Audio smps 700w (IR2153)

Silvio

Well-known member
Hi Steven, I am seeing the layout and looks good but I think you wasted a lot of space on the pcb. I guess it could be made smaller, perhaps turning the output capacitors will get you there. This being a double sided pcb shall not be a problem. Using a pin header for the control circuit will bring an option to change if needed, The pin header also served to get the chip near the mosfets and also eliminating stray capacitances and inductances getting in the way. Creating a PE ground around the pcb will also help out with decoupling and interference suppression making the route shorter. This will eliminate a 3 pin connector on the pcb input as then it will be grounded to the chassis. You did not ground the heatsink. Trace width shall be calculated at 1mm per amp so to handle 8 amp then they must be 8mm wide. You can also use half the width and make a double trace from top and bottom using vias to connect together. I think that the size of the output capacitors are a bit too large, usually these are around 18mm in diameter. I see that they are the same size of the input capacitor :).
Lastly keep clearances between traces adequate 4mm on high voltage rails (320vdc) 3mm on AC input rails (230vac) and 2mm on the low voltage output rails up to 100vdc. Low voltage traces the usual around 0.3 to 0.5mm.
Note. It may be difficult to find a 160v Zener however you can make allocation for 3 x 55v zeners instead.

Regards Silvio
 
Last edited:

Silvio

Well-known member
Hi Steve, that is a big improvement from your previous layout. I would like to point out some more improvements though.
1) you can add another pair of secondary caps making a total of 6. You can put them closer together
These will give more punch on the base notes.
2) try to put the startup circuit on the main pcb and thus keeping the 320v line away from the low voltages on the pin header.
The pin header must not be taller than the trafo.
3)you need a support in the centre of the pcb
4) you must provide a means (screw holes) to tie the heatsink to the pcb
5) the 320 volt rail must be clear of the heatsink. (Cannot be on the top layer if under heatsink)
These are a few tips in designing a better pcb and safer also.

Regards Silvio.
 

yingtso

New member
I cannot see why the circuit does not work with the IRS chip. They suppose to have the same structure as the dip types. Regarding the led its rather difficult really as during operation the aux winding is supplying the voltage and when the SCR fires it pulls the supply to ground. Adding any resistance here may not ground the supply rail in time. As the chip stop oscillating only the start up circuit is in operation but this will only supply a few milliamps and may not be enough to light up the led.
Regarding some more modifications. Well I hope ZD1 is around 150v. To aid start up you can chainge ZD2 to 12v, ZD3 to 15v and change R8 to 120R. Before making changes check the input voltage before R8, this should be 17v to 20v. Check voltage at pin 1 this should read 13v. With the chip removed the startup voltage at pin 1 should not be less than 10v. I encountered some chips that need 11v at start up and refuse to start at a lower voltage. It could be that fake chips draw more current at start up.
Hi, Silvio

I recently got some more IR"S"2153 again.

I am under the current configuration of IR2153

Some ICs can work, some burn up when they go up, and they are not stable.

I re-read the datasheet of IR"S"2153

It seems that the working frequency is not as high as IR2153(IR"S"2153 is limited up to 100k)

I suspect that the IC may be burnt under 3x frequency startup


*In addition, what are the advantages of 3 times the starting frequency? And why?


Maybe it can be started at normal frequency and it can work well(about 50~70k)



Thanks & Regards

Benny
 

Silvio

Well-known member
Hi, Silvio

I recently got some more IR"S"2153 again.

I am under the current configuration of IR2153

Some ICs can work, some burn up when they go up, and they are not stable.

I re-read the datasheet of IR"S"2153

It seems that the working frequency is not as high as IR2153(IR"S"2153 is limited up to 100k)

I suspect that the IC may be burnt under 3x frequency startup


*In addition, what are the advantages of 3 times the starting frequency? And why?


