1000w smps based on LUDO3232

Silvio

Well-known member
I think you will be better off using 2 X ir2110 instead of gate drive. A gate drive transformer is a delicate thing to do and it has to be symmetrical in all ways with the 5 windings. It also needs to have a high permeability core other wise you will have to do a lot of turns to get to the right inductance. One other thing I want to mention is that if you load the sg3525 more than necessary the wave form starts to deform. On the other hand with IR2110 the wave form comes out nice and clean. For isolation the GDT is better. I guess for a first experience it would be better to start first with IR2110.
I experimented with gate drive transformer and what I told is from my experience.
 

stewin

Member
I think you will be better off using 2 X ir2110 instead of gate drive. A gate drive transformer is a delicate thing to do and it has to be symmetrical in all ways with the 5 windings. It also needs to have a high permeability core other wise you will have to do a lot of turns to get to the right inductance. One other thing I want to mention is that if you load the sg3525 more than necessary the wave form starts to deform. On the other hand with IR2110 the wave form comes out nice and clean. For isolation the GDT is better. I guess for a first experience it would be better to start first with IR2110.
I experimented with gate drive transformer and what I told is from my experience.

i have used gdt with very good success , always used small e:e cores , moved away from using ir2110 because of failure, ir2110 always burnt and are expensive to replace . but thats my experience
 

stewin

Member
I think you will be better off using 2 X ir2110 instead of gate drive. A gate drive transformer is a delicate thing to do and it has to be symmetrical in all ways with the 5 windings. It also needs to have a high permeability core other wise you will have to do a lot of turns to get to the right inductance. One other thing I want to mention is that if you load the sg3525 more than necessary the wave form starts to deform. On the other hand with IR2110 the wave form comes out nice and clean. For isolation the GDT is better. I guess for a first experience it would be better to start first with IR2110.
I experimented with gate drive transformer and what I told is from my experience.

i have used gdt with very good success , always used small e:e cores , moved away from using ir2110 because when smps output failed the ir2110 always burnt and are expensive to replace . but thats my experience
 

Silvio

Well-known member
i have used gdt with very good success , always used small e:e cores , moved away from using ir2110 because when smps output failed the ir2110 always burnt and are expensive to replace . but thats my experience

Hi Stewin, well using EE transformers brings the need to use ecw. Isolation is not so good at 320v, I am saying this is that usually when mosfets blow they tend to short out the gate with the source and drain. If the smps is not short circuit protected then there will be danger of the input DC voltage shorting out with the control circuit leading to a catastrophic damage. For a gate drive trafo I would rather use a small toroid around 15 to 20mm dia and using insulated wire. This will make a better trafo more safe and more symmetrical. The wave form needs to be nice and clean at all times especially on heavy load. My experience with these gate drive trafo is that the the SG3525 must not be loaded. I had a current booster made up of BD140 and BD139 after the SG but when loading the output of the GDT to about 1A the wave form from the SG3525 was not so pretty. It had an overshoot on the rising edge. I tried to tame it as much as I could but I could not get rid of it completely.
Regarding IR2110 I did get some from aliexpress and they where quite cheap. I tried them out and they all worked satisfactory.
 

stewin

Member
Hi Stewin, well using EE transformers brings the need to use ecw. Isolation is not so good at 320v, I am saying this is that usually when mosfets blow they tend to short out the gate with the source and drain. If the smps is not short circuit protected then there will be danger of the input DC voltage shorting out with the control circuit leading to a catastrophic damage. For a gate drive trafo I would rather use a small toroid around 15 to 20mm dia and using insulated wire. This will make a better trafo more safe and more symmetrical. The wave form needs to be nice and clean at all times especially on heavy load. My experience with these gate drive trafo is that the the SG3525 must not be loaded. I had a current booster made up of BD140 and BD139 after the SG but when loading the output of the GDT to about 1A the wave form from the SG3525 was not so pretty. It had an overshoot on the rising edge. I tried to tame it as much as I could but I could not get rid of it completely.
Regarding IR2110 I did get some from aliexpress and they where quite cheap. I tried them out and they all worked satisfactory.

hi silvio thanks for the reply ,very useful info and thanks for sharing your experience . even though i still have phobia of using ir ic as gdt,i will revisit it.
but i have to have many ir2110 as spare , cause when amp fails , the smps also fails ; leading to gdt, i.e ir 2110 failing.
 

