tebci
New member
I sent you a private message, please respond.
1kW smps project (based on MicrosiM design)
- Thread starter ludo3232
- Start date
I sent you a private message, please respond.
badboy_6120
Member
Hello Jagd.Panther
I did all your tips about the sudden voltage drop and the only thing that worked is {{connecting heatsink to the low voltage ground (Middle leg of lm7815 regulator)+ connecting an oscilloscope probe to the gate of Q1 and the ground clip to the low voltage ground}} at the same time ( If I remove one of connection above the problem would come back)
sometimes Without connecting the probe and just with the heatsink connection to the ground the smps run normal for a while and then the voltage would drop( at this point when I turn on and off the input AC line to the smps very fast the voltage became normal and after a few second it will drop again)
I personally think that the problem is from IR chip (But I do not have anymore of that to test it)
what do you think?
Thank you in advance
I did all your tips about the sudden voltage drop and the only thing that worked is {{connecting heatsink to the low voltage ground (Middle leg of lm7815 regulator)+ connecting an oscilloscope probe to the gate of Q1 and the ground clip to the low voltage ground}} at the same time ( If I remove one of connection above the problem would come back)
sometimes Without connecting the probe and just with the heatsink connection to the ground the smps run normal for a while and then the voltage would drop( at this point when I turn on and off the input AC line to the smps very fast the voltage became normal and after a few second it will drop again)
I personally think that the problem is from IR chip (But I do not have anymore of that to test it)
what do you think?
Thank you in advance
Last edited:
Jagd.Panther
New member
Please share pics of your updated SMPS and share the updated schematic with all of the fixes you did.
badboy_6120
Member
everything are the same as ludo schematic except for these:
1- I replace D9 that was MUR120 to MUR460
2- for input rectifier I use KBPC3510W
3- Changed the FETs to IGBTs
4- Replace the output bridge diodes from MUR810 to MUR1520
5- Instead of using 2x1000uf cap for each rail of input I use 3x680uf
6- I just install 2x1000uf for each rail of output instead of 3x1000uf
7- Connection between Heatsink and Low voltage ground
8- Connecting Oscilloscope probe to gate of Q1 and low voltage ground
Those are the whole changes from my current smps than Ludo's
Last edited:
badboy_6120
Member
Problem solved:w)
Changing R6 to 10K (as you mentioned)+Connecting Heatsink to the low voltage ground(as you mentioned)+Removing R1
Last edited:
badboy_6120
Member
You're right
thanks for the tip
badboy_6120
Member
I want to start with the pcb and I want your helps because I never design a pcb dealing with high voltage and high frequency
Is ludo pcb ok? (V3.1)
I reupload it for you for ease of access
View attachment smps v3.11.pdf
Is ludo pcb ok? (V3.1)
I reupload it for you for ease of access
View attachment smps v3.11.pdf
Jagd.Panther
New member
It's better than the one you use for sure.
badboy_6120
Member
please show me every weak point in this design so I will make a good pcb of it
If I remember right you said to closer the caps to the chips??right??
badboy_6120
Member
I am working on a simpe smps - just for learning. And I am wondering why so many SMPS units (especially unstabilised ones) have no output inductors. Today I have measured current going from the diodes to the capacitor bank. The load at the secondary is only 2,3A, the resistor for measurements I have used 0,1R for easier calculation. Just please take a closer look, how does the current curves look like with no output inductors, ruding OFF phase current is going to 0 and at the ON state the current spikes are really high (nearly 8A !!!!). According to my capacitors (Panasonic FC) datasheet the max ripple availible I have is approx 6A so the capacitors will ''wear'' out very quick !!. Further more, higher current spikes at secondary means higher current spikes at primary too --> less power, more prone smps to fail.
The conclusion is --> building an unregulated SMPS without output inductors is kind of pointless.
badboy
Output inductor ''charged/magnetized'' during ON state and it is ''giving back'' the current during OFF state, so the current flow is more smooth. Disadvantage is larger voltage drop at the output, so the secondaries must be winded for much higher voltage (regulated SMPS).
