Help needed in push-pull circuit for Inveter.

[audioguru]

I already told you that a cheap ordinary multimeter cannot measure the blocky modified sinewave waveform voltage accurately because it is calibrated to measure a real sine-wave from the mains. My expensive Fluke multimeter can measure "true-RMS" so it might measure the voltage accurately.

 

[koolguy]

At least after changing the 12 0 12V to 9 0 9V transformer should raise it output voltages...
and i have connected the Oscilloscope i notice:
1. when the BNC cable was connected to transformer input side i.e. 12 0V it was like spikes not proper wave,why?
2. The oscilloscope can measure max. 5V so, i can't see the proper voltage(wave form) on screen of CRO.
so, please tell what to do??

 

[MrAl]

Hi,

For one, use a voltage divider to divide down the signal for the scope. If you have 20v signal use a 4 to 1 resistive voltage divider to get down to 5v max. This works ok for signals that are not sourced from super high impedances. You could probably get away with 3k and 1k for a 4 to 1 divider. If you need higher impedance, try 30k and 10k.

 

[audioguru]

The spikes on the 12VDC power are caused by the series inductance of the cable from the power supply to the inverter circuit (the circuit should have a 100uF low frequency supply bypass capacitor and a 0.1uF ceramic high frequency supply bypass capacitor).

The modified sinewave circuit does not produce "shoot through" like a simple square-wave circuit.

 

[koolguy]

OK, can i measure current of inverter output in Oscilloscope..

 

[MrAl]

Hi,

Yes.
To do it right, you need a current probe, but they are a little expensive so what most people do is take a low value resistor like 0.1 ohms and use the scope to look across the resistor and then use Ohm's Law to calculate the current knowing the voltage on the scope and the resistance value:
I=E/R
So with R=0.1 ohms if you read 0.1 volts on the scope that means it is 1 amp.
Sometimes you have to use a lower value resistor so it doesnt overheat or drop too much voltage, but the signal starts to get a little noisy as you get into lower R values, but you can usually still see what is noise and what is signal with a little practice.
It also depends what sensitivity your scope input channel has. If you can read down to 0.010 volts that is good, but 0.001 volts better.
Non inductive resistors are the best for this kind of thing, or actual current shunts you can buy that are not too expensive.

 

[koolguy]

I have never seen .1ohms resistance.

 

[audioguru]

I have never seen .1ohms resistance.

Audio amplifiers use it as an emitter resistor for complementary output transistors (NPN and PNP emitter-followers).

 

[koolguy]

Hi,

I have changed 12 0 12V 500mA transformer to 30A max transformer of UPS which has 4-5 tapping at output for AVR (external) and the circuit is working fine i have tested 200Watt BULB and 300Watt Drill machine running.
the amp meter was showing more than ~15A at heavy load from 12V 7AH battery of UPS. so i guess the output power tested till now is 24Vx15A =360VA.
The no-load current taken by Transformer is 900mA. as there is no voltage regulation circuit so voltage fall down at load.
Someone suggest me to use 3524 PWM modulator. please help me with suggest for further improvement...

 

[MrAl]

Hi,

Gee sorry i didnt see your other reply until just now. The 0.1 ohm resistor is a very common value that can be found almost anywhere they sell resistors. It's a small value to keep the voltage drop very low during testing. It's also a common value used for current sensing in permanent applications like UPS systems or battery chargers (might be even lower like 0.05 or even 0.01 ohms).

What is wrong with what you are doing now, what are you trying to do?
The chip you are talking about is a standard PWM modulator made for power supply applications. There is a lot of info in the data sheet.
TI makes chips TL494, TL497, TL594, TL598, etc., if i rem right those numbers, for switching power supplies, and there are other parts close to that number that have slightly different features. You have to look at them and decide what approach you want to use. Many of the PC power supplies use something like the TL494.

