RustyBolt.Info/wordpress

On Youtube, this guy named Xee2 or Xee2vids posts short videos of schematics, projects and circuits.  There is no motion, just slides, and no narration, no voice in his videos.  There is no way to comment to tell him if I’ve found a mistake.  But so far, most of his vids are very good with very few errors.  And most of the stuff is original and innovative.  One project caught my eye, a blinking Joule Thief.  The video gets to 35 seconds then freezes and won’t go any further, so I don’t know what the rest of it is about.  But I got the schematic and now I’ve built on to see how it works.

It’s a simple circuit, with all the parts of a conventional JT and adds a 10uF capacitor to get the circuit to flash.  I built it on a small perfboard but I’ve been having difficulties with it.  I put a 68k resistor instead of the 100k resistor called for in the schematic.  When I finished and powered it up, I got irregular flashes at about 2 per second.  But I couldn’t find the irregularity in the flash rate.  The schematic in his vid does not show a bypass capacitor across the 1.5V battery, so I installed a 10uF bypass capacitor.  While I was doing that I removed the 68k and put in a 120k which I figure should be closer to 1 a second.  When I finished and fired it up, nothing.  Voltages look normal and I didn’t change anything since the first time it flashed.  I’m puzzled why it’s not working.  Maybe the heat from unsoldering then soldering it back has damaged something.

Try It Again  I decided that it wasn’t worth troubleshooting the problem and built a second one, same as the first.  I’ll go over a few things.  These changes apply to both.   I changed from the MPSA06 to a PN2222A, which is not that different.  The original resistor was 100k, I changed to 120k, mainly because the flash rate was too high.  Even with 120k the flash rate is still fast, about 2 per second.  A better choice might be 180k or more. The capacitor was 10 uF, just like the original.

The Coil  The original used a “1 inch core from Electronic Goldmine, 5 for $1.00”, which has been sold out for some time, and will no longer be available, since they sell mostly surplus and it’s extremely doubtful they will ever be able to find them surplus again.  I still have some of these, so I wound 20 turns of telephone wire onto one core and measured the L and got 4.52 millihenrys.  But this core is too large for a low power Joule Thief.  Instead, I used a tiny 1/4 inch core from surplussales.com, which was about the same price.   It is the (ICH) T231212T, available for $.25 apiece.  As stated in the ad, they are high permeability, 5000 or more, so it doesn’t take many turns of wire to make a reasonable amount of L (inductance).  I used 7 inches each of 28 and 30 AWG enameled wire, wound bifilar onto the core.  The inductance was about 480 microhenrys, which is just fine for a Joule Thief.  I used a yellow LED,  mainly because a flashing white LED is totally boring – white LEDs are made for illumination – and I have a dozen Blue Blinky circuits madly flashing so I really don’t need another blue one.

The second one went together quicker than the first, mainly because I had the original to use as an example.  The perfboard was only 6 holes by 6 holes, where the hole spacing was 0.1 inch.  Everything fit nicely.

Test It Out  I connected it to the power supply and it flashed, about ten flashes every 7 seconds.  Still too fast.  But what was disappointing was the light output.  I could barely see the yellow flashes on the palm of my hand in a fully lit room.  The light output is so dim I can look directly into the LED chip and clearly see it flashing.  At this low current, the prediction he made about the AA cell should last for 8 years is very likely true.

But with the 120k and 10 uF, the flash rate was more than twice a second.  I removed the 10 uF and put in a 33 uF, and it now flashed considerably slower, but still flashed at 1.4 flashes per second.  The capacitor needs to be at least 47 uF to get down to close to 1 flash a second.  In order to increase the light output. the capacitor must be much larger – he suggests 10k and 100 uF, but I think 100 uF is still too low.  This brings up an important point.  This resistor is the only source of base bias current for the transistor.  The normal JT uses 1k, but here we have 100k, so it stands to reason that the base bias current is much less than a normal JT.  This looks to me like the main reason the LED is so dim.  The solution to more base current is a lower value resistor (the reason why he suggested using a 10k in the video).  But the more base current, the faster the capacitor charges.  So we have to increase the value of the capacitor to keep the flashes at about 1 second.

Conclusion  I don’t see a future for this circuit, due to the low light output and need for a large, expensive capacitor.  My Supercharged JT Flasher uses a 4.7uF cap, my Blue Blinky uses a 1 uF, and some other flasher circuits use similar sized capacitors.  Also, we are back to a circuit that uses a single transistor, so the LED output, the frequency and length of the flash are all dependent on a single transistor and a few other parts; changing one may change the circuit in unexpected ways.  Right now, I like the Blue Blinky circuit because it puts out bright but very short pulses, and does a respectable job of making the AA cell last a long time (2 years).