The conclusion is --> building an unregulated SMPS without output inductors is kind of pointless.
badboy
Output inductor ''charged/magnetized'' during ON state and it is ''giving back'' the current during OFF state, so the current flow is more smooth. Disadvantage is larger voltage drop at the output, so the secondaries must be winded for much higher voltage (regulated SMPS).
Attachments
Last edited:
Basicly you can putt the value from the forum calculator but you will never know for what condition it is calculated.
To do it right way you have to calculate inductance value for certain amount of avarage output current, choke must be charged during ON time and there must be enough inductance to store energy for off time.
In my case it was around 80uH for 2A avarage output current at 70kHz. Look at my scope reading two posts above, the current with 80uH choke is nearly constant.
Price you have to pay is larger voltage drop.
To do it right way you have to calculate inductance value for certain amount of avarage output current, choke must be charged during ON time and there must be enough inductance to store energy for off time.
In my case it was around 80uH for 2A avarage output current at 70kHz. Look at my scope reading two posts above, the current with 80uH choke is nearly constant.
Price you have to pay is larger voltage drop.
badboy_6120
Member
Thank you so much for your thorough explanation
Jagd.Panther
New member
There are topologies that don't require an inductor. Unregulated DC/DC converter running at maximum duty cycle / small dead time is one of them (LLC with primary side resonance is another example). Checkout Car amplifiers rated over 500 W - 1 kW (example: http://soundstream.com/store/media/...9df78eab33525d08d6e5fb8d27136e95/4/_/4_19.jpg ), you won't find any "output inductors" there (in 99% cases)
Are you using full wave rectification? What is dead time/ sw frequency?
you can add another cap in parallel as well if you are so concerned about capacitors lifespan. Or a choke.
Not true. You are oversimplifying things. BTW if you put 0.1ohm resistor between diodes and the cap you were measuring ripple current on capacitor wrong.
badboy_6120
Member
Hi everyone
I'm back with good news about the transformer
I Change the Primary and secondary winding and now with 1250W power the transformer temperature without any cooling system is less than 80°c and with 1600W power and a pc fan the temperature won't go near 60°c
It's the best result I could obtain so far and it's pretty good for me
So I go further and now I'm working on the protection board
now I have a problem with startup. because the protection kicks in as soon as the Relay contacts closed and shorten the input resistor
If I short circuit the current sense secondary at startup and after relay contact closed release it, the smps starts normally
I use 10Ω for R1 and 10KΩ for R5 in protection board (picture is attached below)
and with above resistors value I set the protection to kick in, in about 1700 to 1900W
am I doing this wrong????
Should I set it so it does not active at startup???????or can I do something else to prevent startup activation????
I'm back with good news about the transformer
I Change the Primary and secondary winding and now with 1250W power the transformer temperature without any cooling system is less than 80°c and with 1600W power and a pc fan the temperature won't go near 60°c
It's the best result I could obtain so far and it's pretty good for me
So I go further and now I'm working on the protection board
now I have a problem with startup. because the protection kicks in as soon as the Relay contacts closed and shorten the input resistor
If I short circuit the current sense secondary at startup and after relay contact closed release it, the smps starts normally
I use 10Ω for R1 and 10KΩ for R5 in protection board (picture is attached below)
and with above resistors value I set the protection to kick in, in about 1700 to 1900W
am I doing this wrong????
Should I set it so it does not active at startup???????or can I do something else to prevent startup activation????
Last edited:
badboy
I have had exacly same problem. Dynamics of the protection circuit is very tricky.
You can try to filter it out with RC networks but I may be too slow in operation after.
In my case I decided to start the SMPS while main capacitors are charging, the SMPS was starting at about 200V DC at main caps + it was going in kind of brust mode (I call it buzz mode) for a few ms.
Belive me you will spend more time with proper protection than with thrensformer
I have had exacly same problem. Dynamics of the protection circuit is very tricky.
You can try to filter it out with RC networks but I may be too slow in operation after.
In my case I decided to start the SMPS while main capacitors are charging, the SMPS was starting at about 200V DC at main caps + it was going in kind of brust mode (I call it buzz mode) for a few ms.
Belive me you will spend more time with proper protection than with thrensformer