optocoupler signal

[fifahaywire]

Hi, i'm new here and currently carrying on a project on buck converter. and i'm having trouble. i use IGBT ( https://www.alliedelec.com/search/productdetail.aspx?SKU=70017065 ) as the switch but there seems to be a problem with the signal. my driving circuit is as attached below. i'm using TL494 and HCPL 3140. i got the output from the PWM controller as i want it to be but when i connect it to the optocoupler, i got a straight DC. what am i doing wrong? please help.

Thanks

 

[MrAl]

Hi,


Is that a 12k resistor on the input? Isnt that a little high? Are you getting enough input current to drive the internal LED? You need a certain minimum current to get the device to switch and it doesnt look like you have that, although you dont show the drive voltage.

 

[ronsimpson]

Do you have the IGBT-E connected to the opto? I can see the IGBT-B connection.

 

[Boncuk]

Hi fifahaywire,

you really haven't left out much to do wrong with your circuit.

The TL494 connections are wrong by about 90% according to TI datasheet page 6.

I suppose you chose that circuit for your design.

An optocoupler, even if overkill for your application requires a certain Vf and If for the on chip IR LED. The one on your chip requires typically 1.5V at If=10mA. With a 12KΩ resistor it won't do anything.

Pin5 of the optocoupler should be connected to ground. Before doing that you should remove the Zener diode between pin8 and pin5 to avoid the magic cloud of black smoke.

Boncuk

 

[fifahaywire]

Hi,


Is that a 12k resistor on the input? Isnt that a little high? Are you getting enough input current to drive the internal LED? You need a certain minimum current to get the device to switch and it doesnt look like you have that, although you dont show the drive voltage.

hi MrAl, according to the datasheet https://www.electro-tech-online.com/custompdfs/2012/02/0900766b804e419d.pdf max If is 25mA. and i did try with 10Ω but it doesn't seem to work too. what do you think?

Do you have the IGBT-E connected to the opto? I can see the IGBT-B connection.

yes, i am using an IGBT at the end of the circuit for the switching for my converter

Hi fifahaywire,

you really haven't left out much to do wrong with your circuit.

The TL494 connections are wrong by about 90% according to TI datasheet page 6.

I suppose you chose that circuit for your design.

An optocoupler, even if overkill for your application requires a certain Vf and If for the on chip IR LED. The one on your chip requires typically 1.5V at If=10mA. With a 12KΩ resistor it won't do anything.

Pin5 of the optocoupler should be connected to ground. Before doing that you should remove the Zener diode between pin8 and pin5 to avoid the magic cloud of black smoke.

Boncuk

according to the datasheet i need If 25mA. i did try 10Ω.
do u mean that i don't really need the zener at all? hmm..

 

[alec_t]

i did try with 10Ω but it doesn't seem to work too. what do you think?

I think that might have caused too much current to flow and fried the opto-diode. Did you see the magic smoke?

 

[fifahaywire]

I think that might have caused too much current to flow and fried the opto-diode. Did you see the magic smoke?

maybe but i didn't see the magic smoke. probably because i normally let the power supplies on not more than 10 seconds.
i'll try changing to a higher resistor value.

 

[ADWSystems]

Do you have a meter with an LED or diode tester? The internal LED could have blown in 10 milliseconds after the power came on. There's also a chance the LED circuit blew open, check the current flow through the LED with a steady voltage and resistor to provide 25mA (200ohms @ 5 volts).

 

[duffy]

and i did try with 10Ω but it doesn't seem to work too. what do you think?

First, you need to ground pin 5 of the HCPL-3140, like Boncuk said. Your opto drive has no ground! That's the most glaring error.

Second, that's a bad configuration with the zener and no current limiting resistor and no cap. What are you trying to do there? Limit the peak voltage to the IGBT? If so you need a resistor from Vcc to pin 8, not a diode, -and either way you need a .1μf cap from pin 8 to ground, that driver pulls high peak current pulses.

 

[fifahaywire]

Do you have a meter with an LED or diode tester? The internal LED could have blown in 10 milliseconds after the power came on. There's also a chance the LED circuit blew open, check the current flow through the LED with a steady voltage and resistor to provide 25mA (200ohms @ 5 volts).

i did check. and you were right. i changed the IC and my circuit works fine now

First, you need to ground pin 5 of the HCPL-3140, like Boncuk said. Your opto drive has no ground! That's the most glaring error.

Second, that's a bad configuration with the zener and no current limiting resistor and no cap. What are you trying to do there? Limit the peak voltage to the IGBT? If so you need a resistor from Vcc to pin 8, not a diode, -and either way you need a .1μf cap from pin 8 to ground, that driver pulls high peak current pulses.

omg. you were right. there's no ground. what was i thinking?
so, i corrected the circuit (like u said), eliminate the zener and i finally got the output! thanks!


Thanks to everyone who replied! you guys are so much help. i can't thank you guys enough