Back to experimenting…

I came across this document on scribd, and found it very interesting.  The guy’s name is Fernando Garcia and it is dated May 2009. Unfortunately there is no contact information, for there are several unanswered questions I have that he could answer.

He discusses how he measured the efficiency of a Joule Thief and how he measured the actual light output of the LED without using a luxmeter.  Then he compares the first to another JT.  I especially liked the graph he made of the LED current versus the CdS photocell resistance.  It shows that the CdS  resistance is not linear in response to the current to the LED.  But it really isn’t that important because once he finds the CdS resistance, he only needs to refer to the graph to find the current.

The important point about using this photocell is that it will remove the possibility that the pulses through the LED are causing an error in the measurement.  In other words, once the photocell is graphed with the LED current from a DC power supply, the pulsating current causing light pulses from the LED will be filtered out by the photocell, presumably because the photocell cannot respond to the pulses, which is basically the same way our eye responds.  Then if you get a photocell resistance of 100 ohms for example when it’s lit by the Joule Thief, the JT should be putting out the same amount of light as when it was lit with DC and the photocell resistance was 100 ohms.

Measuring the input current   He discusses how he went about measuring the power supply current at 1.5 volts.  He inserted a current sensing resistor in series with the positive supply line, and measured the voltage across it and calculated the current.  But he went to elaborate lengths to filter out the “high frequency ripple” pulses in the supply line caused by the Joule Thief’s switching on and off.  He used a low ESR 2.2 uF tantalum capacitor, two more capacitors and a 2.2 uH choke to remove the pulses.  I highly recommend using a bypass capacitor across the supply leads.  But instead of filtering out the pulses in the supply, I would simply put a low pass filter on the voltmeter’s input: a 1k or 10k resistor in series with the positive voltmeter lead, and a 1 uF capacitor across the voltmeter leads.  The voltage drop across the 1k or 10k is almost zero because the typical digital multimeter has a very high input resistance.  Also, the great thing about doing it this way is that you can move the filter with the  meter leads to another part of the circuit and have the same benefits of the filter.

The Joule Thief   He shows the circuit and admits, I’m happy to see, that the 2N3904 he used was not a good choice of transistor.  I think that this choice may have influenced his measurements in a detrimental way.

He goes on to explain how he built the Joule Thief.  He explains how he made the coil, which he mistakenly called a transformer; I quote:

The transformer was bifilar-wound on a Micrometal’s T30-26 core. Fully wound it gave about 8.5 µH, which is also a little on the lowside.

The T30 toroid core is only 0.3 inches or 7.6 mm outside diameter, which is a very small core. The 26 type ferrite material is low permeability, so the inductance is very low.  He said fully wound, but what size wire did he use?  He also didn’t say how he wound the two windings.  He said they were bifilar wound.  Were they both the same size wire?

He goes on to say that the inductance was 8.5 microhenrys.  But I would say that that is a lot more than just a little on the low side.  In my experimentation with dozens of JTs, I would say that the coil should be 30 microhenrys minimum with an optimum point between 100 and 300 microhenrys.  A lot has to do with the size of the core and the core material, and the size of the wire.  With the tiny core that he used and fine wire, he might be able to get 30 uH, but the wire will have to be very thin.  And the wire will have higher resistance, which will cause more losses and lower efficiency.

I think he should have chosen a higher permeability core, or else larger to allow thicker wire.  He later goes on to use a larger core, a T50-8 core.  But then he “adjusts” the turns to give the much too low 8.5 microhenrys.  Apparently he thought that was necessary to make the comparisons.  I’m fairly certain that the result of using the low 8.5 uH inductance is that his Joule Thief ran at a frequency that was very much higher than a typical JT.  A typical JT runs at frequencies between a few tens of kHz to a few hundred kHz, with a broad optimum point around a hundred kHz.  He gave no information about what frequency at which his JT was running.  I think it was running at as much as a megahertz, or maybe even more.  I’ll explain below why this is important.

Further Measurements  He then adds a 1N5817 Schottky diode and 1 uF filter capacitor to the output between the transistor and the LED, to light the LED with DC without the pulses.  In his comparison, he showed that adding this diode and cap made a large increase in the “figure of merit”.  Later he says that adding the diode and cap is a no brainer, and to always include it (see note at the bottom).

