How to build a GIGANTIC Joule theif for 288 V DC electric car?

[hartiberlin]

Better go with standard NiMH Batteries.
These you can use to charge up with a big sized Joule Thief.
Li-Ion batteries will overcharge and could blow up.
They don´t like the high voltage pulses.

But with NiMH it is no problem.
Looking forward to see your implementation.

Regards, Stefan.

[acmefixer]

Hello Group, I've never taken an electronics course, I use to tinker with toys when my dad wouldn't buy the correct size batteries, and figured out electronics the old school way...

So I've seen videos of Joule thiefs and know how they work...

I am building an electric car, I will be buying Thundersky 100 amp hour cells, they are 3.2 volts each, and for my total pack I will purchase 90 units...   

90 x 3.2V @ 100 amp hour = 288V @ 100 ah,  28.8 kW battery pack...

Now here is the issue with using lithium batteries, if you ever drain them, they loose life, they are best used up until 40% is remaining... and then charge them back up.  This extends their life cycle 2000 cycles until 80%, and then 3000 cycles @ 70% then after that they become ... ?


I am going to spend thousands of dollars on these batteries has anyone ever made a joule thief that will for work for my application?  That is my question for you bright people... 

When I deplete my batteries to 50% can I turn on my joule thief circuit... or have it constantly on?  Can I charge with the joule thief connected? 

[if anybody would like to learn more about building Electric Vehicle simply go to EVTV.me, or youtube: GAVs EV]

I look forward to your responses,

Don

This kind of power demands a controller made for that purpose.  I've seen the insides of a hybrid controller, and they use copper bus bars the width of 2 to 3 cm and thickness of 1 cm, so we are talking about huge currents.  You would have to have a thousand transistors to get a Joule Thief to handle that much current.  The Joule Thief is for AA cells, so you can forget about using that.

[acmefixer]

The Wikipedia article on Joule Thief is a mish-mash of errors, changes, attempts at corrections, but it's a mess.  Many others have added their "view" and it's just like the blind men and the elephant: one blind guy feels the tail and says this elephant is like a rope, another feels the trunk and says it's like a snake, another feels the leg and says it's a tree, etc. If you want to find out how bad it is, click on the discussion tab and read as much as you can stomach.  Better yet, just forget it.

Also, the conventional Joule Thief is lucky if it gets 67 percent efficient, usually it's somewhere between 50 and 70 percent.  The boost converters used in high power applications have a bank of IGBTs or high current MOSFETs and are designed to handle very high currents.  You need to have huge copper bus bars to handle that kind of power.  Remember that 1 horsepower is 0.746 kW, so if you have a 50 HP motor, that's 37 thousand watts!! 

To answer your question, yes that is a Warp 9, from a conversion that was recently finished.  I am using the Warp 11 HV with my conversion, waiting for the adapter plate to come.  My car weighs 5000 lbs and I am going to become the benchmark, nobody else I know has used one of those big boats that was created in 1985.

As per the JT, from wiki:

(Joule Thief) The circuit uses the self-oscillating properties of the blocking oscillator, to form an unregulated voltage boost converter. As with all power conversion technology, no energy is actually created by the circuit. Instead, the output voltage is increased at the expense of higher current draw on the input. As a result, the amount of power entering the circuit is the same as the amount leaving, minus the losses in the conversion process.

(Boost Converter)
Battery powered systems often stack cells in series to achieve higher voltage. However, sufficient stacking of cells is not possible in many high voltage applications due to lack of space. Boost converters can increase the voltage and reduce the number of cells. Two battery-powered applications that use boost converters are hybrid electric vehicles (HEV) and lighting systems.
The NHW20 model Toyota Prius HEV uses a 500 V motor. Without a boost converter, the Prius would need nearly 417 cells to power the motor. However, a Prius actually uses only 168 cells and boosts the battery voltage from 202 V to 500 V.

[Pirate88179]

The Wikipedia article on Joule Thief is a mish-mash of errors, changes, attempts at corrections, but it's a mess.  Many others have added their "view" and it's just like the blind men and the elephant: one blind guy feels the tail and says this elephant is like a rope, another feels the trunk and says it's like a snake, another feels the leg and says it's a tree, etc. If you want to find out how bad it is, click on the discussion tab and read as much as you can stomach.  Better yet, just forget it.

Also, the conventional Joule Thief is lucky if it gets 67 percent efficient, usually it's somewhere between 50 and 70 percent.  The boost converters used in high power applications have a bank of IGBTs or high current MOSFETs and are designed to handle very high currents.  You need to have huge copper bus bars to handle that kind of power.  Remember that 1 horsepower is 0.746 kW, so if you have a 50 HP motor, that's 37 thousand watts!!

I agree with most of what you say except for the efficiency of the JT.  We have seen many at well over 90% and some the data is still being collected (Like Dr. Jones's  8 times more out than in JT circuit)  Gadgetmall's circuit was tested at about 99%.  Most of my circuits are very efficient as well.  The problem with the catch all term joule thief, is that it now encompasses many similar but different circuits.  To me, it has always been much more than a simple blocking oscillator.

Bill

[acmefixer]

I agree with most of what you say except for the efficiency of the JT.  We have seen many at well over 90% and some the data is still being collected (Like Dr. Jones's  8 times more out than in JT circuit)  Gadgetmall's circuit was tested at about 99%.  Most of my circuits are very efficient as well.  The problem with the catch all term joule thief, is that it now encompasses many similar but different circuits.  To me, it has always been much more than a simple blocking oscillator.

Bill

I was following Jones' experiments a the early stage, but I gave up after awhile, there was just too much noise (other comments, etc.) to make any sense. I don't know what the final outcome was.  All I know is that the circuit he showed early on was not a JT in that it used the emitter for driving the load.  I have had efficiencies in the low 90's with my Supercharged JT, but never anything higher than 70s with a conventional JT. This link is to an "apples to apples" comparison using a luxmeter for measuring actual light output.

BTW, I have my doubts when I measure the LED current using a "CSR" resistor.  That is why I bought and used the luxmeter.  If I can measure for example 1000 lux for 20 mA pure DC through the LED, then if I measure 20 mA pulsed DC and get 1000 lux, I know that the pulsed current measured by the meter is accurate.  I tried this and I didn't get what I thought I would.  The meter measured higher with the pulsed DC.  Even so, the pulsed DC measurement is still a good relative measurement.