homemade inverter question

WildRice

Flashlight Enthusiast
Joined
Nov 30, 2003
Messages
1,135
Location
SW Michigan
I suppose it would be more efficient and easier to just get some of those booster chips to run stuff of a single AA/AAA, but where is the fun in that. I have been trying different satcure configurations.
dual coil, single NPN
single coil dual NPN
single coil single NPN single PNP
but I keep hitting a power wall at about 50mA output current. I was thinking, the average ECv drop on a bi-polar transistor is what .7v? total voltage across coil would be .8v in 'charge'. from what I remember, germanium components only have a .3v drop but are VERY static sensitive. Are there any other style of bi-polar transistors that world have a low collector/emmiter drop?
Just thought I would toss this one out here...
Jeff
 

Doug Owen

Flashlight Enthusiast
Joined
Jan 30, 2003
Messages
1,992
Jeff,

Probably a question better asked on the Electronics forum?

There are no longer Ge based parts made (at least that I know of), but Vbe is not what is at issue, it's Vsat. You can easily get Vsat bipolar transistors under .1 Volt using parts made for the use.

There is no 'power wall', at least not at this level. I used to work for a 40 or so employee company that made 1500 fluorescent lamps a day. They drove two 15 Watt tubes with about 1.8 amps of 12 VDC. A '3055 class' NPN (Beta greater than 20 at 4 Amps and 4 Volts). 25 turns for primary, 12 turns for feedback, 625 for secondary on a dual C core of half an inch or so cross section with a .005 inch gap. A resistor, two diodes and two caps. We made a bunch of 'em.

Doug Owen
 

tonyb

Enlightened
Joined
Aug 31, 2002
Messages
311
Location
PA
Paralleling two similar transistors with a close beta (current gain) may help. The coil can effect operation also. An Oscilloscope to veiw the waveform is neccesary to check its operation.
tony.
 

WildRice

Flashlight Enthusiast
Joined
Nov 30, 2003
Messages
1,135
Location
SW Michigan
i wish I had a scope. All I got is a DMM. I have been putting a pot across the various resistors and trying different inductors, finding the sweet spots. the best one so far is a single inductor (170uH) and 2 NPN configuration. I was able to get 31mA output current (running a nichia white) off a AA. I am going to be trying the single transistor setup with a dual wound choke (feedback loop on it). just trying the configuration with the highest output power. The 2 NPN setup is nice because it uses a standard inductor. this one will also push 150mA at 3v input (similar to the circuit used in the long life brinkman flashlight). PVC, dollar store flashlight cases, and scrap parts are all I can really afford right now.
Jeff
 

Kram

Newly Enlightened
Joined
Jan 21, 2003
Messages
86
Location
Texas
If you're talking discrete (TO-92 pkg) transistors, I like the 2N3391A. The VCEsat at 200ma is about 0.4V. It's less for smaller currents. This transistor also features a relatively high hfe (over 200 for small signals).

You can get these at Mouser for about 6 cents each in quantity 100 (or PM me and I'll send you a few).

I'm sure there are others.


In my experimentation with similar circuits, I'm not sure that the transistor is causing the current limit you're seeing. The inductor seems to me to be a more likely culprit. A transistor with the characteristics I mentioned helps extend the voltage range you can effectively boost (i.e. will work at a lower voltage) than most of the Radio Shack transistors.

I guess this thread does work better in the electronics forum, doesn't it? /ubbthreads/images/graemlins/blush.gif

Good luck!

Mark
 
Top