Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Flyback SMPS power MOSFET getting hot? uc3843 pwm converter

Status
Not open for further replies.

codemaster11

Advanced Member level 4
Joined
Sep 2, 2019
Messages
115
Helped
9
Reputation
18
Reaction score
7
Trophy points
18
Activity points
1,218
Photo0327.jpg

today i built this flyback smps transformer for this circuit as shown above using PQ3230 core

with Nprim = 24T (swg 26 x 4) , Nsec = 4T(swg 23 x 4) & Naux = 5T(swg 26 x 1). but the smps power

FET is getting hot even without load as shown in below image.

Photo0329.jpg

the circuit has Vout = 16vdc & Iout = 12amp.

how to stop heating of the smps power mosfet?
 

this power mosfet getting extremely hot even without any load connected.

the box of the power supply act as heat sink not shown here.

but mostly battery charger or power supply is not getting hot without any load.

this may be due to high flux density taking the core into saturation.

i use power esim software for determining the numbers of turns it show Bmax greater than 0.2500 T
 

i change the frequency of uc3843 oscillator by putting 10K variable resistor on the Rt/Ct pin.

still it does not work and the power mosfet still getting hot without load. i think this is not the problem

with frequency. it may be due to saturation caused by less numbers of primary turns. because i also have

a calculator for determining the numbers of flyback xfromer turns which calculate Nprim = 40 T(swg 26)
 

Hi,

still it does not work
You did not mention this before.
So "what" does not work as expected? Give a detailed error description: What do you expect, what do you see/measure instead, how and where did you measure this...

Maybe there is an error in your schematic, maybe the PCB layout is bad, maybe you used unsuitable devices....how can we know?
You don't give much information that we can validate...

Klaus
 

without loading why mosfet heat up??? and when loaded mosfet get hot extremely. i 'm using 20NM60 power mosfet

with this flyback battery charger circuit. it is a readymade chines switch mode battery charger circuit?

one thing i noticed with this circuit is the 2vdc pulse generated at the gate pin of the mosfet which is low

for the mosfet to fully turn on. i want to stop mosfet heating.
 

2v is too low for a fet to turn on properly.
What is your vin.
What is the gate series resistor value.
Can you show scoep shot of gate voltage on no load.
Can you show scope shot of source sense res voltage?

- - - Updated - - -

Is there a primary clamp/snubber?
What is fet drain voltagew (scopes please)
 

i measured the pulse voltage of uc3843 in mu circuit today. it is 0.7v not 2v.
Vcc of the uc3843 at the 47uf capacitor is 21vdc but pulse voltage = 0.7v

primary clamp/snubber include 47K-2W resistor, C = 103pf & and fr207 fast recovery diode.
drain voltage is about 190v to 200vdc.


but in properly working power supply. this gate voltage Vgs = 2.5vdc

Vcc of uc3843 was about 16.5vdc. & voltage at the FET drain to source was about 300 to 330vdc.
 

Hi,

pulse voltage = 0.7v
this gate voltage Vgs = 2.5vdc
This makes me assume you used a DVM to measure this voltages.

But a DVM value is average value, which is not useful.
Measure it using a scope. You need to differ between V_GS when ON and OFF.

Klaus
 

yes i measured it with DVM, but unfortunately i have no scope. i didn't understand why Vgs = 0.7v when Vcc = 21vdc.

the gate series resistor at the uc3843 pulse is 22R/1w. and the Bmax for the transformer i wind--> Bmax = 0.1400T.

but there is no air gap in transformer due to the total numbers of turns.
 

Hi,

yes i measured it with DVM, but unfortunately i have no scope. i didn't understand why Vgs = 0.7v when Vcc = 21vdc.
Then just tell what you expect.

As already said: a DVM (in DC mode) measures "average". Obviously the output voltage is less than VCC and of variable duty cycle.

Klaus
 

but there is no air gap in transformer due to the total numbers of turns.
This explains a lot. Operating a flyback transformer with no air gap is a safe way to saturate the core.

Without an oscilloscope, you can't debug circuit operation. There's a certain chance that it works from the start, but this requires a trustworthy circuit design. You have ruined it by using wrong transformer parameters.
 

the figure above show an xformer have air gap & numbers of turns mentioned above.
and the same problem occurs from the beginning that mosfet heating up. i want to stop heating of the power
mosfet. but how a transformer can be designed to solve this problem? stop mosfet heating.
the mosfet is not heating a little but heat up extremely that no one can touch it when load connected.
and if the circuit continuously operate for a minute the mosfet & the ic will short as expected. the above circuit
works correctly with a readymade transformer. but not with my own transformer. the problem is with the xformer.
the circuit has 14.2V dc output & 3 amp dc current in series. it's right but the only problems is the mosfet heating.
Vcc of the uc3843 of a properly working same circuit as mine is 16.5vdc with DVM & Vgs = 2.5vdc. but mine is Vcc = 21vdc &
Vgs = 0.7vdc
 

flyback Tx needs to have correct air gap, it is a storage and release transformer - all the energy per pulse is stored in the air gap ...
 

Capture.JPG

this air gap of transformer as shown in the image ?
 

you'll find there is the same gap in the centre leg of the core too - sometimes only in the centre limb ...

- - - Updated - - -

if you wind your own transformer for flyback - you need to know the Lpri you want, the proper turns ratio - which sets the flyback volts on the pri - and the gap to keep the peak flux < 0.15 T at peak design current in the pri
 

Hi,

this air gap of transformer as shown in the image ?
It's not clear where your arrows point to..
The "air gap" is between the two halves of the ferrite core. (With air gap, the two ferrite pieces don't touch each other).
(It's not the gap beween coil and ferrite)

Klaus
 

Capture1.JPG

i have PQ3230 core with fixed air gap of about 0.979mm. for this air gap Bmax = 0.2600T which is much higher.

i have no tool to change the air gap. the frequency fs = 77 KHz, core area Ae = 1.42sq-cm & Vin(min) = 141vdc.

primary inductance Lprim about 135.89uH. the Vout = 14.2 Vdc & auxiliary voltage = 18Vdc comes out correct according to my calculations.

i tried different numbers of turns with different frequencies Vout, Vaux & Output current all right. but every times the power mosfet

starts heating without any load & with load if runs for a minute the mosfet will damage.
 

if you wind a simple sec over primary the leakage inductance will be quite high - thus the mosfet has to absorb a spike every time it turns off - this leads to heating.

please show detail of the turn off snubber for this flyback - if you have none - your mosfet will always overheat - due to its clamping above Vds max
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top