He didn’t say anything about what kind of LED he used.  The schematic says “Wht LED” so I presume that that is what he used.  I have my own theory as to why the diode and cap increase the figure of merit.  This tends to get involved but it is important, so I hope you can bear with me.

All semiconductors have a junction capacitance, because the junction (or junctions) acts like the plates of a capacitor.  In a transistor this is called the Miller capacitance.  In a LED, there is a small amount of capacitance, which, I will have to try to find out what it is for a typical white LED.  Also, diodes have a recovery time during the reversal of polarity that is when the hole and electrons are swept out of the junction.  In a Schottky diode, the junction is half a diode, because half the junction is metal, not semiconductor, therefore it has faster recovery time. The 1N4004 rectifier is a “standard recovery” diode, because it operates at power line frequencies and has very little losses during the recovery time.  If we operate a 1N4004 at tens of thousands of kilohertz, the losses are much greater, so these are not used at the frequencies found in switching power supplies.  Instead the fast recovery and ultrafast recovery diodes are used, and also Schottky diodes.  These have very low losses at a hundred kHz or more.

The LED is also a diode, and will have a similar recovery time.  If this is slow, it will have an effect on the efficiency at high frequencies.  The Schottky diode has a very fast recovery and this may be why the Schottky rectifier and filter gives better performance than the LED alone.  One thing is certain: there is a half volt drop across the Schottky as it rectifies, and this has to be accounted for as far as loss goes.  Assuming 30 milliamps peak current (just a guess) and a half volt drop, that is 15 milliwatts of loss that ends up wasted as heat in the Schottky diode.  This is why I chose to eliminate the LED current from being rectified in my Supercharged Joule Thief.

Back to the document.  Since the waveform at the LED is complex, I do not measure the LED voltage.  I assume that the LED voltage is 3.3V for calculating the power to the LED.

In the formula he gave, he did not say current, which is confusing.

At the beginning of the document the author stated that power out divided by power in gives efficiency.  Then he uses the figure of merit, which looks to me like he is saying it is the ratio of the (equivalent) LED currents (his formula doesn’t say it’s current) between the original circuit and the modified circuit.  In making the graph the author did not measure the voltage across the LED for each point on the graph.  Therefore the power to the LED is not known.  With LED current but without the power, the efficiency cannot be calculated.

One could assume a constant voltage of 3.3V across the LED, which is how I calculate the LED power.  I have attempted to read the LED voltage with a DMM, but I get the battery voltage, not the LED voltage.  I have also read the voltage across the LED with an oscilloscope, which gives peak and RMS readings. The RMS reading applies only to sine waves, but the JT does hot have a pure sine wave, so RMS readings are not valid.  I do not measure the LED voltage since the waveform at the LED is complex, and may cause the meter to give a false reading .  I assume that the LED forward voltage is 3.3V for calculating the power to the LED.  The LED forward voltage stays relatively constant as the current changes so this assumption is a reasonable one.

Difference of opinion  I must state that I do not agree with the conclusions the author gave in the “Comments On The Results”.  The reason I do not agree is that the coils (he mistakenly calls them transformers) in the JT he used have much lower inductance than a typical JT.   In his 3rd comment he says (I quote) “for maximum efficiency it pays to use magnetic cores optimized for switchmode supply usage”.  Ironically, the small micrometals cores he used are for radio frequency use, not for switching power supply use.  He said the coil was 8.5 microhenrys; the typical JT may be 100 or more microhenrys.  Therefore the frequency of his JTs is much, much higher than a typical JT, and it is my opinion that this causes the JT’s operation to be less than optimum and the results that he found were not typical of a conventional JT that the average experimenter might build.   I think the author should change the ferrite core to a higher permeability core so that the coils measure at least 100 microhenrys, and then make the measurements with these coils.

The author stated that adding the capacitor across the base resistor was beneficial.  My opinion is that as I said above, his Joule Thief was running at a frequency much, much higher than a typical JT.  Therefore adding the capacitor to speed up the transistor’s switching was much more noticeable in his results because the transistor was running at too high a frequency, which caused the transistor’s switching times to be more influential on the circuit.  I think if he was using the switching type of cores and the frequencies were much lower, the capacitor’s effect would be much less, and the improvement may not be worth adding the capacitor.  In my own experiments, I have found that when I add the capacitor (depending on its value), it sometimes helps a small amount, or sometimes it reduces the LED light output. I have used values from a few tens of picofarads to over 1000 picofarads, and I don’t find that it is of much help.

