what ive tried so far doesn't really give much output... 47uF capacitor with 50ohm 5W resistor connected between both bases...... it generates pulses but very weak output.
You seem overly focused on HAVING to do this with the two transistors you have in your hand. really, just get a PAIR of transistors suited to the task in the first place. I am speaking only of your oscillator, not the overall impracticality of the concept.
I "cant" get a pair of identical power transistors.. I dont have a way to buy any.. is there any way at all to do it similarly with NPN and PNP 40W power transistors? i cant afford to get a second NPN or PNP. or even have a way to do so.. unless i'm going to attempt to use 2N3904 signal transistors to drive a power transformer... because they wouldn't last a second before popping like bubble wrap... even if I had 20 of them in parallel to share the current. it wouldn't be worth trying to use 2N3904's
if I had to.. would getting 2 pairs of these be good enough to use? http://www.amazon.com/TIP35C-TIP36C..._UL160_SR160,160_&refRID=0XQFGZS2WXP9SABBH58G
As long as they are what they say they are then they would be fine for experimenting with.
Just heed all the advice given earlier about this not being an easy project to build successfully. What sort of battery are you going to use ?
Just heed all the advice given earlier about this not being an easy project to build successfully. What sort of battery are you going to use ?
In the 40's and 50's car radios and mobile transmitters used a mechanical multivibrator, a transformer and a rectifier to generate the B+ for the tubes. This can be found by Googling "vibrator power supply"
In the 60's and 70's the mechanical vibrator was replaced with a two transistor circuit. The transistors should be the same type, and rated for the amount of power needed. With the right parts 200 watts or more is possible. A pair of PNP's or NPN's can be used by swapping the power input polarity. To avoid losses the transistors should have a low saturation voltage (VCEsat). The best parts for this at the time were large germanium power transistors. 2N3055's will melt and generate low power. Power mosfets should be great, but didn't exist when I made these.
This circuit was commonly used in linear amplifiers for two way radios in the 70's with 2 or 4 large germanium transistors generating 500 volts or more for a 100 to 300 sweep tube amplifier. The transformer was a large tape wound toroid. I successfully used the circuit to power tube audio amps, and with rewound secondaries, transistor amps in my van back in the mid 70's.
You can get it to work with a reverse connected filament transformer, but power will be limited due to magnetic losses. Still, 50+ watts should be possible.
In the 60's and 70's the mechanical vibrator was replaced with a two transistor circuit. The transistors should be the same type, and rated for the amount of power needed. With the right parts 200 watts or more is possible. A pair of PNP's or NPN's can be used by swapping the power input polarity. To avoid losses the transistors should have a low saturation voltage (VCEsat). The best parts for this at the time were large germanium power transistors. 2N3055's will melt and generate low power. Power mosfets should be great, but didn't exist when I made these.
This circuit was commonly used in linear amplifiers for two way radios in the 70's with 2 or 4 large germanium transistors generating 500 volts or more for a 100 to 300 sweep tube amplifier. The transformer was a large tape wound toroid. I successfully used the circuit to power tube audio amps, and with rewound secondaries, transistor amps in my van back in the mid 70's.
You can get it to work with a reverse connected filament transformer, but power will be limited due to magnetic losses. Still, 50+ watts should be possible.
Attachments
I was thinking of using a limiter for current in series with the battery and center tap for more efficiency for lower power devices for longer battery life.
about 4-6ohms non-inductive. with a switch to short it for higher power devices.
That way I could use it to charge a phone or power small laptop for emergencies and switch between the two for longer battery life and more efficiency or more power.
I was thinking using either two 12V electric scooter batteries in parallel or buying a few cheap 6V torchlight/lantern batteries from the dollar store for an emergency use like if our power goes out and we need to charge a phone or something/. and putting two of them in series and two in parallel
about 4-6ohms non-inductive. with a switch to short it for higher power devices.
That way I could use it to charge a phone or power small laptop for emergencies and switch between the two for longer battery life and more efficiency or more power.
I was thinking using either two 12V electric scooter batteries in parallel or buying a few cheap 6V torchlight/lantern batteries from the dollar store for an emergency use like if our power goes out and we need to charge a phone or something/. and putting two of them in series and two in parallel
Any resistor in series with the power supply will only waste power and lower efficiency since it converts electrical power into waste heat. The more voltage across the transistors, the better since the saturation voltage is pretty much constant.
