I am making a simple inverter using the circuit below. I am using a 3amp 12-0-12 Transformer. I am aware of the cons of the circuit. So, I never expected regulated Pure Sine Wave output from it. The circuit works fine. But I am able to lit up a 5 Watt CFL Lamp only. I was trying to lit 10 or 15 Watt CFLs, but they didn't work.

How can I increase the output power of the inverter? Should I reduce R3 and R4? I guess, 3 amp Trafo is good enough for my expectation.

And I never saw the LED to lit up. If someone could explain the purpose of the LED and what changes would lit it up, it would be great.

enter image description here

  • 1
    \$\begingroup\$ Well, you'll have to figure out why the circuit isn't providing enough power. On the way to that, you'll probably notice that this is actually a bad idea and you'd want to use an IC that was meant to drive an inverter. \$\endgroup\$ – Marcus Müller Nov 30 '16 at 19:59
  • 2
    \$\begingroup\$ Also, this design is not a very good one. For example using R1, OK but then there should be a decoupling capacitor on the supply line. Using a 390 k in series with 1 k pot, geez, the tolerance on that 390k alone is way more than 1k from that pot. All in all, not designed by someone who know what they're doing. Driving such big NMOS straight from a single CMOS output, sure it "works" bit not something that I would do. \$\endgroup\$ – Bimpelrekkie Nov 30 '16 at 20:12
  • 1
    \$\begingroup\$ @FakeMoustache exactly. I mean, using a simple vibrator to drive an inverter, sure, that's how things started, right? But if one does it the low-tech way, it's probably a good idea to work with a well-designed circuit. There's just too many "scary things" in this circuit. For example the combination of T1 with totally unknown characteristics and C4 and rectangular switching … \$\endgroup\$ – Marcus Müller Nov 30 '16 at 20:19
  • 2
    \$\begingroup\$ also, R6 is 100kΩ in the schematic he probably copied (which itself is probably a copy from somewhere else, and that was a copy of the schematic from the cousin of the guy's wife, who knew someone who had the magazin that printed the article taken from the book, which back then came with an explanation of the circuit and its limits in the late 1970s. I really don't get why people copy things they don't even remotely understand, or have sources for). \$\endgroup\$ – Marcus Müller Nov 30 '16 at 20:20
  • 1
    \$\begingroup\$ @MarcusMüller I really don't get why people copy things they don't even remotely understand That's because they do not know that they do not understand the circuit. They know it "works / does something" so they think it is good design. Those with knowledge are full of doubt while those without knowledge are full of confidence. \$\endgroup\$ – Bimpelrekkie Nov 30 '16 at 20:36

In all probability you have insufficient voltage on the output to run your CFLs.

You are driving the 12-0-12 transformer with a 12 V square wave. That is giving you approximately a 228 V square wave on the output side. The domestic 230 VAC used to power a load is about 322 volts peak and since almost universally there is a bridge rectifier in the circuit of the CFLs they are designed to run from the peak voltage and not the RMS.

You have IMO two choices (and I'm only suggesting this for driving CFLs!!!!):

1) Increase you your input voltage. The turns ratio for your transformer is about 19:1 so to get closer to the 300 V peak you'd need about 17 VDC. You could use a boost convertor to provide this from your 12 V battery.

2) Change your transformer, which may be harder. You'd need about a 9-0-9 volt transformer.

Note.... big red flashing light ......you are setting the output voltage to 300 V RMS by implementing either of these solutions, so don't try to power anything other than the CFLs (or LED 230 VAC lights) https://en.wikipedia.org/wiki/Root_mean_square


How to increase output power of this simple CD4047 based inverter

Probably: replace the design with something that actually is fit to drive an inverter – there's plenty of ICs that are actually designed to drive inverters, and have things like feedback loops. Without a lot of time on your hands, and an in-depth understanding of what's happening and why, you won't be able to improve this very minimalistic "demo" circuit.

On the other hand, there's a lot of application notes from semiconductor manufacturers that actually explain the theory of operation of a device – and list the limits of a design. It'll be easier to do this well from scratch than to save this particular design.

A typical design would use a flyback controller to feed a full-bridge MOS set at more than 100 kHz, keeping the transformer small (cost reduction without reducing power).

The resulting high-frequency AC would be rectified using a classical diode bridge rectifier, generating the intermediate circuit's DC voltage.

From that DC voltage, another four-FET controller would "push" current through one or two inductors and your load, with some kind of feedback mechanism that strives to achieve a sinusoidal output, or just pre-calculated 50 Hz switching.


HF inverter

It's significantly more complicated than your CD4047 inverter, but you actually get a lot more efficiency out of this – so you don't have to buy a higher-rated transformer or overdrive your output voltage to power regular loads.

  • \$\begingroup\$ While this is a good solution it is very much more complicated. The transformer and inductors are Ferroxcube and the driver is very complicated with switching (PWM) in the 10's - 100's kHz. Certainly not a one off DIY project. \$\endgroup\$ – Jack Creasey Dec 1 '16 at 15:42
  • \$\begingroup\$ @JackCreasey very true! Still, having a > 15W inverter for 12 V -> 220 V that is neither a live nor a fire hazard isn't what I'd call a one-off DIY project, to start with. \$\endgroup\$ – Marcus Müller Dec 1 '16 at 15:46
  • \$\begingroup\$ @JackCreasey to be fair, one could replace the CDxxxx with some kind of microcontroller that gets a voltage feedback signal from the secondary side of the transformer – and use that to more accurately switch the two MOSFETs. but then you'd quickly realize that only energizing half of the primary coil probably is a waste of copper, and you end up doing at least one of the two four-quadrant bridges... \$\endgroup\$ – Marcus Müller Dec 1 '16 at 15:50
  • 1
    \$\begingroup\$ Muller. The FET's are operated in switched saturation mode....not PWM. He's using a simple 50/60 Hz transformer. You could certainly use a bridge drive (at 50/60 Hz not PWM) to a single 230/9 V transformer, but it would be more expensive than what he already has working. While it's correct to point out all the downsides of the implementation path he's taken. My objective was simply to show what might make the project as implemented work at some minimal level. The project need not be a fire or live hazard ...you could certainly suggest the use of fuses. \$\endgroup\$ – Jack Creasey Dec 1 '16 at 16:08
  • \$\begingroup\$ @JackCreasey I know – the frequency at which the CD4047 oscillates is roughly 2/100 Hz :) yep, not saying a better thing would be cheaper! And I fully agree, if OP's driving a CFL from that voltage, increasing Vpp will most likely help. We both agree that OP's claim "this circuit works fine" is purely based on the smaller load's resilience against the less-than-perfect output of the inverter.I think our answers complement each other rather nicely: Yours shows how to,with a bit of understanding of what OP's circuit does,improve it,and mine shows that alternatives exist and reading up is cool :) \$\endgroup\$ – Marcus Müller Dec 1 '16 at 16:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.