Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

inverter not working using cd 4047.

Status
Not open for further replies.

mugheesnawaz

Member level 1
Joined
May 1, 2013
Messages
39
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Activity points
1,605
i am using the attached image cct for square wave inverter but i dont get any output.i have the simulative model of cd 4047 and design file of proteus as well attached kindly check it and help i have to submit it soon.

this is the cct diagram
Low-Power-Inverter-Circuit-12V-DC-to-230V-or-110V-AC-Diagram-using-CD4047-and-IRFZ44-Power-MOSFE.png


and this is the design proteus file and the simulation model for cd 4047 which can be added to proteus>Models folder for it to work.thanking in advance

View attachment Inverter.rar
 

I can't use the Proteus file but the schematic looks OK except for lacking a capacitor across the incoming supply. Do not expect good efficiency or a clean output signal with this very crude design.

Brian.
 

Dear Mr M..z,

Let me share some decades of experience.
There is such a thing as (Only) I call inverse impedance gain in switching circuits. Like current gain hFE is typically rated for Ic/Ib for Vce(sat) =10 typ, 5 old, 50 best-in-class.

Similar performance is required Rgs/Rds for fast switching of Ciss which is uniform, more or less with constant RdsOn*Ciss, where best is lower at same rated Vds max and worse is higher but fcn(Vds.max)

CD4047 is old HC4xxx 15V CMOS which has an Zout or RdsOn of 250Ω @12V and drops to 200Ω at 15V. Add this to 100Ω and you get an excessively high (>>50) R ratio.

Fix that with intermediate RdsOn buffer for better transition power loss.
 

I think the present problem is more of the trivial kind. It's that the CD4047 simply isn't oscillating in this Proteus simulation.
 

I think the present problem is more of the trivial kind. It's that the CD4047 simply isn't oscillating in this Proteus simulation.
I love examples like this one that show clearly that simulators are a waste of time.
 

My other advice is ignore sites like this full of student (rookie) examples of design. Use OEM designs like from professional demo cards that work, like TI, Analog Devices, Toshiba, Hitachi, Intersil, etc etc.

Study these instead.

Can you see an obvious error in this one, from your same example site?

http://www.circuittrue.com/a-compac...uit-wich-optimised-for-cfl-loads-uses-cd4047b

I think there may be some setup error in your Sim., which I wont bother since many other problems ignored in your design like load regulation, OVP, OCP, etc which does not exist.

- - - Updated - - -

But Simulators are not a waste of time, but inexperience takes time to learn to use existing working designs.

It is better to find a modern product that works very well, get schematic and test it and understand what each component does and how it can fail or why it works well.

Reverse Engineering is much better for learning than trying to copy unproven designs with NO SPECs.. always start with a spec. Unless you are just getting familiar with Sim.

Best tool is the one you learn to the best.
 

If you build the circuit it will work. The output will be a squarewave that some modern electronic products cannot use and the output voltage is not regulated. Also no overload protection.
The CD4047 IC should be fed 12V through a series 100 ohm resistor with a 16V zener diode and capacitor to ground to prevent supply voltage spikes from zapping the CD4047 like this:
 

Attachments

  • CD4047 inverter oscillator.png
    CD4047 inverter oscillator.png
    95.6 KB · Views: 448

what makes you think you will get 230Vac out?

Is that RMS p-p or average and under what load ?

Without specs, without proper design, nothing

This circuit will just be an alternating current spike with no load giving obvious poor results for intended output.

this is a rookie design but maybe useful for teaching...


a bad design.
 

Yes I agree, this is not suitable for anything useful. To do something useful it would have to generate a sine wave based on PWM.

Rob.
 

About 10 years ago I helped modify a squarewave inverter that used the CD4047 oscillator, driver transistors and common 2N3055 transistors in parallel at its output. It was (and probably still is) being used by people in The Philippines to power their incandescent and tubular fluorescent lights and TVs. The batteries for the inverters were replaced every week with freshly charged batteries. Modern Mosfets were not available there.
 

Hi,

The error description isn't very informative...

I agree with the posts, especially #6, #8, #9.

Klaus
 

It is obvious that the simulation software knows nothing about the oscillator in the CD4047 IC.
The circuit produces a squarewave that is 240V peak-to peak which produces exactly the same power to a heater or incandescent light bulb as a sinewave that is 240V RMS. Many electronic products that rely on the higher peak voltage of the sinewave and electronic motor speed controls probably will not work from the squarewave.
 

It is obvious that the simulation software knows nothing about the oscillator in the CD4047 IC.
The circuit produces a squarewave that is 240V peak-to peak which produces exactly the same power to a heater or incandescent light bulb as a sinewave that is 240V RMS. Many electronic products that rely on the higher peak voltage of the sinewave and electronic motor speed controls probably will not work from the squarewave.

Prove it.

... as transformer is merely an Inductor with no load, and V is full of harmonics not kind to iron core losses.

So output is anything but a squarewave, but rather the integral of the primary current \N turns ratio. * RL....i.e. triangle voltage at max loads and high voltage square wave with tiny load that increases with R Thus bulbs with 10% to 1 R will start soft and regulate voltage.
 