This circuit is simple, but old. The main efficiency losses will be in the transistors and the transformer. The transistors should be chosen for low saturation voltage and the resistors on the bases should be chosen by trading off idle (no load) power consumption VS maximum power output. The frequency of operation is determined by the transformer and the capacitors on the bases. The best efficiency will be obtained at one frequency, which must be determined by experiment, and will likely not be 60 Hz. The frequency doesn't matter much for most loads, with motorized (AC motors) being an exception. The toroidal transformers we used worked best around 500 Hz which made filtering the DC output a pain for running an audio amp.
This circuit is simple, but old. The main efficiency losses will be in the transistors and the transformer. The transistors should be chosen for low saturation voltage and the resistors on the bases should be chosen by trading off idle (no load) power consumption VS maximum power output. The frequency of operation is determined by the transformer and the capacitors on the bases. The best efficiency will be obtained at one frequency, which must be determined by experiment, and will likely not be 60 Hz. The frequency doesn't matter much for most loads, with motorized (AC motors) being an exception. The toroidal transformers we used worked best around 500 Hz which made filtering the DC output a pain for running an audio amp.
I don't think cheap lantern batteries are going to be up to it tbh. A 10 watt CFL would draw 1.6 amps from a 6 volt battery assuming the convertor were 100% efficient. In practice you would be lucky to get 40 or 50 % (I bet even that is very optimistic) unless you had the tools to develop the circuit to suit the parts and transformer.
A battery running a bank of high brightness LED's would be far more practical and the battery would last infinitely longer.
A typical laptop is going to draw around 1.5 to 2.5 amps @ say 12 volts (give or take) so that is 30 watts. So that's 5 amps draw (again at 100% efficiency) from a 6 volt battery.
Its just not workable imo.
A battery running a bank of high brightness LED's would be far more practical and the battery would last infinitely longer.
A typical laptop is going to draw around 1.5 to 2.5 amps @ say 12 volts (give or take) so that is 30 watts. So that's 5 amps draw (again at 100% efficiency) from a 6 volt battery.
Its just not workable imo.
really? I managed to get a 120V stereo amplifier to run off my previously poorly 5-10% efficency positive squarewave generator circuit.. at quite loud volumes as well.. Even on just 9.5V or less it still worked.. the way I did it was instead of using a 1K between the base of the NPN I used a solid wire instead. with a 4uF non-polarized 250V capacitor for the oscillator and a 1K for the oscillator. in all only used a 1.5A 9.5V input to 120V AC transformer single phase and used the output to power the stereo. It had enough power with one channel being used to drive a 75W 6.5inch subwoofer to decent bass levels
It worked great even off only a 9.5V 2A power supply. (adjustable switch on the front from 3V to 11.7V) was using it to test-power the circuit
I believe the circuit might of been posted earlier on this forum thread.. Pretty inefficient but worked.. The stereo must not draw very much idle current for it to work seemingly.. It has a transformer in it which is where it gets its power from.
A better design with two 90W power NPN transistors would be massively more efficient at worst.
And would using darlington/sziklai pairs be more efficient?
It worked great even off only a 9.5V 2A power supply. (adjustable switch on the front from 3V to 11.7V) was using it to test-power the circuit
I believe the circuit might of been posted earlier on this forum thread.. Pretty inefficient but worked.. The stereo must not draw very much idle current for it to work seemingly.. It has a transformer in it which is where it gets its power from.
A better design with two 90W power NPN transistors would be massively more efficient at worst.
And would using darlington/sziklai pairs be more efficient?
Last edited:
i managed to build it using two NPN TIP35D 25A 125W transistors
only used two 1K's and no capacitors and it works great!
can even be powered by a set of 4AAA batteries. it works exceedingly well
with 12V it can even power a 120V oscillating table fan directly!
I'm thinking of changing the two resistors to something lower like 200ohms or so. would that help? or would a darlington pair work better than lowering resistance of the resistors?
only used two 1K's and no capacitors and it works great!
can even be powered by a set of 4AAA batteries. it works exceedingly well
with 12V it can even power a 120V oscillating table fan directly!
I'm thinking of changing the two resistors to something lower like 200ohms or so. would that help? or would a darlington pair work better than lowering resistance of the resistors?
I'm also able to power a 20W fluorescent bulb from it! (tried a 23W bulb and it tried but just wasn't quite able to power it)
but would using lower ohm resistors get better performance out of it? or even changing the transistors to darlington transistors?
wasnt able to edit post since its been a few hours
but would using lower ohm resistors get better performance out of it? or even changing the transistors to darlington transistors?
wasnt able to edit post since its been a few hours
That's good for you but only vaguely informative as to what mean for someone reading your posts. Can you sketch and post your circuit diagram so we know exactly what you are talking about? If that's too difficult, can you copy and paste similar circuit from the thousands out there on the net?
http://www.chinaicmart.com/uploadfile/ic-circuit/2009793442234.gif
I used that diagram. but what I have is 1/2W 1K resistors. and TIP35D's with 11.7V DC (from a small wall adapter power supply to test with before using a battery or something)
somehow ive managed to cram the TIP35D's onto a breadboard to make the circuit easier to build
one of them (top transistor) gets very slightly warmer than the lower one for some odd reason... dont know what could be causing that... possibly the transformer isn't wound right im guessing?
