Bedini transistor Selector

[nvisser]

Here's a diagram from an Application Note for a simple circuit enhancement to speed up transistor turn-off.

It's possible to easily adapt the same principle to NPN circuits as well - by reversing the polarities of the semiconductors.

When Base Drive ceases, the circuit removes 'base charge' quickly from the switching transistor which expedites its turn-off and greatly enhances the efficiency of the switch.

This technique may be used in any circuit where a switching transistor is utilized in order to improve its ability to turn-off rapidly.

Hi Seamonkey
Could you please explain this circuit as it makes no sense to me at all
Vissie

[SeaMonkey]

Here's how it works: (See #11 above for circuit diagram)

When a positive input is applied to the base of Q2 (Drive) it will go into conduction thereby providing base current to the PNP Switching transistor Q3. The base current path is from Ground (common) upward through the 68 Ohm emitter resistor of Q2, through Q2 emitter to collector, upward through L1 and into the base of Q3. This will cause Q3 to turn on to provide power to the buck regulator circuit to the right of Q3.

This base current flow through inductor L1 will produce a small magnetic field with resultant energy storage.

Now, when the base input to Q2 is taken Low (0 Volts) transistor Q2 will cut off and the base current upward to Q3 will suddenly cease. At this time the magnetic field of L1 will suddenly collapse inducing and electrical output across L1 which is applied to the base-emitter junction of transistor Q1. This brief pulse of electrical energy will be sufficient to take transistor Q1 quickly into saturation which effectively produces a short circuit across the base-emitter junction of Q3. By transistor Q1 momentarily becoming a short-circuit from base to emitter of Q3 it effectively allows an escape route for the carrier charge in the base of Q3 to 'discharge.' It is this rapid removal of the base charge of Q3 which speeds up its turn-off dramatically.

By speeding up the turn-off of transistor Q3 the circuit will operate with higher efficiency and with less power dissipation. In normal operation transistor Q3 will be switching at a fairly high frequency as determined by the controlling Pulse Width Modulator chip (not shown) of approximately 40 KHz.

Transistor Q3 is one of the new family of small, highly efficient Low Vce transistors which are as efficient as a MosFet in applications requiring up to about 6 Amperes of collector current at low voltage. The advantage of using this type of transistor in this relatively low power switching supply application is that it is smaller than an equivalent MosFet.

The circuit as it is shown will operate at less than two volts input, something that would not be possible with a MosFet Q3. So while the MosFet is greatly superior to the bipolar transistor in most applications, this is one example where the new generation of efficient bipolar transistors is the superior choice.

Low voltage devices are becoming very common today with the advent of new low voltage innovations in chip design and fabrication. Low voltage switching power supplies are also very common.

It should be noted that these new Low Vce transistors (Q3) work very well in low voltage devices such as the Joule Thief and the Low Voltage/Low Power Tesla type coils that many in this forum are experimenting with.

[nvisser]

Amazing!
Nice description.
Could this circuit be used to do switching with large Q3 transistors in the tesla switch and scalar charger or any switch that is used to switch bedini coils to give us that sharp cutoff
What is the battery for on the right hand side. Does this circuit change a low dc voltage into a higher one to charge the battery. 1n5820 zener is 20v.

[SeaMonkey]

The speed-up circuit could be used with high side power transistors, yes.

Its primary value though is with low voltage applications (5 Volts or so and less). For higher voltages up to about 15 Volts it is still good but for
applications where the voltage is 12 Volts or more the MosFet, and
associated driver chip, is easier to deal with.

The 'buck' circuit is used as a 'step-down' power supply - it is often called
a DC transformer even though it is just a single inductor. This circuit is
a battery charger for a single cell re-chargeable up to about 3 Volts and
would be driven by a 5 Volt supply.

It would function at higher voltages, up to 15 or so, and would produce a
lower voltage, higher current output as determined by the PWM chip
(not shown) which regulates the output to the desired voltage.

The 1N5820 is a low voltage Schottky Diode which provides the discharge
path for inductor L2 and is called the 'free-wheeling diode.'

