Simplest Ever DC-DC Boost 12v to 73v

A simple DC-DC boost converter:
- 12-18v in = ~73v+ out
- Uses only 2 transistors, no IC's
- Can be made out of one CFL light bulb
- Uses fully discreet components
- 17.2kHz switching frequency

Output is only low current but could probably be used to power a small triode preamp stage (Very high impedance)

UPDATE: Thanks to "neumanngregor" for pointed out: ONLY USE ELECTROLYTIC CAPACITORS IN SERIES AS A LAST RESORT, I am not responsible to any damage caused. If possible, use properly voltage rated capacitors. Thanks :)

If you are unsure of the the inductor size, just try tweaking the oscillation capacitor until you get the highest voltage :)*

*A general rule, since this is a constant duty cycle, larger inductors will need a lower oscillation speed.

Step 1 Parts and Specs

This is a very simple 4 component oscillator made from very cheap and easy to access components. It works through the breakdown of the transistor, as the capacitor charges, it gets to a certain voltage then the transistor conducts from the emitter to the collector. This process causes the capacitor to discharge through the transistors emitter, creating a crude sawtooth wave.

Most standard transistors should be able to cause an avalanche on 12v but some transistors dont (as I have found out!) I find the BC337 and BC547 to breakdown perfectly.

The circuit oscillates at ~17.2kHz, nearly above the audible range.

The parts required:
TIP122 or Power Darlington transistor
BC337 or Compatible
0.1uF/100nF Ceramic Capacitor
100uF Electrolytic Capacitor (Preferable 16v if the circuit is ran on 12v but 10v should do fine)
1k resistor
Any high voltage capacitor. I used two 63v capacitors in series to double the voltage rating.
Any suitable diode with high enough voltage (1N4001 isn't high enough voltage, its rated at 50v)
An inductor with inductance between 1-5mH should be fine

Comments

[GuardJanissary]

How can I get a higher voltage with a lower input(or not)? With a higher mH's inductor? Thank you.

[akshay.31890]

How feasible is it to boost a 0.1W I/P to about 3 or 4W? The input is an AC supply at UHF range which would be rectified with schottkey diodes (as hey support high freq). i need to boost this rectified DC to a range sufficient to charge an average cellphone (maybe the NOKIA 1100 model)

[qwerty156]

No.
You cant make more energy out of nowhere!

[hlmasterchief93]

Can i modify it to 12v to 30-32V only ?

[anvit]

Hi,
the TIP 122 becomes very hot just after i apply 12 v .

Is it getting short ?

[KITKAT69]

yeah dude was that the same thing that happens to you???? my VU meter only jumps to about 20V+ then suddenly decreases, then when i touch the TIP122 i got a minor burns man even if my input was on 6v..... and I wonder how many amps does that max 73v comprises....... I got a V-DUB using 555 timer and my theory works even if the input was only 6v 4.5a, I may say its a V-quad not a V-Doubler, since my input 6v was double up then the 2nd v-dub even doubles it up so i got 18v+ but it only got low current only for small circuit applications......

Anyone who wants my circuit about V-Quadrupler is welcome...... hence I though it would power up my gun....

[ankitgarg2005]

I need to convert a 6v 4.5A battery into 12v that can give around 2A or more current
any schematic? Voltage doubler circuits on the web gives around 100mA which is not sufficient for motor driving. Thanks...

[KITKAT69]

hey dude do you want my V-Quad circuit (Voltage Quadrupler circuit) its has two DC voltage doubler circuit based on IC LM555 hooked up in parallel in a small circuit board, it only got around 800ma current though.....

[KITKAT69]

the schematics also looks confusing about which of which was TIP122 and BC337 and the polarities of the series capacitors......

my god I think I have burned my TIP122....

[KITKAT69]

WTF.....!!! this circuit was pretty perfect as a Battery Step Up circuit for COIL GUNS, Thank you very much for this DUDE..... I can't wait to test your stuff...... I appreciate this..... :P

[reehan]

hi there .. i am new to this site.. and i find this instructables very interesting ..i would like to knw who many amp's it draws??thanks in advc

[ahassan123]

this is something great , but i want to know what the " NC " means, does that mean that it is not connected??

and one more thing i didn't find TIP122 in Multisim ... is there any equivalent transistor?

thanx alot

[pyrohaz]

Hi, yep NC means not connected :) yeh, unfortunately you probably wont be able to simulate my oscillator circuit as it works off the breakdown of a transistor junction causing a brief short circuit, and therefore oscillating etc... as far as I know, not many simulators can accomplish this.

[CoolKoon]

Oh well, this is quite an interesting design. So far I've only seen designs that rely either on a 555 timing circuit for rapidly turning the transistor on and off. How unfortunate that since your circuit's basically "cheating" I couldn't simulate (and tweak) it in Multisim (just like the others). I could really use a design like yours, unfortunately (with no simulation option) I have to understand it, in order to be able to modify it and cut the resulting voltage to 48V. Your recommendations are quite good (and I'm planning to use a voltage regulator too), but there are some problems with proceeding with the unmodified circuit: first, the Zener diode decreases the overall efficiency (after all, it conducts above the Zener voltage, the higher voltage, the more), second, the solid-state voltage regulators' ratings drop quite sharply as the Vout to Vin difference increases (and besides the adjustable LM317HVT I have has a maximum rating of 60V).