The resistors get a little bit warm. with a 20W light bulb on the output of the center tap transformer
20 watts is literally the max. anything more and it just seems to stop outputting useable power
but it can actually drive a 20W induction AC motor efficiently! and im very impressed by that.
I used an enormous 75W 120V +12V/0/-12V transformer with 0 being center tap
i assume lower voltage on input of the transformer could make the circuit function at a lower input voltage? perhaps a more efficient transformer?
I used that diagram. but what I have is 1/2W 1K resistors. and TIP35D's with 11.7V DC (from a small wall adapter power supply to test with before using a battery or something)
somehow ive managed to cram the TIP35D's onto a breadboard to make the circuit easier to build
one of them (top transistor) gets very slightly warmer than the lower one for some odd reason... dont know what could be causing that... possibly the transformer isn't wound right im guessing?
The resistors get a little bit warm. with a 20W light bulb on the output of the center tap transformer
20 watts is literally the max. anything more and it just seems to stop outputting useable power
but it can actually drive a 20W induction AC motor efficiently! and im very impressed by that.
I used an enormous 75W 120V +12V/0/-12V transformer with 0 being center tap
i assume lower voltage on input of the transformer could make the circuit function at a lower input voltage? perhaps a more efficient transformer?
Last edited:
Ok, so it's this with 1k resistors and similar transformer, right? The transistors won't be sharing current equally in a simple circuit like that, without emitter resistors or very close matching so yes, the waveform will be skewed and one transistor will be dissipating more power due to a couple of reasons. A 'scope would show you in an instant but I guess that's not possible unless you take your circuit to a college, school, repair shop, friend etc.
An externally hosted image should be here but it was not working when we last tested it.
The transistors don't share current, they switch alternately. It works due to differences between pass devices. Search ZVS - Zero Voltage Switching.
Sure Jason, I stand corrected. I realised as soon as I posted that I should have said something like "supply equal power" rather than "share current" but the term "sharing" communicates the idea better.
Is there a way to make it more equal or slightly more efficient? and possibly have a higher output frequency? how could I use capacitors or different resistor values to get a higher output frequency?
Could I use my computer mic in with a 10K resistor and a select capacitor and record "Audacity" and view the waveform?
or would voltage spikes from it switching destroy the mic in? on the 12V winding not 120v winding!
and if it did work would it be accurate to any degree?
Could I use my computer mic in with a 10K resistor and a select capacitor and record "Audacity" and view the waveform?
or would voltage spikes from it switching destroy the mic in? on the 12V winding not 120v winding!
and if it did work would it be accurate to any degree?
Last edited:
You may be able to use Audacity in a limited manner depending on oscillation frequency, because the high harmonic content of switched waveforms means you may need a wider bandwidth than the sound card can handle. In any case, you will need to make a precise voltage divider, not just fit a series resistance to the input, which won't protect the sound card's safe input max. of around 1V nor give any certainty to the scaling factor.
I think it's been decades since BJTs and low frequency, iron core transformers were used in this type of inverter circuit (ca. 1970) Mosfets are the universal device for increased switching efficiency and Ferrite core transformers are essential for use at the high frequencies required. Many articles on this form of power supply are pretty demanding reads for the layman and unless you are able to source modern components, it makes sense to stick with what you have, if it works and already meets your needs. If you are interested in old style power supplies and experimenting with getting better performance, why not post in the Power Supply forum here?
I think it's been decades since BJTs and low frequency, iron core transformers were used in this type of inverter circuit (ca. 1970) Mosfets are the universal device for increased switching efficiency and Ferrite core transformers are essential for use at the high frequencies required. Many articles on this form of power supply are pretty demanding reads for the layman and unless you are able to source modern components, it makes sense to stick with what you have, if it works and already meets your needs. If you are interested in old style power supplies and experimenting with getting better performance, why not post in the Power Supply forum here?
Be very careful trying to connect a PC to this.
If you really want to try then connect a 10k AND a 1 ohm in series. Have the 1 ohm connected to battery negative and the 10k to the collector of either transistor. Connect the mic input across just the 1 ohm.
You could easily have spikes of 50 to 100 volts across the windings.
Its maybe not a good idea on balance... not worth zapping the PC.
- Status
- Not open for further replies.