocking oscillator circuit anyway and if you look up some schematics for blocking oscillators most hardly use a bifilar coil if at all. JT (Blocking Oscillators) have been around for longer than you are old but only in 1999 did some fool call one a Joule Thief. I hate the term but it is what the modern people know it as so stuck using it.
That's my understanding as well. A fellow in England named Big Clive coined this term and it has been used on youtube and elsewhere online ever since. A little further research shows the circuit was developed by a Russian in the 60's. (I forget his name)
I "thought" I knew a lot about these circuits and then I come to this topic and learn of the existence of radial inductors and a tiny blob covered chip. This means that I can take one of those chips, a resistor and one of those inductors and have a JT circuit that will have a footprint the size of a dime. Now I am wondering why I wound all of those 1" toroids all those years ago!!!! Of course, this is for a low voltage JT. I am still working on my circuit board high voltage JT. (Similar to the modified flash circuits)
Can you put enough of those radial inductors in series (or parallel?) to output 400 volts from 1.5 volts in a blocking oscillator step up transformer (JT) type device? That would be really cool.
Bill
You can easily step up the voltage but the current is another story.
All of the led bulbs and modules that I use in my lights respond very well to high freq./high voltage and very low current. That is why I stick to 1.5 volts or "dead" batteries for the input. I light my apartment for free using the dead batteries given to me by my friends. 350-600 volts seems to work well. About 200 mA draw usually.
I feel so stupid. I had no idea those type of chips were available. Oh well, I love to learn.
Bill
With a JT you can easily get over 1500 volts but I will be damned if I know how to even measure that as everything I have would die (I lost my first DMM due to negative ion generators I used to make).
You can make an indirect measurement for the peak voltage. You swap the LED for a low value resistor, say 20 ohms, and look at the peak voltage across the resistor. That tells you the initial current. Once you know the initial current then you implicitly know the peak voltage for any resistive load. What ultimately puts a limit on how high the voltage can go is the parasitic capacitance associated with any high-resistance load and the amount of energy the inductor has available to discharge into the parasitic capacitance.
In plain English, if you see a 7 volt peak when you use a 20-ohm resistor as a load, then if you use a 500-ohm resistor, you will scope a 175-volt peak that is narrower. You know what the peak voltage is expected to be before you scope it.
MileHigh
Hey Bill, I just posted a video of a quick test of the "Pirate Light"!
http://www.youtube.com/watch?v=f81ZEB_zGVA
8) (Need shades, the light is so bright!)
Very cool. I see you got bit by the circuit already, ha ha. I usually cover the HV outputs with hot glue for insulation but I thought you might want to solder something different someday so I left them bare. I still get zapped by these even though I am mostly careful, ha ha.
Bill