1000w smps based on LUDO3232

blasphemy000

New member
That is not a PFC stage. That is just a high voltage boost converter with voltage regulation and no current limit protection. That circuit has 940uF of bulk capacitance after the rectifier, where as a PFC converter usually uses somewhere around 1uF or less, so that the input voltage to the boost inductor is able to track the sine wave of the mains voltage. The TL494 PWM chip cannot be used as a PFC controller because it does not have the ability to track the mains voltage and cannot do current-mode control. There is a very wide selection of controllers that can be used to drive a PFC. They range from relatively simple to very complex designs and they vary widely in price also. The UCC28019 and the NCP1654 controllers are both 8-pin devices that are quite simple to design a PFC around.

The UCC28019 operates at 65KHz and the datasheet provides all of the calculations so that you can design to whatever power level you need and there is also an example design to explain how to use the calculations.
The NCP1654 is available in 65KHz, 133KHz, and 200KHz versions. The datasheet for this chip also provides all of the information you need to do the calculations and it also shows an example design and example calculations.

I've personally used the NCP1654 in a PFC stage and the calculations that they provide work great and it is fairly easy to get it working as long as you follow the design procedures outlined in the datasheet. Both of these chips are very, very similar though, and I imagine the UCC28019 is equally easy to use.

- Brad
 

Upik

Member
That is not a PFC stage. That is just a high voltage boost converter with voltage regulation and no current limit protection. That circuit has 940uF of bulk capacitance after the rectifier, where as a PFC converter usually uses somewhere around 1uF or less, so that the input voltage to the boost inductor is able to track the sine wave of the mains voltage. The TL494 PWM chip cannot be used as a PFC controller because it does not have the ability to track the mains voltage and cannot do current-mode control. There is a very wide selection of controllers that can be used to drive a PFC. They range from relatively simple to very complex designs and they vary widely in price also. The UCC28019 and the NCP1654 controllers are both 8-pin devices that are quite simple to design a PFC around.

The UCC28019 operates at 65KHz and the datasheet provides all of the calculations so that you can design to whatever power level you need and there is also an example design to explain how to use the calculations.
The NCP1654 is available in 65KHz, 133KHz, and 200KHz versions. The datasheet for this chip also provides all of the information you need to do the calculations and it also shows an example design and example calculations.

I've personally used the NCP1654 in a PFC stage and the calculations that they provide work great and it is fairly easy to get it working as long as you follow the design procedures outlined in the datasheet. Both of these chips are very, very similar though, and I imagine the UCC28019 is equally easy to use.

- Brad
it seems like very good information, can you just give a picture with v out 320vdc
 

blasphemy000

New member
Not at this time, no. The datasheet provides all of the math required to design a PFC stage. My PFC stage is part of a much larger project that I don't want to release until the rest of the project is done. My output voltage is 390V and it's rated for 10 to 12KW. My topology isn't really going to be useful for the kind of supply in this thread.

- Brad
 

Silvio

Active member
hi, Mr. Silvio ... how to make the isolation of the probe on an oscilloscope, ??

Hello Upik, To make isolation of the oscilloscope you need a differential probe. This tend to be very costly. There is another way to do this is to use an isolation transformer for the SMPS on test and if you want you can also use isolation transformer for the scope.
( I use both at the same time)
You can look up the internet and see what a differential probe is.

You can read SMPS warnings on top of this page and download the pdf file to see what Microsim has to say about it. Please also read the tread and see what people said about this subject so that you will learn.

Regards Silvio
 

Upik

Member
Hello Upik, To make isolation of the oscilloscope you need a differential probe. This tend to be very costly. There is another way to do this is to use an isolation transformer for the SMPS on test and if you want you can also use isolation transformer for the scope.
( I use both at the same time)
You can look up the internet and see what a differential probe is.

You can read SMPS warnings on top of this page and download the pdf file to see what Microsim has to say about it. Please also read the tread and see what people said about this subject so that you will learn.

Regards Silvio
what about this oscilloscope, how to isolate itShopee_5d80c4ae6b58f97d754a60b111c3949f.jpg
 

Silvio

Active member
@ UPIK

The scope looks good to me but you need to give me model number so that I can see what it can handle. You need to use this with the laptop.

If you are using This kind of scope you will not need an isolation transformer as the power supply for the laptop is already isolated.

IMPORTANT NOTE Make sure that the laptop ground (usb case) and mains earth are not connected together
Please verify and check with multi meter.

If you have a normal PC with tower etc then you will have the mains earth and usb ground connected together. In this case you must use an isolation transformer

Isolation transformer is a transformer that has input 220v and secondary winding is also 220v. This will isolate scope from ground

Regards Silvio
 

Upik

Member
@ UPIK

The scope looks good to me but you need to give me model number so that I can see what it can handle. You need to use this with the laptop.

If you are using This kind of scope you will not need an isolation transformer as the power supply for the laptop is already isolated.

IMPORTANT NOTE Make sure that the laptop ground (usb case) and mains earth are not connected together
Please verify and check with multi meter.

If you have a normal PC with tower etc then you will have the mains earth and usb ground connected together. In this case you must use an isolation transformer

Isolation transformer is a transformer that has input 220v and secondary winding is also 220v. This will isolate scope from ground

Regards Silvio
model 6022BE ... how to test the large voltage Mr.Silvio, because I don't really understand how it works
 

Silvio

Active member
@ Upik

This kind of oscilloscope works in conjunction with the laptop. It uses the screen of the laptop for display.
It also uses the power from the USB port to power the oscilloscope. (You will not need to isolate it when using laptop on battery)
I tried to see the maximum voltage it can handle but could not find it. (mine for example can take a maximum of 300v when using 10X probe.)
At the primary of a half bridge transformer there is more or less around 160v so with a 10X probe it can take it very easy because limit is 300v.