Update 12-24 at 7 PM:  I wound a low permeability core with 12 turns of 24 AWG wire and measured the inductance at 8.9 microhenrys.  I connected this up to a 2N3904 (same as he used), a 1k resistor and a white LED.  I measured the frequency and got 780kHz, which is ten times higher than a typical JT and is much too high, it’s in the middle of the AM broadcast band.  The supply current at 1.5 volts was about 45 milliamps, which is low for a JT but typical of a 2N3904.

One other factor that I’ve thought about but found very little info on is the response of the phosphor used in white LEDs.  The white LED is actually a blue LED but the chip is covered with the phosphor that converts part of the light from blue to the other colors, which then looks like white light.  The blue LED can respond very quickly to the electric pulses, putting out similar pulses of blue light.  But what happens when the blue light hits the phosphor?  From what little I’ve read, the phosphor is much slower and cannot respond quick to the light pulses.  Is it possible that this phosphor’s slow speed is responsible for the loss of efficiency as the JT’s frequency increases?

Note: The 1N5817 Schottky diode is an excellent 1 amp rectifier for high frequency power, but it is not commonly available at the local electronics stores.  So the neophyte, not realizing how important the recovery time is, decides to substitute the 1N4002 1 amp rectifier, or the 1N4148 diode.  The 1N4002 and similar has the slow recovery problem and this will cause excessive losses.  The 1N4148 is not made for high currents and the forward voltage drop will be much higher and the losses will be much higher.  In both cases, the benefits of using the 1N5817 are lost and it would most likely be better if the LED was used alone with no rectifier.

Back to experimenting…

The Coil: Explanation  The coil, the heart of t he Joule Thief, is somewhat mystical in that those that understand basic DC and AC theory do not understand how a coil works.  I will admit that I am no expert on magnetic theory but one does not need to know magnetic theory in order to understand how the coil works or how a transformer works.  I’ll try to avoid any magnetic theory or heavy math.

A Joule Thief coil has two windings, only one of which is needed to do the voltage boosting.  This winding is commonly called the primary mainly because that is what the input winding is called in a transformer.  The second winding, called a secondary in a transformer, is called the feedback winding in a JT coil, because it does not perform a function needed to boost the voltage, it just inverts the signal which is fed back to the base of the transistor to keep it switching on and off.  The transistor inverts the signal coming into its base, therefore the coil must invert it a second time in order for the feedback to be positive, and cause the circuit to oscillate on and off.

Basic Single Winding Inductor  Let’s use a singe winding coil, also called a choke or inductor as an example of how a Joule thief boosts the voltage.  We have a circuit that is a coil and a LED in series, with one lead of the coil connected to the positive of the 1.5V battery and the cathode lead (flat spot) of the LED connected to the negative.  No current flows and nothing happens because the battery is only 1.5V and the LED requires 2V to 3.4V to conduct and make light.  We put a push button switch in parallel (across) the LED leads.

Every time we press the push button, current flows from the positive through the coil and switch to the negative.  In a short time the current builds up in the coil, storing energy.  We release the push button, and the current flow is stopped, but the energy in the coil must find somewhere to go, so it tries to continue flowing, but the switch is no longer closed.  The voltage across the coil rises, until it reaches the forward voltage of the LED, and the current then starts to flow through the LED until the energy has dissipated, and the LED makes a brief but bright flash of light.

But we want the LED to keep on producing light, so we press and release the pushbutton switch rapidly and the LED flashes quickly come, one after the other.  This quickly gets tiresome on the finger pressing the pushbutton.

We remove the pushbutton and replace it with a transistor, with the emitter connected to the negative and LED cathode (flat spot) and the collector connected to the coil and other lead of the LED.  Now we have a transistor that can switch so rapidly that the eye can’t see it, and the LED looks like it is continuously on.  But we need to feed back some of the energy from the coil to keep the transistor switching.  We wind another feedback winding over the original coil, and connect it so that the feedback to the transistor’s base lead is positive feedback.  We connect one of the feedback winding’s wires to the positive of the battery, and the other lead to one end of a 1000 ohm resistor to limit the current to the base to a safe level.

Now when the battery is connected, the  LED lights brightly and the circuit keeps switching the transistor on and off tens of thousands of times a second.  It’s as if an invisible finger is pressing that pushbutton switch at more than ten thousand times a second.

