Another IR2153 smps

borysgo2

New member
Hello
Last year I was playing a bit with IR2153 and I have burned a good few mosfets and IC's. Have had also a problem with over current protection (used MCR100 to shunt Ct pin to ground).
This month I decided to start the project from the scratch again.
I found a few key factor to make good and stable circuit for this simple controller.

First issue, IC power supply.
Must be stable at all conditions, startup resistor is not too good solution so I have supplied the controller from the constant current source (voltage independent - extremly important during startup!!), the CCS is turning on after main caps are charged to approx 170V, the ccs is powering up the IR2153 threw stabilised supply unit (MPSA42+zener), after SMPS is ON it is supplying itself from the AUX winding and the supply unit with CCS is turned OFF.

Second issue, not stable startups.
In all IR2153 projects I found the bootstrap capacitor is approx 10uF as in datasheet, I have checked the startups with 10uF, 4,7uF and many lower values, the problem is time to charge it.
At switching frequency approx 50-70kHz the 100nF is really enough to drive the IRF740 mosfet. I left on the board 470nF wchih is working very well. If you go to the IR2110 datasheet You can calculate bootstrap capacitor value yourself more accurate.

Soft-start.
I have used NTC thermistor parallelled with triac, single turn at the main trafo does the job for turning it on. For lower power aplications it should be OK.

Yes I have burned a few mosfets but it was only my fault. I have failed with protection circuit using MCR100, so I decided to use transistor. Now SMPS does not need to be plugged out from the mains to reset. The OCP works fine, it hepls even at the startup with high load connected (or short circuit) to the secondary.

I have no idea if the project is technically OK but it works very well.

primary secondary.jpg
NewFile26.jpg
NewFile5.jpg
DSC01103.JPG
przetworka v2.11.jpg
 

Attachments

  • schematic.pdf
    22.5 KB · Views: 1,132
  • TOP.pdf
    74.7 KB · Views: 676

wally7856

New member
Very impressive design. Can you explain more about what the constant current supply for the ir2153 helps with?
You already have a thermistor why do you also need the triac? BTW cleaver way to turn it on using the transformer.
Also clever using Velcro to hold transformer together.
Current protection circuits are always difficult to make work right. Did you try shorting the output on and off a few times?
 

borysgo2

New member
The upper mpsa42 transistor, voltage between emmiter and base is clamped with 160V zener, the 159,4V / 33kR = 4.8mA of current. With the CCS the supply for IR2153 is very stable, do not forget that the CCS is supplying the stabilised voltage regulator and that regulator is supplying the IC. So the supply for the controller is rock solid (it filters out the maind DC caps ripple), no need to use separate small trafo.


In case You do not want to use AUX winding You must use some more rubust transistors (like MJE340).

I have tryed many times to destroy this SMPS with screwdriver and failed, it still works OK :D
The key factor was protection timing, with wrongly matched RC networks from current sense transformer the SMPS was blowing up instantly !!
 

MicrosiM

Administrator
Staff member
Looks nice to me. I really like the protection section. Smart idea.

Can you post a video of the amps while showing the short circuit tests?

Good luck
 

borysgo2

New member
I have sorted the output with screwdriver a few times, the 3rd one killed the mosfets :(.
I have adjusted the RC a bit (two test with differend values bellow).
The over load protection works fine.

test with 220nF x2
[video=youtube;H71jUYXba2w]https://www.youtube.com/watch?v=H71jUYXba2w&feature=youtu.be[/video]

and second test with 470nF x2
[video=youtube;yNA6wqN65Zc]https://www.youtube.com/watch?v=yNA6wqN65Zc&feature=youtu.be[/video]
 

mone

New member
Good job!
One question I found a diagram
"Current sense transformer 1:50" - "TR_SENSE"

Can you describe what it is? What power do we have? 600W?
 

borysgo2

New member
The current sense trafo ratio is 1:50, one turn of primary and 50 of secondary (kynar wire). If U have 1.5 turn of primary just multiply it by 50 for secondary winding.
 

borysgo2

New member
Jagd.
The problem for me was to wind a pure one turn, I had to recalculate my current sense transformer. I taught that I have ''one turn'' primary and I have wunded 50 turns of secondary after I have measured the currents it turned out that the current ratio is not 1:50 but 1:30 !! So half or a quarter of turn of the primary was making a big difference.


one turnJP.jpg


And this is the current sense trafo tests I have done (i have used audio amplifier and resistive dummy load). There is a mystake on the screanshoot, I taught the trafo is 1:50.
12V /4R=3A of curren amplitude at the primary.
2.1V/22R=0.095A current amplitude at secondary.

ratio 3A/0.095A=31.5 times

Przekladnik pradowy sprawdzenie JP.jpg
 
Last edited:

Jagd.Panther

New member

borysgo2

New member
Jagd.Panther

Big thanks for explenation!!
As far I remember the core was at the secondary of ATX psu unit with single winding so I persume it was a small output inductor.
I will prepare another current sense transformer for some education purposes.
This is my first smps and there is so many things I do not know. Learning process takes some time.


Regards
 

Jagd.Panther

New member
As far I remember the core was at the secondary of ATX psu unit with single winding so I persume it was a small output inductor
Then it's very likely to be a saturable reactor from magamp for 3v3 line.How does it look? does it a have a plastic outer shell made of two pieces? Solid green epoxy color? Solid, gray, plastic-like coating?
 

borysgo2

New member
The core is coated with black plastic. I was checking it up to 5A peak current and it was not saturating (didnt look like).
Should I make another one with green core larger in size ??

Thanks
 
Top