What I'm planning to do is to slash the voltage down to about 50V, then using the LM317HVT to keep the voltage at 48V. Could you add some theory or point me to some regarding this circuit? I'd really appreciate it.

[anvit]

hi. thnx for ur help.
I have following doubts.

1) I couldnt get 1N4005. So i got 1N4007. Will it work ?
2) Do we have to connect the 12V battery between +12v and the ground ?

I am new to Electronics, so plz help me.

[pyrohaz]

Hi, a 1N4007 will work absolutely fine :) and yes, for point 2, the positive side of the battery goes to +12v and the 0v of the battery goes to ground.

[anvit]

I tested this circuit in multisim, but the Tip 122 was not there, so i connected another NPN darlington transistor. The output is showing only 13.75 V...
Can any one explain ?

[pyrohaz]

One of the problems with simulating this circuit is the fact that it uses a transistors "negative resistance" region, hence why the transistor is connected backwards. Its this function that allows it to oscillate and I just dont think that simulators can simulate this function (as far as I know; if your simulator features a UJT, you could try a ujt relaxation oscillator or something in place of the transistor for simulation purposes)

[mysteriouz]

i have tested the circuit or do the simulation in multisim (software), but it didn't work! i made sure that all the connections were properly connected. when i connect the oscilloscope to the output (that is the positive of the oscilloscope to the diode's cathode or to the capacitor then the negative one with respect to ground), the voltage reading is just 12 V straight line w/c is only the input. i need help on how to produce or make see the desired output by this method!

[Ke-Bob]

If I don't have a 100uF Capacitor will a 220uF one work ok?

[pyrohaz]

Absolutely! infact, the circuit really should work if you replace that capacitor for a resistor, 100ohm?

[Ke-Bob]

Ok great! One last question, how common is the TIP122 and the BC337? I'm kinda new to electronics so can they be found in most electronic supply stores?

[pyrohaz]

Well.. I think in my opinion, they're quite popular due to them being quite cheap BUT any power NPN transistor and small signal transistor should do, i only used these parts since they were in my electronics box :)

[coilsinamotor]

will transistors in a cfl work?

[pyrohaz]

Yes if you can work out the emitter base and collector :)

[ntewinkel]

 Thanks for this great instructable! I'm a newbie still, and wonder if your circuit is tweakable to get a lower voltage. My goal is to get around 48 volts from 12 volts. Thanks!

[pyrohaz]

Hi there, sorry for the late reply! Yes, 48v is possible. The least efficient way of doing this would be to use a linear regulator or use zener diode regulation. This will mean that when the output voltage is above the zener voltage, it will get shorted to ground temporarily to keep it to the zener voltage. The closest zener voltage to 48v is 47v, so less than 3% off :) A zener diode would be connected with the Cathode to the output POSITIVE side and the anode to 0v. (A zener diode works in reverse to a normal diode).

[atofangchi]

Thanks for your great help. I need to convert a higher input voltage ( 17 V) to lower output voltage( 4-6 V) in order to generate high output current. Can I use this device? Thanks

[pyrohaz]

Hey, that would require a buck/buck boost converter. If you change where the inductor is in the circuit, it will create a lower negative voltage at the output. here is an example of how you might make this one: Since the diodes cathode is pointing towards the emitter of the transistor, this denotes the output voltage will be negative. Buck boost is a combination of converters: Buck (lower the voltage) Boost (boosts the current) meaning instead of just lowering the voltage, the current increases too. this will increase efficiency :)

[neumanngregor]

If you couple capacitors in series, you get half the capacitance, and double voltage rating. But coupling electrolytic capacitors in series to get higher voltage rating must generally be discouraged. For this to work, you must be sure that the two (or more) caps share the voltage load properly; a resistor network can augment this, but if leakage currents are markedly different or the capacitors age differently, you are looking at a potential disaster, so do this only as a last resort, if at all. This is a variant of the Joule Thief. Btw those inductors you can recover from old PC CRT monitors, you get from 4-470mH and they are nicely marked with the value too.

[pyrohaz]

Ok ,Thanks for that :) I was just saying, from personal experience they haven't exploded or such so you are right, I should add in the safety rating.

Yep, but taking apart computer monitors is something that I don't think everybody would do. Computer monitors have much more high voltage components than a simple CFL light bulb and not make people have old CRT's at hand, I just said CFL's because there normally around the house (I can think of a few places that I have one).

BUt thinking about it, if the voltage is only ~73v and the capacitors are both rated at 63v, that gives 126v (if perfectly balanced obviously) so:

If the first capacitor was taking 80% of the voltage due to being imbalanced, that would only be 58.4v, wouldn't this be under the 63v rating?

[neumanngregor]

Shure it will work that way, but i wanted to poit that out, somone cud mod your schematic for bigger voltage output and get som bad surprise. The CRT are from my company and thet trashed them in smaller pices befor they go recyling, they dident need some inductors to be trashed xD. Btw in a CFL you work with somthing near 1000-2000V depending on the CFL, in CRT you have from 3-15kV depending on the schematic, they even have 555 IC Timers and a lot of variable resistors.
Keep the good work and have fun.

[pyrohaz]

Thanks again dude :D

[john3f]

There is no load. If you add a load, (with some heavy load) output voltage will drop down to a bit higher than supply voltage. All the circuits like "joule thief" and this are based on the simplest version of "step up pwm convertor".