Maybe it can be started at normal frequency and it can work well(about 50~70k)



Thanks & Regards

Benny
Hi again Benny, the high frequency startup is because the IR2153 does not have a soft start. Charging the secondary capacitors at startup present a nearly dead short especially if having more than 2000uf on each of the center tap windings. However starting the trafo with 3 times the frequency cannot deliver that much current as this is wound for a lower frequency. This will reduce the stress on the mosfets and prevent them from burning.
Another way to eliminate this problem can be arranged by putting a 5R resistor on the VS line to limit the current and then this resistor can be bypassed by means of a relay after a couple of seconds after startup. Another idea comes in mind is adding an NTC thermistor on the VS line. However with an NTC you cannot start the smps while the thermistor is still warm as it will have no effect as its resistace would be very low.

Regards Silvio
 

Silvio

Well-known member
Control board V1.2:
SMPS board V1.3:


All seems well. Just one small detail. The primary snubber capacitor usually has a pitch of 7.5- 10mm (ceramic type 1 or 2kv) and the snubber resistor is usually is of the 3watt type. Make sure that these will fit. I cannot determine the pitch used but assuming by proportion. As for the secondary snubber these are usually of a lower wattage but still use a ceramic type.
Above everything else all seems well and you did a fine job this time.
Well done. :)
 

Silvio

Well-known member
Control board V1.2:
SMPS board V1.3:


I see no output for a fan motor which normally is essential in an amplifier. This will also help if needing a separate 12-15v dc for the speaker protection relay. Ideally this part does not share a common ground with the preamp stage.
Another thing I noticed are the diodes used for the rectification of the preamp supply. They seem large of the 3amp type. The 100uf/35v capacitors after the diodes also seem large. I think they are oversized. I hope I am not mistaken here. 1 amp diodes will surely suffice here.
I am attaching a schematic for a temperature sensed fan speed control.
Another note for you, the ground of the preamp supply must be lifted. I advice you to insert a 10R resistor when connecting the center grounds of the main output to the preamp center ground. This will help to eliminate interference in the preamp stage.
 

Attachments

  • fan speed control.PNG
    fan speed control.PNG
    35.8 KB · Views: 29
Last edited:
To tell you the truth although I designed the schmitt trigger version I never tried it for real on the smps.
Things you can try
1) You can remove the input thermister and put a 5watt 47R resistor instead. To help a slower start up
2) Following my schematic for schmitt trigger there is a decoupling capacitor which is coming from the sample diodes (1N4148) and has a value of 220nF, You can rise the value slowly 1uF- 4.7uF until you get a start up without tripping on the surge current, Keep to the minimum possible otherwise during a short circuit your fets will not be saved.
A picture of a center tapped current transformer. Note the 1.5 turn coupling loop in the bottom.

View attachment 7395
Silvio, how are you? It's been a while, but I've resumed work with this SMPS and now I'm going to use the current limiting module.
In my smps I planned to place the CT, could you help me with the construction of it?
I start with double thread for the central key and after 22 turns I have the other 2 threads left. Do you have an order to connect on the PCB?
I would also like to know the values that go on this PCB and the orientation of the transistor.

I am attaching some photos so you can see how the variable resistor adapts to adjust current, it really is comfortable, cheaper than multiturns.


I put 10K for RT and 560pF for CT.
Just to know....the variable resistors for frequency and current can be of other values?


Thanks!
 

Attachments

  • Silvio.jpg
    Silvio.jpg
    125 KB · Views: 25
  • WhatsApp Image 2023-01-25 at 15.37.15 (1).jpeg
    WhatsApp Image 2023-01-25 at 15.37.15 (1).jpeg
    99.5 KB · Views: 23
  • WhatsApp Image 2023-01-25 at 15.37.15.jpeg
    WhatsApp Image 2023-01-25 at 15.37.15.jpeg
    103.3 KB · Views: 25
Last edited:

Silvio

Well-known member
If you are choosing to use the Current transformer an option for a short circuit protection is available. The pin header however is different from the one using a current limiter and I am afraid to say that you have to make another one. The last version of pin header uses a high frequency start up and also caters for the short circuit protection which uses a small SCR to shut down the supply rail for the IR2153 This is the best option I found as the IR2153 does not have a soft start. The soft start is created by running the smps at around 3 times the frequency. This will not let the trafo expel its maximum power hence limiting the output current. The circuit is designed so that firstly running at the high frequency at startup then resumes to the running frequency after about 1.5 seconds
The current transformer will sense the current in the half bridge and when a threshold is reached fires the scr to pull down the supply to the IR2153.
The current transformer is wound with 0.2mm enameled copper wire and has about 75 turns bifilar wound on a small core about 12 to 15mm dia.