Silvio

Well-known member
hi silvio thanks for the reply ,very useful info and thanks for sharing your experience . even though i still have phobia of using ir ic as gdt,i will revisit it.
but i have to have many ir2110 as spare , cause when amp fails , the smps also fails ; leading to gdt, i.e ir 2110 failing.

Hi Stewin, Well as far as the smps is concerned this must have a good working short circuit protection. It must sense both high and low side pulses. The one from Ludo is a good example of this. The only mishap is that it latches and you have to switch off for it to come on again.

I am working on a HB smps which is LLC and the protection is self resetting. This will give me a chance for a good output capacitance value especially on start up. I am using an EE55 and my hopes are for 2000w continuous and 2500w on peak. I still have to wait for some more parts to come. I just finished the pcb etching and drilling and fitting the heat sinks. The main oscillator and protection is on a separate pin header. This makes it more compact and also ease of repair in the event of failure. I also got the auxiliary smps on a pin header. This was published here a month or so ago.

The input capacitance is 4 caps @ 680uF X 450v this gives 2700uf across the 320v rail. The output voltage is to be around 85-0-85 and a capacitance of 5000uf on each leg of the output rail. it also include a double regulated 15v rail for preamp and a heat regulated fan speed control for 3 X 80mm 12v fans. Thermal protection is also included. The only thing that it does not have is that the fan motors will stop on thermal overload as these are powered by the main transformer and not from an auxiliary supply.

here is a pic of my work so far for the prototype

DIYsmps 2000.JPG
 

stewin

Member
Hi Stewin, Well as far as the smps is concerned this must have a good working short circuit protection. It must sense both high and low side pulses. The one from Ludo is a good example of this. The only mishap is that it latches and you have to switch off for it to come on again.

I am working on a HB smps which is LLC and the protection is self resetting. This will give me a chance for a good output capacitance value especially on start up. I am using an EE55 and my hopes are for 2000w continuous and 2500w on peak. I still have to wait for some more parts to come. I just finished the pcb etching and drilling and fitting the heat sinks. The main oscillator and protection is on a separate pin header. This makes it more compact and also ease of repair in the event of failure. I also got the auxiliary smps on a pin header. This was published here a month or so ago.

The input capacitance is 4 caps @ 680uF X 450v this gives 2700uf across the 320v rail. The output voltage is to be around 85-0-85 and a capacitance of 5000uf on each leg of the output rail. it also include a double regulated 15v rail for preamp and a heat regulated fan speed control for 3 X 80mm 12v fans. Thermal protection is also included. The only thing that it does not have is that the fan motors will stop on thermal overload as these are powered by the main transformer and not from an auxiliary supply.

here is a pic of my work so far for the prototype

View attachment 6963


very lovely and nice layout , after assembling it please post pictures and results.
 

Silvio

Well-known member
Change the file to PDF go to advanced on the website and upload the file. I cannot open it with my 7zip file opener

Silvio
 

mituadas

New member
Hi

I have made the PCB and assembled it. Now i am facing a new problem. When SMPS current is on I am not getting correct voltage. First of all by using multi meter i am getting some random values. After switch off my SMPS i am getting correct voltage and able to measure the voltage also it showing correct voltage. I am attaching my SMPS photos for reference.

WhatsApp Image 2019-08-22 at 10.54.23 PM (1).jpeg

WhatsApp Image 2019-08-22 at 10.54.23 PM.jpeg

Please guide if anybody came across the same Issue.
 