For voltages exceeding 300v you must use a 100X probe instead. You can buy 100X probe separately.

You can also use a resistive divider across the transformer primary so that the scope will not see the full voltage when taking measurements with 10X probe.

Example 90k resistor in series with 10K resistor and probe and ground clip is attached across 10k resistor. This will divide voltage by 10.

The Hantek model 6022BE only has DC coupling and no AC coupling. AC coupling is not used much but you will need it to measure the noise level on the output of the smps. You can go for Model 6052BE instead. Maybe it cost a little more but it is more suitable.

You should see some videos on youtube how to use oscilloscope and learn a little bit about the subject.

Regards Silvio
 

Upik

Member
@ Upik

This kind of oscilloscope works in conjunction with the laptop. It uses the screen of the laptop for display.
It also uses the power from the USB port to power the oscilloscope. (You will not need to isolate it when using laptop on battery)
I tried to see the maximum voltage it can handle but could not find it. (mine for example can take a maximum of 300v when using 10X probe.)
At the primary of a half bridge transformer there is more or less around 160v so with a 10X probe it can take it very easy because limit is 300v.

For voltages exceeding 300v you must use a 100X probe instead. You can buy 100X probe separately.

You can also use a resistive divider across the transformer primary so that the scope will not see the full voltage when taking measurements with 10X probe.

Example 90k resistor in series with 10K resistor and probe and ground clip is attached across 10k resistor. This will divide voltage by 10.

The Hantek model 6022BE only has DC coupling and no AC coupling. AC coupling is not used much but you will need it to measure the noise level on the output of the smps. You can go for Model 6052BE instead. Maybe it cost a little more but it is more suitable.

You should see some videos on youtube how to use oscilloscope and learn a little bit about the subject.

Regards Silvio


hi Mr.Silvio ,,,, for the model you say, I have a hard time getting it, maybe I found it with a different brand. For quality I don't really understand.

sketch-1536033470157.jpg
 

Upik

Member
hi mr. silvio ,, i have ee55, with a frequency of 70-80 khz ,, what is the power that is released, with a full bridge topology
 

Silvio

Active member
hi mr. silvio ,, i have ee55, with a frequency of 70-80 khz ,, what is the power that is released, with a full bridge topology

Hi Upik the maximum power for an EE55 core and material PC40 you can get around 2500 to 2800 watts in full bridge topology.

Do not forget what I told you about the input capacitance now you will need around 3000uF at 400v in the input. You will also need a good soft start in the input and also while charging the secondary capacitors. In the output you will need 5000 to 6000uf for audio. A 6uH inductor will help out limiting some current during the secondary capacitor charge up. If you use inductor for output then you have to put snubbers across the diodes in the output as this will reflect spikes in the input winding.

The current at 160v is around 20A for this power. wire diameter is just below 0.5mm (0.48mm) at 75Khz.

Good luck Silvio
 

Upik

Member
thanks mr. silvio, i want to make with pfc, use GDT to drive mosfet. and use flyback transformer to supply pwm. it seems i have to use etd59 to get more power.

FB_IMG_1534130971860.jpg
 

Silvio

Active member
Hi upik May I ask you where are you going to use this amplifier? Keep a few things in mind that say a proper 3K smps can power a 5K amplifier.

The pcb shown cannot deliver more than 2.5k peak with an ETD49 or EE55. With the poor grid voltage at home it would be impossible to try this out properly. The PFC will help a bit but it does not do miracles when the input voltage drops drastically. You did not show the schematic maybe I can figure out better the setup.

If you are looking for a 3K smps then an ETD59 is more suited, another option is the make 2 smps with an ETD49 or EE55
One smps for positive and another smps for negative rails.

You can feed them from one large input capacitor bank You will still use half bridge and you still get your 3k or more. They can be on the same pcb if you want. I sent you some pics last time for a 3K smps. It is half bridge and use only ETD59 with 4 IGBTs rated at 30 amps each.

I would suggest that if you are going to build this smps it would be better to open another tread.

Good luck

Silvio
 

cerrem

New member
Hi upik May I ask you where are you going to use this amplifier? Keep a few things in mind that say a proper 3K smps can power a 5K amplifier.

The pcb shown cannot deliver more than 2.5k peak with an ETD49 or EE55. With the poor grid voltage at home it would be impossible to try this out properly. The PFC will help a bit but it does not do miracles when the input voltage drops drastically. You did not show the schematic maybe I can figure out better the setup.

If you are looking for a 3K smps then an ETD59 is more suited, another option is the make 2 smps with an ETD49 or EE55
One smps for positive and another smps for negative rails.

You can feed them from one large input capacitor bank You will still use half bridge and you still get your 3k or more. They can be on the same pcb if you want. I sent you some pics last time for a 3K smps. It is half bridge and use only ETD59 with 4 IGBTs rated at 30 amps each.

I would suggest that if you are going to build this smps it would be better to open another tread.

Good luck

Silvio

I would also add...if using two power supplies for + and - ...... try to use dual tracking feedback to keep the separate rails locked together...
 

Upik

Member
hi mr. silvio ,, i want to ask for your help, can you tell us how many turns on the primary ee55 and ee65 at 72khz. for the full bridge topology ... thank you very much in advance
 

Silvio

Active member
hi mr. silvio ,, i want to ask for your help, can you tell us how many turns on the primary ee55 and ee65 at 72khz. for the full bridge topology ... thank you very much in advance

Turns for EE55 full bridge 230v = 19 turns (72Khz) B max at 245v 1450 guass

Turns for EE65 full bridge 230v = 15 turns (72Khz) B max at 245 v 1450 guass

Silvio
 
Last edited:
Top