UC3843 Slightly heat Up

[padoxx]

Is it normal for UC3843 ic in SMPS circuit to slightly heat up while on working operation within some seconds without load?

 

[cupoftea]

No its not. What Vcc are you giving it?
Do you have load on the Vref pin.
What timing components do you use.
What is connected to the gate drive pin?

 

[padoxx]

Is it normal for UC3843 ic in SMPS circuit to slightly heat up while on working operation within some seconds without load.​

timing components 82k and 2A472j

At the gate I have 22R and 10k pull down Resistor'​

I attached the schematic here , the timing is components are 8.2k and 472pf

 

[cupoftea]

You feed vcc with 100k from 48v....that can deliver max of 48/100 milliamps...nowhere near enough for uc3843.
You need to re-do.

 

[scopeprobe]

Is it normal for UC3843 ic in SMPS circuit to slightly heat up while on working operation within some seconds without load.​

timing components 82k and 2A472j

At the gate I have 22R and 10k pull down Resistor'​

I attached the schematic here , the timing is components are 8.2k and 472pf

View attachment 178547

I'd also say it unusual to see significant heat but a slight warming up i would say would be normal. Is this a functioning design?

C2 Looks unusually placed, its function isn't too obvious to me. Its normal to have a small RC on the current sense which is your C29 & R7 but C2 i struggle to identify its purpose.

I'm not familar with the UC3843 but I also find the shorting of the FB to ground quite an interesting idea i've not seen before. This will pull the error amplifier 5V making the COMP pin a fixed current source but eliminates the error amplifier functionaly in the IC to instead be created from the TL431 & opto.

T6 looks to be providing an Auxilary VCC via the 5R1 resistor obviously the higher the voltage on this the more heat you'll be disipating.

 

[cupoftea]

Oops sorry, i didnt see R6, yes, as ScopeProbe says, the VCC is being provided from a bias coil....what is the coil turns and vouts?.......its probably the product of vcc and the 13mA or so of operating current rhat the UC3843 takes...but if you have gone above the internal zener voltage, then it'll heat up

 

[Easy peasy]

If the mosfet gate is damaged / leaky then the gate drive currents from that pin may be higher and cause the chip to heat up.

At 100kHz, warm to the touch is not uncommon.

As above if you go over 34V of the internal zener from the bias supply - you will cook the device - also higher than 16V Vcc will heat the internal gate drivers - and you can damage the mosfet you are trying to drive ....

 

[padoxx]

No its not. What Vcc are you giving it?
Do you have load on the Vref pin.
What timing components do you use.
What is connected to the gate drive pin?

The VCC is giving me 15vdc instead of 8. v that is heating up i guess. VREF is 5v which is normal in this design. I want to know what happen i am seeing 15v at PIN 7

 

[padoxx]

If the mosfet gate is damaged / leaky then the gate drive currents from that pin may be higher and cause the chip to heat up.

At 100kHz, warm to the touch is not uncommon.

As above if you go over 34V of the internal zener from the bias supply - you will cook the device - also higher than 16V Vcc will heat the internal gate drivers - and you can damage the mosfet you are trying to drive ....

I tested the Fet's and is working fine

 

[padoxx]

I'd also say it unusual to see significant heat but a slight warming up i would say would be normal. Is this a functioning design?

C2 Looks unusually placed, its function isn't too obvious to me. Its normal to have a small RC on the current sense which is your C29 & R7 but C2 i struggle to identify its purpose.

I'm not familar with the UC3843 but I also find the shorting of the FB to ground quite an interesting idea i've not seen before. This will pull the error amplifier 5V making the COMP pin a fixed current source but eliminates the error amplifier functionaly in the IC to instead be created from the TL431 & opto.

T6 looks to be providing an Auxilary VCC via the 5R1 resistor obviously the higher the voltage on this the more heat you'll be disipating.

yes, the Auxilary VCC is suppose to be 12v but 15v is there

 

[cupoftea]

So vcc is 15v, and i(operating) is approx 13mA......so theres 200mW.....then add the gate drive current.......what is the gate charge of the nfet? What is f(sw).

Also, how hot are the things around it?...they may transfer heat to it.....

Also, how do you measure temperature?....remember those thermal cams need you to know how to handle their "nuances"......reflected heat, emisivity, etc.

Did you put a sense resistor into the vcc capacitor yet?..and actaully measure the current thats going into the uc3843?......this could be your nxt investigatory step.

Try and get the average current reading...so rc filter a sense resistor voltage.

The chip could be ESD damaged...but who knows......ive had a spate of esd damaged chips so far this year.

 

[scopeprobe]

yes, the Auxilary VCC is suppose to be 12v but 15v is there

Ok, so the the circuit is operating but your vcc is high? Is that correct? This points me back to the voltage of you auxiliary supply once the supply is running which powers the Vcc via R6. Is the output voltage correct and also are you sure Vcc is 12v by design? whats the transformer turns ratio. Given your saying its 15V if the output is 15v i would expect A 1:1 ratio between the windings for a 15v vcc

 

[cupoftea]

You coudl get another UC3843 chip......say a dip one...and solder it into a stripboard.
Then put 15v into its vcc.
Put the timing RC in there too.
And a cap on VREF pin.
Tie FB to ground.
Check its got 5v in its VREF pin.
Leave gate drive open
Tie CS pin to ground

......Then see if it gets hot.

But how are yoiu measuring temperature?...and when you say its hot...are you sure its not heat from some nearby component

UC3843
UC3843

 

[cupoftea]

AYK, txfmr aux coils can drift about, and vout of them varies as per the loading on the main output winding.
15V sounds fine for Vcc for UC3843.......ive used this chip a number of times.