 

[koolguy]

TI makes chips TL494, TL497, TL594, TL598, etc., if i rem right those numbers, for switching power supplies, and there are other parts close to that number that have slightly different features. You have to look at them and decide what approach you want to use. Many of the PC power supplies use something like the TL494.

Is there any difference in SG3524 and TL494 working, and which will work best...??

 

[koolguy]

In the datasheet of sg3524 the PWM freq range is 100-400khz but i need 50Hz, AG reply please...

 

[MrAl]

Hello again,

You really need to look up the part numbers and get the data sheets and go over them yourself so you can start to see what you want and what you dont want. The chips are similar, but they are sometimes push pull and sometimes just single ended, and sometimes they can drive MOSFETs and sometimes made for bipolars.
Most of them have built in dead time as well, but watch out for the min dead time because it may be longer than you want at a given frequency.
Obviously if the data sheet specifies a given frequency range and you intend to work outside that range then it's not a good idea unless you know what to modify to make it work. Generally it's easier to get low frequencies than higher frequencies.

Did you say 50Hz? Wow that's low. Most chips are made for 500Hz or above, but the TL494 will work down much lower. The SG chip is not recommended for frequencies under 500Hz.

 

[audioguru]

In the datasheet of sg3524 the PWM freq range is 100-400khz but i need 50Hz, AG reply please...

I don't chat about inverter circuits anymore because you can buy excellent ones for much less cost than you can make them.

 

[koolguy]

So, SG3524 can't be used?
will TL494 will work??

 

[MrAl]

Hello,

Yes that's what it looks like, unless they have an improved version of the SG chip. This is because the graph for the timing components only allows going down to about 500Hz and also the onboard amplifier may not work well at too low of a frequency either. It could be because the chip can not deal with the high capacitance and/or high resistance required to get down to low frequencies like that.

It looks like the TL chips would be better. There are a couple of them that should work maybe you should check them all out and see if they look like what you need.

 

[koolguy]

i found lot of ckt in GOOGLE

 

[MrAl]

Hello again,


Oh yes very interesting. You may want to check what values of timing components are being used to get that low. They may be taxing the chip too much. Im not sure though if they improved this chip over the years so that bigger caps can be used, but larger resistances could lead to too much noise in the circuit so you'll have to be careful about that.

[LATER]
I checked out one of the schematics now and i see that they way they are getting to the lower frequency is by using a higher value resistor than is specified on the data sheet. It probable works, but you should check for noise in the waveform and see if it works out ok. The highest value they spec on the data sheet is 100k, and they are using 100k fixed resistor (which takes us down to around 200Hz) plus another variable resistor to add to the resistance. The extra resistance can be as high as 100k making the total resistance twice what the data sheet recommends. As long as it works i guess it should be ok, but again watch for noise.
I would not try raising the capacitor value higher though as that could tax the chip itself too much. It's up to you of course what you want to do here, but there are other chips that are spec'd for lower frequencies right on the data sheet. The data sheet i looked at was from TI.
The pulse width modulation mechanism depends on the rising capacitor voltage which comes from the series resistance, so one of the side effects of a very large resistance is going to be increased jitter in the output pulse widths. This may or may not be objectionable in your application.

 

[koolguy]

The SG3524 and TL494 ICs are used in pure sine wave inverters (not in simple modified sine wave inverters). Their oscillator operates at a high frequency, maybe 100kHz and is Pulse-Width-Modulated with 50Hz or 60Hz.

Because the oscillator frequency is high then a small ferrite stepup transformer can be used.

AG said this on another forum...
So, pls tell what to do or should we start it using pic16f877a??

 

[audioguru]

If you program a PIC so it makes PWM then it can produce a pure sinewave that drives logic-level Mosfets. Additional parts (they are inside the SG3524 and TL494) can vary the PWM for voltage regulation of the sinewave.
There are plenty of circuits in Google and most are from your country (where the electricity might have poor reliability).