Joule Thief ZEEWEII DSO1511G Scope Analysis
April 07, 2023
Blog
In the GreatScott! video seen in my last Joule Thief post, he notes that this circuit is difficult to understand. While a full understanding may indeed still elude me, here I’ll use the $60 ZEEWEII DSO1511G scope and signal generator to get a hands-on view of each of my Thief setups.
Experimental Setup
Input power for each setup is via a 1.5V AAA battery, and measurements are taken across the positive LED input and ground.
Wrapped Toroid
As one might suspect, it appears that the 2mH wrapped toroid Joule Thief configuration works quite well. This device used with a 1kOhm resistor generates the following LED power inputs:
- 12.6kHz, max 2.5V, min 59.1mV
Perhaps we can improve this design, or change its performance? Substituting in a 1.8kOhm resistor changes the situation, but perhaps not as much as you would suspect:
- 13.7kHz, max 2.5V, min 98mV
.jpg)
Caption: 10Ohm resistor added
Image Credit: Jeremy Cook
What about adding a resistor on the other leg? A 1k resistor in series with the base coil, plus a 10ohm resistor in series with the collector/LED coil resulted in:
- 12.3kHz input, max 2.24V, min 38.4mV.
.bmp)
.bmp)
Caption: 1kOhm resistor only, compared to adding a 10Ohm reisistor on the base coil
Image Credit: Jeremy Cook
The extra-resistor waveform is skinnier, and provides a lower duty cycle. Given this, perhaps this extra component could be the key to produce an even longer-lasting Joule Thief, though how this performs during different battery stages is an open question.
Axial Inductor Results
Another view, this time with a scope: pic.twitter.com/VZIPvrqilw
— Jeremy Cook 🤖 (@JeremySCook) March 20, 2023
Using two axial inductors yielded much different results. Two side-by-side 220uH axial inductors were used in the video shown above, with a 1.8kOhm resistor, resulting in the following when side-by-side:
- 100kHz input range, max 2.02V, min 256mV
Besides a much higher frequency, what’s interesting is that the gap between the top and bottom voltage readings is smaller. Light is inconsistent with a 10Ohm resistor on the collector/LED coil. The duty cycle, however, is in the 31% range, which could present some advantages.
As the inductors are separated, the frequency goes up, the maximum voltage goes down, and the minimum voltage goes up until the LED no longer lights. In fact, it appears to go lower than the nominally 1.5V battery input with a wide enough gap.
PCB Inductor Results
.JPG)
Caption: 1.8kOhm resistor and PCB coil
Image Credit: Jeremy Cook
Results here were inconsistent with this type of inductor. As pictured with a 1.8kOhm resistor, results were as follows:
- 753kHz frequency, max 2.12V, minimum 256mV
Taking all this data together, it would seem that the lower the inductor value, the higher the frequency, and the narrower the gap between the maximum and minimum voltages. How strongly each inductor coil is magnetically coupled with the other also appears to make a huge difference.
Also notable here is that experimental numbers fluctuated greatly during the same and repeated experiments. The position of components and even one’s hand appears to affect performance.
ZEEWEII DSO1511G Performance
For this experiment, the DSO1511G performed admirably, as the frequency tested and voltage range was well within its parameters. The reference wave function was especially helpful to compare between parameter changes. Being able to store images and transfer them to the computer is nice, and they include on-screen information like frequency and max/min voltages.
.JPG)
Caption: Red reference wave with 1kOhm resistor, yellow live view adds 10Ohm resistor to collector coil
Image Credit: Jeremy Cook
That being said, it doesn’t save the reference wave in the stored images, which would have been helpful here. It would also be nice to be able to label your images, as analyzing even a handful at a time (such as for this article) quickly gets confusing.
Joule Thief: Not Free Energy, but a Fascinating Contraption
The Joule Thief has been known–and certainly misunderstood–for many years. Based on this experimentation, it appears that using a small resistor on the collector/LED inductor side could potentially help trim power usage. It also seems that the larger the inductor’s Henry rating, and the better the magnetic coupling, the lower the frequency and the greater the voltage swing.
Beyond this limited analysis, one could also examine different inductor values, as well as the transistor type, and LED choice for optimization. In fact, exploring this circuit fully could be the subject of an entire research paper, and you will find more than one with a quick websearch. That being said, using the ZEEWEII DSO1511G scope did give me a better understanding of how the Joule Thief works, and it could be an invaluable tool for other such analysis.
