Question on toroids and Joule Thief

Status
Not open for further replies.
You can buy a solar garden light for $1.00 at some places. It has a solar panel that charges a 1.2V battery and uses a little joule thief IC with an inductor that looks like a 1/4W resistor to boost the voltage for the LED. The cost is so low that the poor quality battery is free. I use many solar garden light circuits to drive colors-changing LEDs.
 

Attachments

  • solar garden light.png
    67.3 KB · Views: 1,041

I have some of those QX5252F IC's on order and will be here next week I assume. I also have some inductors coming in the resistor looking package (axial) but was not sure on the specs so I just ordered some 220uH and 330uH versions. Any idea what the inductor spec is on those solar garden lights? And give me the link where you buy them for #1.00 please.
 
You ordered inductors that are not suitable because the datasheet for the QX5252F lists the peak current of the high frequency pulses. The average current is what you see that is much dimmer than the momentary peak current of the pulses. Most of my solar garden lights were bought at Walmart and at Giant Tiger and use 82uH. Dim ones used 100uH and 150uH. My brightest ones use 47uH.

The high inductance ones have a high resistance that also reduces the brightness.
 

Attachments

  • solar garden light IC specs.png
    12.2 KB · Views: 357
Ok, great! Thanks for the info about the inductors. I will order some more with the more suitable specs. I guess I should have looked at the spec sheet but I believe I ordered the inductors for use without the QX5252F but I was still guessing. I will have to check Walmart and see if they have a sale here and there.
Thanks

p.s. Another question. On the SMD/SMT type inductor that are a round coil bobbin shape, for the use we are discussing here, must they be "unshielded" to work?
 
Last edited:
The high value inductors also have a high resistance that further dims the LEDs.
It is winter now and solar garden lights work poorly when covered with snow and they are impossible to insert into the frozen earth. So now they sell snow shovels instead.
The Covid virus and antivaxers extending its duration are causing the cost of everything to be higher.
I have never made an inductor for a joule thief.
 
So you don't know if shielded vs unshielded would be better? I am guessing unshielded for some reason.

Ok, fair enough. I found this site to be interesting and just downloaded the free software. I checked it for viruses etc and it is clean.

 
Last edited:

I think inductor winding is a pretty simple matter. Here is some general information about winging toroidal cores:
http://cappels.org/dproj/Toroid_winding/Toroid_Winding.html

If you have to wind a lot of wire, use a bobbin:

You can probably make one of these out of a scrap of plastic if you need one.

I find that high permeability cores work well, mainly because they give a lot of inductance with relatively few turns of wire. The high inductance is helpful in that the switching losses of the transistor are usually lower with the higher inductance because the oscillations are at a lower frequency. These things can work into the tens of MHz but efficiency is pretty poor and besides, you might cause radio frequency interference.

By the way, the Joule Thief circuits that use two windings in series and the ones that use a tapped inductor are pretty much the same, but the tapped inductor is easier to wind because there is only one winding needed.

A fellow named Wolfgang Driehaus pointed out that there are ferrite cores to be had from dead compact florescent lamps. Just be careful to not break the glass envelope while carefully dissecting the lamp.
 
Thanks Dick. You have an interesting selection of projects on your website. I have a 3D printer so I could make some cores or bobbins from plastic if need be. But I have purchased some cores that I think will work well. At least the first ones I bought based on this site's authors findings. https://tinyurl.com/m79wwva8 I don't have any issues winding a coil but more in wire size selection and numbers of turns. You know, all the complicated stuff. Do you know if the shielded versions of a toroid like the surface mount variety work as well as the unshielded ones? Just curious as sometime I see them cheaper or available in the US on ebay. I assume the shielding just keeps the magnetic field from messing with nearby components on a pcb. But otherwise do they behave the same as a regular toroid in a Joule Thief circuit?
 
Last edited:

When you wrote "shielded" were you thinking about this style (above)?

I've used this style as the inductor in buck converters, not because they are shielded but because they were the cheapest chokes my distributor had with the value I wanted.

There are applications in which you don't want a transformer or choke to induce currents in other circuits, such as in audio amplifiers (I imagine that is why you find large toroids in audio equipment power supplies. audioguru can tell you more about this.

They should work the same as unshielded chokes in Joule Thief circuits.

That's a nice webpage that you dug up, the fellow went out of his way to investigate several aspects of Joule Thief design.
 
Yes, those are the ones I was referring to. And I have run into the shielded ones cheaper than the unshielded one on ebay so that is why I figured I would ask. Win win situation. That website I found is indeed full of a lot of work and testing. I am glad I found it as it saved me a bunch of time with toroid choice for one.
Thanks
 
I bought a bargain pack of 30 turn choke coils to build a Joule Thief. 6 turns of wire makes it work perfect with a Blue color LED. Later I added the LDR circuit so it only comes ON when it gets dark outside. Notice DOTs on circuit drawing for polarity. Don't worry if you wire it wrong 1.6v will not smoke the LED or the transistor.









 
Last edited:
A Joule Thief is supposed to use high quality inductors so that it has high efficiency which uses a low battery current. A "bargain" choke with few turns of wire will probably be very inefficient which uses a very high battery current.
 
Hi Gary,
What LDR did you use? I have some GM5539 but that is all I have. I never figured out the best way to choose those so I bought one in the middle range. Otherwise I was going to also mention that the toroid looks like a not so efficient design but I am sure it works just fine otherwise. The wire gauge looks too large for one. What AWG wire is that anyhow?Thanks for posting the circuit.
VS
 

Look at the 2nd circuit drawing it shows LDR resistance. Someone in another thread mentioned if you have 30 LDRs they will all test different resistance will be all over the place and that is correct. R = 150 to 500 ohms in sunlight. R = 200K to 400K in the dark. R2 resistor is light sensitivity adjustment when sun is low & it is getting dark. No part numbers on my LDRs. Set R2 so LED does not come on until very dark cheapo dollar battery will last 6 months. NOTICE the date on dollar store batteries sometimes the new packages are dated 2 years old those OLD batteries are junk they go dead very quick. BLUE color LED is higher voltage than RED LED change the 6T coil 1 turn to lower voltage for RED LED.

I bought 50 eneloop rechargeable AA batteries on Ebay they work good and cost less than any other battery in the long run.
 
Last edited:
Ok thanks for the added info Gary. I can figure the rest from there. I may have to buy some different LDR's or do some other tricks to make mine work. Good thing they are cheap. I too love Eneloop batteries and have a bunch of the rechargable AA sized 1.2V ones and a couple of chargers from Amazon.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…