Hi,

Does that include post #5 ?

:)

Although 5 is not included in 6, 8, 9..
I agree with you.

With simple low frequency, low current circuits a simulation gives good results.

But it is very important to understand the critical points of a circuit and to know how to route a good pcb layout.

I once saw a simple step down DC/DC circuit. The designer told me it doesn't work like expected and i should try to find out why.
It was pretty obvious to me: No ground plane and the whole switched current path was routed "around" the pcb with a trace length of about 400mm. ...

Klaus

- - - Updated - - -

Hi,

... as transformer is merely an Inductor with no load, and V is full of harmonics not kind to iron core losses.

So output is anything but a squarewave, but rather the integral of the primary current \N turns ratio. * RL....i.e. triangle voltage

I only agree in parts with that.
For sure the iron core is not good for the high dU/dt of a square wave signal. Giving distortions and loss
And i expect ringing during dead time (but it is not defined here)
But as long as ther is no DC component at the input and there is no core saturation (i don't expect with the given values) the output voltage waveform should be fairly good square wave.
But the primary "current" will be of triangle shape.

Klaus
 
  • Like
Reactions: Vbase

    Vbase

    Points: 2
    Helpful Answer Positive Rating
Hi,

With simple low frequency, low current circuits a simulation gives good results.

But it is very important to understand the critical points of a circuit and to know how to route a good pcb layout.

I once saw a simple step down DC/DC circuit. The designer told me it doesn't work like expected and i should try to find out why.
It was pretty obvious to me: No ground plane and the whole switched current path was routed "around" the pcb with a trace length of about 400mm. ...

Klaus

Very well put, Klaus.
Of course what you said and the example you gave mean that experience with real components is more important.
 

Prove it.

... as transformer is merely an Inductor with no load, and V is full of harmonics not kind to iron core losses.

So output is anything but a squarewave, but rather the integral of the primary current \N turns ratio. * RL....i.e. triangle voltage at max loads and high voltage square wave with tiny load that increases with R Thus bulbs with 10% to 1 R will start soft and regulate voltage.
Although I have never seen a squarewave inverter, here is a similar one and the 'scope shows a squarewave output:
 

Attachments

  • 110W squarewave inverter.png
    110W squarewave inverter.png
    159.9 KB · Views: 146

That's a full complementary bridge. Not an alternating half bridge with centre tap. But almost the same as I have modelled it. 1:10 turns ratio 0.1H, Rs,= 0.1 Ohm primary loop.

Note the Japan power supply uses a Bulb for load which improves performance only for steady state graphs they made as the 0.1~1 x R PTC resistance of tungsten, makes it a almost a constant current load from 0~50% Vrated.

( which makes the specs look better. :O

But other than the DC offset saturating the transformer more , it does something you expect and a bunch of things you ought to expect like remenance induced saturation currents, ringing.

Try one load at a time and note that if you break before making the next lighter load, it will create a huge transient V=LdI/dt with Lp set to 0.1H but you can change it to 10H if you want.
inverter 120V.jpg

This is an interactive Java design I constructed for show N tell
 

Hi SunnySkyguy.

Yes, by adding a series inductance in the primary path you get the shown results.

A transformer acts like a huge inductance. You see that this inductance limits the current. Especially with no load.
Lets say a 100W transformer with no load draws 50mA at 240V, 50Hz. The power loss is maybe 3W.
3W are active power. And it causes a current of 3W / 240V = 12.5mA.
This means 48,4mA are reactive. Caused by the primary transformer windings and the core.
This is an xL of 4.9kOhms an an inductance of 15.8H.
But this is an parallel inductance. Parallel to an ideal voltage to voltage transformer.
And for sure there is not only this huge parallel inductanc, but also a series inductance.
The parallel inductance is inthe range of nH.

In your simulation there is an serial inductance of 100mH (as far as i can see)
This is too huge. It is a serial xL of 31.4 Ohms.

If you now connect a lod to the transformer, then in the serial inductane will be a lot of voltage drop casing the secondarv voltage to drop. You can see this in your simulation. The output voltage is not 120V anymore. The triangle part of the output voltage is caused by the too huge serial inductance.

And for sure this inductance combined with the capacitor forms a resonant.The capacitance of the secondary side act as a 10^2 as large capacitor at the primary side. 10nF --> 1uF. F_res = 503Hz.

Usually transformers are made to have low series inductivity. This is good for voltage stability under load. This is true for small signal transformers like ethernet, medium small like audio transformers in 100V (here especially voltage distortion has to be avoided) up to huge grid line transformers. Excepions are welding transformers, wher one needs high current for arc ignition, but as soon as current flows, the voltage decreases by the wanted serial inductance.

Klaus
 

arc welders have special transformers for constant current limit with air gap.

Did anyone look at the kV spikes when my Sim load turns off? or the resonance with a capacitive load?

Javas sim may consume 100% of equivalent of 1 core,, ie. 25% of CPU with quad core., 50% of CPU with dual core, that's normal.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top