Joule Thief

[orbut 3000]

I hope external Thunderbolt SSDs become more affordable and available soon. Ideally sooner than Thunderbolt I is obsoleted by the PTB.

[MarkE]

I hope external Thunderbolt SSDs become more affordable and available soon. Ideally sooner than Thunderbolt I is obsoleted by the PTB.

Interfaces march on.  A consumer grade SSD is more than saturated by USB 3.0 or TB I.  I would not worry about the interface so much, but more about the reliability of the drive.

[Pirate88179]

Yeah... I've done that in the past with no success. Bill suggested it to me again last week and I rand about 20 capacitors on that specific
circuit yesterday and per Bill's suggestion, I have had a 20F cap on it for several hours and the voltage reading never got past the source
voltage, which is about 450 mV. The larger capacitance cuts the electric from my earth battery to that circuit's LED. I was thinking of
demonstrating this on a short video later today.

As far as these little inductors/chokes, I measured and noted the resistance, which is significant. I am pretty sure these LC meters can
measure resistance of these things much better than they can measure the capacitance of them. I tore a few apart a few months ago.
They have a tiny ferrite core and very thin wire, which looks like copper, wrapped around them. The covering, which you can see, is
some kind of insulator. I love these things!

I have about 50 circuits all over my place with these chokes/inductors of various inductance - I got a 20-value set of 200 on eBay
earlier this year on eBay for less than $5 USD delivered from China early this year. The 1 mH's show about 30-31 ohms and work well
with the 3904/3906 transistors without the need of a resistor or capacitor in a joule thief. The 1 mH version of this JTC also works very
well with fully charged AA batteries all the way down to dead batteries with about 400 mV, 0.5 mA. The 10 mH seem to choke the light
out of the LED with batteries that have too much power.

I'll try to find and demonstrate a "simple" way to test capacitance on these and my other inductors/coils. In case you haven't seen this,
here's a fun view of my mess (I mean lab)! lol

http://youtu.be/_36V25kDTEU

Just for clarity, what I suggested on the EB topic was to use a large value supercap in parallel between your EB and JT circuit.  This captures those spikes and also allows you to store usable power to be able to run devices that you would otherwise not be able to run.

Example:  My dmm shows about 1.9 volts or so from my EB.  I can charge my 2.7 volt, 650 F cap fully so...either my dmm can not read the spiky voltage properly, or, the cap is storing all of the energy from the EB, including the spikes to get to 2.7 volts.  Maybe both?

What Mark was suggesting, I believe, was a cap inline with the base and the coil.  Gadgetmall did this a long time ago, as did several others.  I believe he was using a ceramic cap for this.  As a matter of fact, I have a bunch of ceramic caps over here...some have numbers on them (code?) some are blank.  This was why I purchased my meter so I could test the values on them and use them like Mark was suggesting.  Of course, my meter is useless for this so I never have tried this to date.

I hope I have explained this ok.

Bill

[allcanadian]

@Bill

What Mark was suggesting, I believe, was a cap inline with the base and the coil.  Gadgetmall did this a long time ago, as did several others.  I believe he was using a ceramic cap for this.  As a matter of fact, I have a bunch of ceramic caps over here...some have numbers on them (code?) some are blank.  This was why I purchased my meter so I could test the values on them and use them like Mark was suggesting.  Of course, my meter is useless for this so I never have tried this to date.

I would agree completely and the current through the base of the transistor can be nearly as high as the collector/emitter in a bad setup. This is why I made the transition to mosfets using an open gate structure because the gate capacitance is generally very small. Make no mistake that initially it's a real bugger to even get it to run but once you do make the jump the gain in efficiency is well worth it.


I started simple, take a N channel mosfet in which a positive charge on the gate turns it on and a negative charge turns it off. Connect the drain pin from the mosfet to the (+) of a 12v battery terminal and a low wattage/high resistance 12v light (say 1-5w) off the source pin which then goes to the 12v battery (-) terminal.


If you use an alligator clip off the mosfet gate and hold the other end of the lead then the positive charge of your body turns the light on. Now keep holding on to the alligator clip lead and touch the negative terminal of the battery with your other hand... the light turns off. This tells us the whole of our body and the alligator clip to the gate now has a negative charge from the battery. You have just invented a charge detector and this same principal can be used to drive a JT.


If fact I don't even make a hard connection in most cases and a simple wrap of wire around the insulation of a medium to high voltage output will switch the mosfet. The thing to remember is that the charged state(+ or -) of the mosfet gate determines whether the source/drain is conducting(on) or not (off). To be honest I haven't even touched a transistor in years because I find them to be a real pain in the optimus rear endus. The real trick is gate conductor placement because we are dealing with the charged state and we have to start thinking on that level, not current but the charged state which is relative because we are speaking of an electric field. Electrostatics...


Try it and I think you will like it, initially the learning curve is high however once you get a feel for it I don't think you will go back to transistors. The simple charge detector I just described changed my perspective completely, it was that other part of the equation which seemed to be missing.


AC

[TinselKoala]

Yep. Here's part of a basic tutorial series I made to try to explain to a "certain someone" how her circuit components actually work. Like water off a duck's back that effort was, but some other people may have found these demonstrations useful.

http://www.youtube.com/watch?v=tKstLQYayNA