TL494 pwm ic , Current sensing.

The boost converter needs a load attached at all times it operates. Each cycle has the inductor sending a burst of current to the output stage. Possibly your resistor network serves as a bleeder load and holds the output voltage down to a manageable level. However if it's only a slight load it might allow the output voltage to reach damaging levels.

you haven't got a proper error amp circuit set up to limit the slew rate and freq response of the comp pin - you need a slower volt loop controller as the boost has an right hand plane zero - which gives the power stage extra delay, lowering gain but also giving phase lag.

research control loops for boost converters - it pays to have an over voltage stop circuit as well in case the feedback loop fails

the current limit ckt might need inverting to fit in with the logic of the TL494

I have made some changes in the circuit.as shown below.



Using approx 50 ohm resistor 10W in output as load.
using 0.1 Ohm , 5w resistor instead of R6 AND R7 for current sensing.

Sharing the scope waveform for
1. TL494 pin (8-11(shorted) to mosfet gate pin 1) in yellow.
2. Mosfet Drain switching on off pulse in blue(mosfet pin 2).


1. I have fixed voltage at the pin 2 of tl494 as 2.5V.
2. Fixed voltage at pin15 of TL494 as 0.5V(assuming max current flow at the load resistor to be 5A . i.e V=IR i.e V = 5Amp x 0.1Ohm = 0.5V.

My concern is that How do i increase the voltage at the output which is still showing only 12.6 V dc.(Not shown in the scope ).

if you turn the pot down ( pot 1 ) and the Vout don't go up - then the current limit must be limiting Vout for fixed load and fixed Imax.

without proper feedback, though, ( which is essentially a modified low pass filter around the error amps ) you'll never have effective control.

have you inverted your pwm by the C & E connections ?

Currently I am able to get :
1. No load voltage : up to 60V.
2. On connecting resistive load of 48ohm resistor at 60V output:
I expect Current flowing in load resistor : V/R = 60 V / 48 ohm = 1.25A.Therefore current sensing resistor (0.1 ohm will give 0.125V) across it.

I have set current limit at 500mv at pin 15. Current sensing is showing 0.125V or 125 mv at pin 16. I believe it should be able to work as the current is within limit.
But the moment i connect the resistive load of 48 ohm . The output voltage shows 12V instead of 60V.I think the switching operation stops on connecting load.
what could be the reason for this .

without load :




With load of resistors the output changes to :

Sharing scope waveform for:


1. No load condition:



2. Load connected 48Ohm resistor.



Yellow is the gate drive .
Blue is the voltage at Drain of mosfet IRFZ44N.

What could be the reason of Converter switching failure at load resistor connection.

You need to read the full data sheet for the 494 which shows a feedback loop in one of the diagrams

Easy peasy said:

You need to read the full data sheet for the 494 which shows a feedback loop in one of the diagrams

Click to expand...

Hello , can you please elaborate what problem do you see in schematic relating to the scope waveforms shown. I want to know what is wrong with the feed back loop.
I am using R13 as 510 ohm and R15 as 51k therefore gain (1+ Rf / Ri) = 101 as specified by datasheet.
attaching current updated circuit as shown below:

Read the complete data sheet !

Easy peasy said:

Read the complete data sheet !

Click to expand...

Dear Friend , My purpose of posting the query here is to get help about something which I am missing out . Kindly either post some suggestion or point out in particular where I should check in datasheet regarding my problem ,any specific page.

For every voltage or current control loop - you need something like this:


to limit the slew rate of the output of Vcomp - but to provide high DC gain so that it will give the correct output regardless of load

typical starting values would be R1 = 100k, Rz2 = 10k, Cz2 = 10nF, CP1 = 220pF

You need to build similar into your control structure using the comp amps inside the IC,

see here:


the outputs of the comp amps go HIGH for LESS power - so you have to get the sense right for the demand voltage and the voltage feed back from the output

for the above if you fed 2v5 to pin 1, then the output would be divided down to 2v5 to match ( say 12V to 2v5 using resistive divider or a pot ) and fed to pin 2 - per the smaller diagram just above - this is called a control loop ...

Similarly for comp 2, for the current limit - although this is generally a much slower loop - e.g. Cz2 = 100nF, so as not to interfere with the volt loop at the point of cross over of control between the two.

I don't understand how the circuit is expected to work at all. You are driving MOSFET gate from C pin which is inverted (active low) output.

FvM said:

I don't understand how the circuit is expected to work at all. You are driving MOSFET gate from C pin which is inverted (active low) output.

Click to expand...

Thanks dear ,

You have pointed the issue correctly in one line. Just yesterday only I added BD139 transistor to be driven by collector of the tl494.The Circuit is alive .

Now that it is working.....I would like to know that if I can generate similar output for mosfet driving using microcontroller.
TL494 is having very refined duty cycle variation .Can i get a similar response if i use the PWM from microcontroller.
If i use PIC16F886 I get 255 steps for duty cycle variation. I.e If i need 12V to 60V dc output then I will get it in 255 steps.

Is there any other chip that i can control from mcu for refined pwm variation. I need Minimum 0.1V step increase in dc output voltage which means
I need minimum 480 steps in pwm to get to 12v to 60v.(60V -12V = 48V ,48V / 480 steps = 0.1V variation per variation of duty cycle ).

jatin12345 said:

Is there any other chip that i can control from mcu for refined pwm variation.

Click to expand...

Indeed almost any switch mode IC can do.
Just "modify" the feedback signal to generate different output voltages.

I´ve even made some kind of tutorial here: https://www.edaboard.com/threads/design-of-pwm-controlled-smps-output-voltage.377215/

Klaus

@jatin12345 - you might share your complete working circuit with all the EDA board members to assist those with similar problems in the future !?

I will share the schematic of a new module pcb which I have made for testing purpose .

I will post it here after testing in 2-3 days.