To create higher voltages you would want to use a 'boost' configuration.
That is the type of circuit which is used in those cell-phone battery
chargers (6V ~ 7.2V) which are powered by one or two AA cells.

[nalgman]

Definitely not 2N3055 or MJ2955. Even cheaper TIP31C/TIP32C is much better. However I found KSC5027 perform best. Although still dreaming MJL21193/MJL21194. My efficiency is more than 60% on any coil I have.

hello

with what transistor you got the 60% of efficiency?
because i use some tip3055 and it's very good..
thank you!!

[sucahyo]

hello

with what transistor you got the 60% of efficiency?
because i use some tip3055 and it's very good..
thank you!!

With those mentioned transistor. My best efficiency now is 75% using two KSC5027 and with stingo circuit:
http://www.energeticforum.com/renewa...ead.php?t=6462

I found the TIP3055 to be better than 2N3055 but is not as good as KSC5027. Not even TIP31C, but TIP31C die quickly.

[faramog]

With those mentioned transistor. My best efficiency now is 75% using two KSC5027 and with stingo circuit:
http://www.energeticforum.com/renewa...ead.php?t=6462

I found the TIP3055 to be better than 2N3055 but is not as good as KSC5027. Not even TIP31C, but TIP31C die quickly.

Currious what you mean by efficiency ? What are you actually measuring ?

thanks

[sucahyo]

Currious what you mean by efficiency ? What are you actually measuring ?

thanks

COP = output power divided by input power with simple amp meter in this way:


with volt meter in parallel with source / charged battery.

Example video:
YouTube - COP = 69% battery charging battery

[radiant]

This thread is still relevant

I reckon I have found a transistor that is hands down outperforming the MJL21194. I use a 3 pin screw terminal to attach my transistors so I can interchange transistors to see what works good and what doesnt.

Ok the transistor I stumbled upon is a FJL6920. It is a H deflection transistor from CRT tv. It is still available on auction sites also...

Specs:

Vcbo = 1700V
Vceo = 800V
Ic = 20A
ICp = 30A
P = 200W
tF = 0.15-0.2usec

When you do "diode tests" with your multimeter, it shows that it has a built in damper diode across the C & E. I was brave enough to remove the neon from the circuit and it survived with no output battery connected!

Im doing more testing with this transistor...but needing no neon makes life a bit easier for power circuits...

To add, as others have found, the H-deflection transistors work well with SG circuits...very forgiving...

Most hated transistor in my book is the TIP3055...(2N3055 is ok)...

That Tip3055 is a joke...if the lead falls off your charge battery...expect death of the weakest link...there is always one transistor that will cop it in a multi coil setup.

[Michelinho]

Tougher Transistor.

Hi radiant,

I use 2N3055 with a 60V Neon and you still have to be careful of those charge battery lead. Mine saturate from the bemf and the transistor blows.
I've kill only one so far but still have around 75 more as backup.

So if you want a tougher transistor, use the MJL21194 or similar in specsatasheet at ALLDATASHEET.COM - Datasheet search site, Datasheet search site for Electronic Components and Semiconductors and other semiconductors.
MJL21194 - 16 AMPERE COMPLEMENTARY SILICON POWER TRANSISTORS 250 VOLTS, 200 WATTS - ON Semiconductor MJL21194 pdf, MJL21194 description, MJL21194 datasheets, MJL21194 view ::: ALLDATASHEET :::

Take care,

Michel

[brian516]

best price I've found on transistors or anything for that matter.....

Hey I saw you guys talking about how expensive the MJL21194's are, and how expensive transistors are in general.... well i found a website that has better prices than I can find anywhere else. Even cheaper than the ones from China that are 80% of the time FAKE. onlinecomponents(dot)com. I was looking at a Velleman PCSU200 USB oscilloscope/signal gen/spectrum analyzer/bode plotter/transient recorder unit (which is less than half the price of the same that I found on ebay) - anyone have any input on whether it would be a worthwhile purchase? I can't afford a standalone o-scope, so getting something like that where its all wrapped up in one would be killer!