Why switching ?

Status
Not open for further replies.

grandestlama

New Member
I have gone a long way with my UPS design and construction. Often times people aim at producing a "sine-like" wave form.

I was wondering why not produce a sinusoidal oscillation and amplify with BJT's or MOSFETs and use this to drive the step up transformer primary ? Instead, the transistors are usually biased to operate as switches (not as amplifiers). Is this because a digital pulse is used as the source of the oscillation ? If yes then why is nobody replacing the NE555 astables or SG3524 PWM osclilators with sinusoidal oscillators and using the transistors in the linear region for amplification instead of switching ?

Please if you know why just post a reply no mater how amateurish my question is !

THANKS
 
I think the major reason is efficiency and the the size of components, mostly the step up transformer can be a lot smaller. Most of the new ones use step waveform that simulates a sine wave. I think they are nearly 90% efficient.
But I know exactly what you are talking about with the use of sine waves.
I had a receiving circuit that needed about 30 volts DC to operate. My power source was 12VDC. Several attempts to build a switching voltage pumps to get 30VDC were unsucessful do to the noise generated by the switching circuits. I finally generated a sine wave at about 800Hz, fed it to a 15 watt audio amplifier and used an audio transformer to step the voltage up to around 30VAC, rectified it and had my DC voltage. Most audio amplifiers may get 50% efficient at best. If you use a pair of transistors in push-pull configuration at class B operation you get 50% efficiency.
 
And all that inefficiency also means that the power is not only used, but turned into a crapload of heat at the transistors.

Few devices NEED a sinusoidal waveform. Computers, most of the stuff the UPS will power, use a switching power supply that only rectifies out the high voltage and filters it before switching it at high frequency through its own transformer. In this case, the modified square wave is actually better, since it charges the cap with equal potential at all times, rather than only charging the cap at the peaks of a sine wave.
 
But many elecronic items rely on charging their cap to the much higher peak voltage of a sine-wave, and therefore don't work when supplied by a square-wave inverter. Also, many ballasts would buzz like crazy.

Can you imagine how many output transistors or Mosfets would be needed, and how huge their heatsink would be in a linear sine-wave inverter?
Even audio amps are using switching devices with Class-D.

Can anyone post a link for a Class-D sine-wave inverter?
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…