Generating clock pulse or PWM with TL494

[hkBattousai]

I'm trying to figure out how TL494 (datasheet) works. I found the datasheet a little confusing. I searched for some online examples, but still I can't make sure if I correctly set up my circuit or not. Please check my design and make corrections or suggestion if necessary.



Is this design correct?

I want to adjust output duty cycle. Assume that I want to do this by hand. Can I do it by simply connecting a potentiometer to the FB pin? Should I use the FB pin or the DTC pin for this? The current source at the input of PWM comparator gives me worries. If there is a 0.7mA current source there, I have to use a very low valued potentiometer for precision, but which also means waste of power. On the other hand, the DTC pin looks like to be specialized for adjusting dead time. Is it OK to mess up with it for changing the pulse width?

Is the generated PWM at the circuit output (i.e.; the A and B inputs of the H-Bridge section) complemented in this circuit?

The datasheet says that, this IC contains only 46 transistors. How is that possible? Even a single opamp IC contains more than 10 transistors. Looking at all those voltage references, logic gates, flip-flop, current source, opamps, oscillator and diodes, I don't understand how they did this with only 46 transistors. Can you please make an explanation?

[c4757p]

The datasheet says that, this IC contains only 46 transistors. How is that possible? Even a single opamp IC contains more than 10 transistors. Looking at all those voltage references, logic gates, flip-flop, current source, opamps, oscillator and diodes, I don't understand how they did this with only 46 transistors. Can you please make an explanation?

Sorry for not directly answering your question - need more coffee! Or some experience with the chip would be nice; I haven't used it before.

But as for this one - that's an interesting question. Check out this TI application note: Designing Switching Voltage Regulators With the TL494. They've got what they claim to be internal schematics for everything. At a quick count: 19 in the reference, 14 in the oscillator, 13 in the dead time/PWM comparator (one combined circuit), 12 per error amplifier, and I stopped counting there! Already a total of 70... Note that TI doesn't appear to make any claims of transistor count in their datasheet. Perhaps theirs is implemented differently? I'd love to see a full internal schematic.

Also note that you don't need nearly as many transistors to implement most of these circuits if they're not "general purpose". For some very specific applications of an "op amp" in a block diagram, you could get away with as few as two transistors. A rough comparator could simply compare against a base-emitter threshold, meaning one transistor! For inverting configurations (where good CMRR isn't needed, as one input is permanently tied to a fixed level), five transistors (two in a differential amplifier, two in a current mirror, and one common-emitter amplifier for the output) can make quite a good op amp.

[diyaudio]

I'm trying to figure out how TL494 (datasheet) works. I found the datasheet a little confusing. I searched for some online examples, but still I can't make sure if I correctly set up my circuit or not. Please check my design and make corrections or suggestion if necessary.



Is this design correct?

I want to adjust output duty cycle. Assume that I want to do this by hand. Can I do it by simply connecting a potentiometer to the FB pin? Should I use the FB pin or the DTC pin for this? The current source at the input of PWM comparator gives me worries. If there is a 0.7mA current source there, I have to use a very low valued potentiometer for precision, but which also means waste of power. On the other hand, the DTC pin looks like to be specialized for adjusting dead time. Is it OK to mess up with it for changing the pulse width?

Is the generated PWM at the circuit output (i.e.; the A and B inputs of the H-Bridge section) complemented in this circuit?

The datasheet says that, this IC contains only 46 transistors. How is that possible? Even a single opamp IC contains more than 10 transistors. Looking at all those voltage references, logic gates, flip-flop, current source, opamps, oscillator and diodes, I don't understand how they did this with only 46 transistors. Can you please make an explanation?

Okay breath take it sloooooooowly... read the data sheet and get an application note. see TL494 application note. (very useful)

The TL 494 is capable of adjusting its duty cycle 3 places, 1 duty cycle adjustment for each of the 2X error amplifiers and its duty cycle pin, it also allows up to 98% duty cycle with 2% DT deadtime. The PIN with the lowest duty cycle will dominate the timing. 


The IC can operate in single ended OR push-pull mode up to 400KHz and 200Khz respectively.
 

As for the internals component count I think you have more important things to understand.