I am posting my final version of this smps which there is some changes in the output capacitors which now contains 6 instead of 4. The schematic is also changed showing the short circuit protection. You can still use your existing main pcb but the pin header has to be changed to cater for the new version. There are also some changes in the values of ZD2 and ZD3 these where done as some chips need a little more voltage at start up otherwise they fail to start if you are not using genuine chips.
There is also the pin header catering for the high frequency soft start and short circuit protection.
In the file you find the description of operation with the short circuit and high frequency soft start.

Regards Silvio
 

Attachments

  • Audio smps 700w V2.11.pdf
    805.7 KB · Views: 92
Last edited:
If you are choosing to use the Current transformer an option for a short circuit protection is available. The pin header however is different from the one using a current limiter and I am afraid to say that you have to make another one. The last version of pin header uses a high frequency start up and also caters for the short circuit protection which uses a small SCR to shut down the supply rail for the IR2153 This is the best option I found as the IR2153 does not have a soft start. The soft start is created by running the smps at around 3 times the frequency. This will not let the trafo expel its maximum power hence limiting the output current. The circuit is designed so that firstly running at the high frequency at startup then resumes to the running frequency after about 1.5 seconds
The current transformer will sense the current in the half bridge and when a threshold is reached fires the scr to pull down the supply to the IR2153.
The current transformer is wound with 0.2mm enameled copper wire and has about 75 turns bifilar wound on a small core about 12 to 15mm dia.

I am posting my final version of this smps which there is some changes in the output capacitors which now contains 6 instead of 4. The schematic is also changed showing the short circuit protection. You can still use your existing main pcb but the pin header has to be changed to cater for the new version. There are also some changes in the values of ZD2 and ZD3 these where done as some chips need a little more voltage at start up otherwise they fail to start if you are not using genuine chips.
There is also the pin header catering for the high frequency soft start and short circuit protection.
In the file you find the description of operation with the short circuit and high frequency soft start.

Regards Silvio
My smps is conditioned according to your smps V2 which has a CT in the primary of the transformer.
What module do you advise me to put? As for the CT, can I use 0.2mm or higher? Is there any drawback with this? Always respecting 75 double laps.
Is the diameter from 12 to 15mm internal? can iron powder be used?
I thank you in advance for sharing and helping.
The module that has this new file does not have an indicator led to know when the protection is activated.


I edit the latter at 9:00 p.m.:

In V2.11 the improvements that I noticed are the increase of capacitors at the output, the secondary MUR, the Z13V, Z15V and the CT that changed the ratio.

I see that the PCB "HF SOFT START + S/C PROTECTION
WITH SCR" in both V2 and V2.11 is exactly the same, so should I make this PCB for my SMPS?
The CT is 80:1 and not 80:1.5?

My SMPS is made like V2 so the module "HF SOFT START + S/C PROTECTION
WITH SCR" will not be difficult to connect since the pins match well.

Thank you!
 
Last edited:

steppler

New member
Hi Silvio & Friends,

I have IR2153 problems again. When starting without secondary capacitors - everything all right. But the problem with the capacitors. Only 330uF is enough to 2153 crash. Sometimes mosfets too. A soft start will probably be necessary.
Please, can I use the circuit from the attached picture? Is it verified in practice? The starting frequency is over 200kHz. Can the IC handle it?
Thanks to all for your help.
 

Attachments

  • SoftStart.png
    SoftStart.png
    43.9 KB · Views: 87

Silvio

Well-known member
My smps is conditioned according to your smps V2 which has a CT in the primary of the transformer.
What module do you advise me to put? As for the CT, can I use 0.2mm or higher? Is there any drawback with this? Always respecting 75 double laps.
Is the diameter from 12 to 15mm internal? can iron powder be used?
I thank you in advance for sharing and helping.
The module that has this new file does not have an indicator led to know when the protection is activated.


