Audio smps 700w (IR2153)

Silvio

Active 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

Active 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

Active 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

Active 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

Active 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: 9
Last edited:
Top