Silvio

Well-known member
Hi

I have made the PCB and assembled it. Now i am facing a new problem. When SMPS current is on I am not getting correct voltage. First of all by using multi meter i am getting some random values. After switch off my SMPS i am getting correct voltage and able to measure the voltage also it showing correct voltage. I am attaching my SMPS photos for reference.

View attachment 7027

View attachment 7028

Please guide if anybody came across the same Issue.

@Mituadas

First of all I see no input filter in your smps, secondly I think your multimeter is getting disturbed by the oscillations of the smps. I suggest you buy a better quality meter. You can also load the output with a 100w lamp to dump out any spikes that may be ending up on the output.

Make a 1 or 2 turn coupling loop around the center bobbin of the transformer and check the waveform with an oscilloscope.
Please post pictures of the waveform.

Regards Silvio.
 

mituadas

New member
I have tested the output with 100w,200w and 500w halogen bulbs. every think is fine no fluctuation. Only think is i have not added feedback circuit in PWM board. I checked the frequency at Transformer Aux output pin and found to be good. I am having two doubts one is my input voltage exceeds 260v and no feedback circuit.
 

Silvio

Well-known member
I have tested the output with 100w,200w and 500w halogen bulbs. every think is fine no fluctuation. Only think is i have not added feedback circuit in PWM board. I checked the frequency at Transformer Aux output pin and found to be good. I am having two doubts one is my input voltage exceeds 260v and no feedback circuit.

This looks quite odd to me that you had no fluctuation in your output voltage especially with 500w load. You see 500w is half the rating of your smps. Do not forget that your smps is not regulated. To be sure you can connect a 100w bulb than add the 500w load. If you notice that the 100w bulb dims a little it means you have some fluctuation.

You did not show any pics of the wave form. If your input voltage is 260v than you will experience a higher output voltage and no drastic fluctuation when the mains voltage will vary a little. With your experience regarding the voltage reading being erratic your multi meter is getting distracted by the smps oscillations. See if you can dump out any spike getting into it by putting a film capacitor with the voltage rated more than the output of the smps. Put this across the terminals at the meter input maybe it will help. If not you need to buy a better meter.

One last thing I want to mention that if your halogen lamp is rated at 500w this load only holds true if your output voltage is the same rating of the voltage of the halogen lamp. If for example the output voltage of the smps is 100v and the rating of the halogen lamp is 230v than its most likely that you will be drawing only around 200-250w. If you want to know exactly what you are drawing from the smps you must connect a voltmeter and an ampere meter. You will multiply the values together an that will be your real load.

Take a look here to see how you should make your measurments

https://www.youtube.com/watch?v=n4R72t2FXFo&t=1s
 

mituadas

New member
Hi SILVIO
As per your advises i purchased a good Multimeter and output voltage is fine. Now i am facing another problem. Without Load i am getting 160v(80-0-80). After load 220V 25w Bulb voltage drop down to 150V(75-0-75) then load of 220v 200W bulb voltage drop down to 80v(40-0-40) again load of 220v 500w halogen bulb voltage drop down to 40v(20-0-20). I have tested wave from AUX winding of Transformer it is good and no distortion. Please advise me how to rectify the fault.
 

Silvio

Well-known member
Check the voltage across the transformer winding during load and see if it is dropping from the half bridge. You can check also the input voltage AC and see if it is dropping there.

voltage expected at the half bridge without load is 155v if the input ac voltage is 220v (220/2=110v X 1.414 = 155v)

check these first and let me know the result.

tell me also what is the type of transformer you are using EX. ETD49 core material N87.

What is the number of turns in the primary winding and also in the secondary.

What switching transistors are you using (part number)

What output diodes are you using ( Part number)

If all is well your voltage drop should be no more than 10 to 15% 10%= (160v / 1.1= 145v) =72-0-72

I await your reply regards Silvio
 

mituadas

New member
I just disconnected the feed back. Now my SMPS output voltage is good. But it is not regulating the voltage. I will check the feed back section once again and post the results.

Regards

S Das
 
Top