I edit the latter at 9:00 p.m.:

In V2.11 the improvements that I noticed are the increase of capacitors at the output, the secondary MUR, the Z13V, Z15V and the CT that changed the ratio.

I see that the PCB "HF SOFT START + S/C PROTECTION
WITH SCR" in both V2 and V2.11 is exactly the same, so should I make this PCB for my SMPS?
The CT is 80:1 and not 80:1.5?

My SMPS is made like V2 so the module "HF SOFT START + S/C PROTECTION
WITH SCR" will not be difficult to connect since the pins match well.

Thank you!
Regarding the pin header I guess the one on the last pdf works best(HF s/start & scr protection) when it comes to protection. It has been tried over and over again and never failed.
The current trafo dia is from the ouside (12-15mm) using thicker wire than suggested (0.16-0.2mm) may not make it possible to fit in the ferrite ring core. 75 or 80 turns will not make a difference in operation. There is adjustment in the preset to set the trip current. Regarding the coupling loop it is mentioned as 1.5 because just passing through the centre will make the coupling rather loose, however passing again from under the ring core gives better coupling. It is still 1 turn around the core.
I hope this helps good luck
Regards Silvio.
 

Silvio

Well-known member
Hi Silvio & Friends,

I have IR2153 problems again. When starting without secondary capacitors - everything all right. But the problem with the capacitors. Only 330uF is enough to 2153 crash. Sometimes mosfets too. A soft start will probably be necessary.
Please, can I use the circuit from the attached picture? Is it verified in practice? The starting frequency is over 200kHz. Can the IC handle it?
Thanks to all for your help.
I don't know what circuit you are using. Charging secondary capacitors present a dead short for the mosfets at statup. They may blow at times. You can do these adjustments as follows.
1) If you have an inrush current ntc resistor increase the value to 20 ohms.
2) put a couple of small inductors (5uH) after the secondary diodes before feeding the secondary capacitors, one on the positve and the other on the negative rail.
3) The high frequency soft start will also help in limiting the charge current in the secondary capacitors.
4) yes the IR2153 can oscillate at 200khz.
In my smps I have 2000uF on each rail and I don't have any problems.
Good luck
Regards Silvio.
 

steppler

New member
3) The high frequency soft start will also help in limiting the charge current in the secondary capacitors.
.
Is it verified in practice?
Isn't C12 the other way around?

 
Last edited:
Regarding the pin header I guess the one on the last pdf works best(HF s/start & scr protection) when it comes to protection. It has been tried over and over again and never failed.
The current trafo dia is from the ouside (12-15mm) using thicker wire than suggested (0.16-0.2mm) may not make it possible to fit in the ferrite ring core. 75 or 80 turns will not make a difference in operation. There is adjustment in the preset to set the trip current. Regarding the coupling loop it is mentioned as 1.5 because just passing through the centre will make the coupling rather loose, however passing again from under the ring core gives better coupling. It is still 1 turn around the core.
I hope this helps good luck
Regards Silvio.
Good morning Silvio! I already started to make the PCB of the last PDF. The electrolytic capacitors 1, 47 and 470uF are 25V? Can the others be ceramic or polyester? I will try to get those values.

Could you tell me the polarity of the electrolytic capacitors?

As for the CT, could you send the measurements of internal and external diameter and thickness? Ferrite or iron powder?
Thank you!
 

Attachments

  • WhatsApp Image 2023-01-26 at 14.55.39.jpeg
    WhatsApp Image 2023-01-26 at 14.55.39.jpeg
    100.5 KB · Views: 10
Last edited:

Silvio

Well-known member
The wrong schematic mark is used. That confused me. I'm sorry.
Hi Steppler I took a look at the schematic again and C12 is a 470 uF. It is drawn the other way round. This controls the timing of the changeover between high and normal operating frequency.
Sorry for the inconveniece.
 
Top