Coil Winding   There are several ways to wind the coil, all of which will do the job adequately.  An easy way to wind two windings is to wind both wires at the same time.  This is called a bifilar winding.  We can take 20 feet of telephone wire, which has two 24 AWG solid conductors, and wind this onto a AA cell.  It will have about 60 turns, and then we remove the AA cell and tie this coil in a few places with some wire ties or tape to keep it from unwinding.  We connect this up just like any other Joule Thief coil, and it works, the LED lights up.  But 20 feet of 24 AWG wire has about a half ohm of resistance, which causes a loss of power as the current passes through it, and causes a loss of efficiency; less of the power gets to the LED.

We can use heavier wire, but then the coil gets too large and bulky, and it costs more.  Instead we wind the coil on a core made of ferrite, which concentrates the magnetic field, increases the inductance, and allows us to use a much shorter length of wire.  This ferrite core can be a toroid (small donut), a rod or bar (see Figs 3 & 4), or a piece that looks like a link of a chain, but a square loop instead of a loop with rounded ends.  With the rod or bar, the magnetic field is concentrated in the core, but it must get from one end of the rod or bar to the other end, so it has to go through the air.  We can get a coil with the same inductance but shorter wire by using a core that has ferrite from one end of the coil to the other end, and looks like a chain link.  The coil is wound around one leg of the chain link, and the magnetic field goes through the ferrite link to the other end of the coil.  We can take out the squareness of the chain link core, and we then have a shape that looks like a donut or toroid as it’s called (see Figs 1 & 2).  This gives us a short magnetic path, a lot of inductance in a small package, and the windings can have a short wire length and very low resistance.

Back to the coil.  The wire can be wound bifilar, or it can be wound as two separate windings on the same ferrite core (see the attached pictures).  In the pictures I reduced the windings to just a few turns so that it can be clearly seen how they are wound.    One picture (Fig.1) shows the working Joule Thief with the windings wound bifilar, the other picture (Fig.2) shows the windings wound next to each other.  They are the same electrically to the Joule Thief.

In the Figs 2 and 4 you can see that both windings are wound in the same direction, as if the connection point is ignored and the turns are the same coil.  If we ignore the connection point and treat the windings as if they are one single winding, we can measure the AC voltage between the ends and we find that it is twice what it would be if measured from the connection point to either end.  In either case, the windings must be connected so they are in series, and the voltages of the two windings add.  If you measure the AC voltage from the end of one winding to the other, it will be twice as much as when you measure from the connection point of the two windings to either end.

The other  important point is that the coil has to store energy in the magnetic field.  This energy is from the current when the transistor is turned on, and when the transistor turns off, it has to go somewhere because the transistor suddenly seems to have been disconnected – it’s as if it isn’t there.  The LED is still there, so the stored energy causes the voltage to rise, and when the voltage gets to the forward voltage of the LED, current starts to flow through it.  The stored energy is dissipated in the LED, and the LED puts out light briefly.

One more thing: the amount of energy stored is equal to the current squared times the inductance, divided by 2.   The important point to remember is that if we double the inductance (wind more turns on the coil), the energy stored will increase.  But the wire has resistance, and the current will decrease.  If the current decreases to 1/2, then the energy stored will be 1/2 squared or 1/2 times 1/2 or 1/4.  Looking at what has happened, we are worse off.  It would be better to reduce the inductance by reducing the amount of wire on the core, which will reduce the resistance and allow more current to flow.  If we reduce the wire resistance to 1/2, then the inductance is reduced, but twice as much current can flow.  The energy stored is twice as much current squared or 2 times 2 or 4 times as much energy as before.  The inductance is less and causes less energy to be stored but we still come out ahead because the energy stored is 4 times the current.

We have to remember that with less inductance, the frequency the Joule Thief switches will increase.  There are losses in the ferrite core and in the transistor and in the LED.  These losses rise as the frequency increases, and will eventually become greater than the gains in energy storage.  We need to find a balance where we get enough energy storage, good efficiency and any magnetic or radio frequency fields that might radiate from the circuit will not interfere with radio reception.  Most of the JT coils I’ve built give a frequency between 30 kHz and 300 kHz.  The higher the L (inductance), the lower the frequency and vice versa.  But the inductance is highly dependent on the ferrite material that makes up the core, with the high mu ( µ or permeability) ferrite giving the higher inductance with greatly reduced number of turns.  This is why in Figs 1 and 2 I can wind only 3 turns on the humongous core but it has 93 microhenrys, more than enough to work great with a JT.