Class D 200 Wrms with 2 mosfet cheap

[norazmi]

Here again, this one i already made and success, idk if this amp can provide true 200 wrms but i aim at least 100 wrms, tested run with +- 45 vdc both linear supply and smps, 4 ohm still running cool, mosfet i use irf540 and irf9540, 1 uf filter cap, t106-2 iron powder core 22 uh. i`ll post later some of my picture finish amp and its small

credit goes to owner edtagle i think hes name at forosdeelectronica forum.

here is the schematic attach.

 

[norazmi]

here is the board, have many version of it.

 

[norazmi]

sample of finish amp, i`ll upload mine later

 

[norazmi]

This is the pwm picture, this amp already tested and finish, this amp run about 6 month without any problem at +-45 vdc, i did not test it with 2 ohm because its complementary n-ch and p-ch mosfet, i do test another module with ir2110 as a driver + tl074 +- 60 vdc and 2 ohm is easy to take.

picture of my amp base on this. anything regarding of this amp u can ask me, i`ll try my best to answer it. regards.

 

[michelle]

nice work

Hi Norazmi
greetings nice compact amp can you share pcb and component values for this amp whats
power in 4 ohms load and how difficult is it for someone trying d class first time
warm regards
michelle

 

[norazmi]

Hi Norazmi
greetings nice compact amp can you share pcb and component values for this amp whats
power in 4 ohms load and how difficult is it for someone trying d class first time
warm regards
michelle

Hi michelle,

Yes sure, i build that amp base on this pcb design, which use bd139 and bd140,. If you re first time to build this amp, im suggest you to use rail voltage +30 0 -30 vdc, 4 ohms it would be easy deliver at least 100 watts, depend on your transformer and supply, how much current you have, i would suggest at least u have +- 30vdc at 5 amp. its not really difficult to build already tested amp, just make sure be carefully when assembly this amp with right parts, double check after finish assembly before power it on, dont connect output to speaker yet after u connect to power supply, and please do check with volt meter DC measurement to make sure there is no dc at output, if u dont this, and ur amp have error and having dc at output , it will destroy your speaker.

the tricky part is to build inductor, this is important to make right inductor because some dont have core T106-2 (iron powder) but u can use toroid core (ferrite) with air gaps about 1-2 mm, DONT use T106-6 core because this core will generate heat much more than 80 Celsius!!, if u can get T106-2 this is better choice, or at least u can use ferrite core torodial or EE or EI core with right winding to get 80-90 % accurate value in uh, i have few sample of core i made it from ferrite toroid and EE @ EI core, i can show it if anyone want build it.

You can take 1uf caps from pc power supply and u can use main trafo to make filter about 22-30 uh, EE33 @ EI should work great, but as i say T106-2 is the best core for best performance.

regards,
Azmi.

 

[norazmi]

sample of toroid ferrite core 22uh and core size almost same as T106-2, this core material F77 i take it out from car power amp transformer, hehehe , easy part lol ((@ , and this core have air gaps about 2 mm , i cut it with mini drill.

 

[michelle]

thanks

Hi Norazmi
greetings thank you for the reply and pcb i will make it tomorrow only problem output
inductor ferrite core i have some ferrites i will show u
warm regards
michelle :UP:

 

[norazmi]

Hi Norazmi
greetings thank you for the reply and pcb i will make it tomorrow only problem output
inductor ferrite core i have some ferrites i will show u
warm regards
michelle :UP:

allright, no problem, i`ll show u one more, E core, i`ll take it picture and measure core size and winding, that toroid core have 23 laps awg 16.

 

[zeus_threat]

Hi noramzi your amp looks really interesting could you please post a bigger schematic so that the part numbers are readable thanks

 

[norazmi]

hi zeus,

sorry late reply, im working outstation and now back home, hummm, i`ll try get better schematic to view, today i`ll post my new smps pawn this amp , my smps base on ir2153 new one i`ve made today, output +- 42.5 and at max volume only drop 1.5 volt, mosfet for smps ir740 (cool), ir2153 (cool), output diode mur420 x4 (cool) the only warm is trafo, maybe huge current drawn from it. Based on the schematic this smps have short circuit protection, and it works. i`ll make another class d amp base on this schematic and i`ll post here too.

regards,
Azmi.

 

[zeus_threat]

Hi noramzi thanks for the reply looking forward to see your schematics

 

[norazmi]

here is the clear schematic, this is original one from ejtagle. according to the owner, lc filter must be 30 uh and cap filter is 640nf.

regards,
Azmi

 

[zeus_threat]

Hi noramzi thanks for the reply i will have a look at the circuit and sym it. Just once question have you checked or got any signs of heating due to cross conduction? Correct me if am wrong there is no deadtime in this circuit right?

 

[norazmi]

hi zeus,

yes, there is heating on bd139 and bd140 but tolerate below +-50 vdc, as i read from the forum and with my experience this amp hard to stable with more than 50vdc and need adjustment. but take a look with small amp and class D at least we can get 100 wrms at 8ohm and more watts with 4 ohm, i dont expect much power from this circuit since its discrete part are used and using nch and pch mosfet, we will face little problem with rds-on on both side which have lots of different. with my testing and experience since 4 month i use this amp and i build 3 amp from this circuit, there is no overheating at mosfet with pc power supply heatsink even u play hard, but yes bd139 and bd140 will produce heat but its tolerate with that voltage rails below +-50 vdc.

yep there is no dead time on this circuit, Microsim or other ppls who have experience on class D may give more info about it, since this amp is NOT ucd. Another version is use LM311 but as i test i`m prefer more with tl074 and not face much problem.

Btw if u want more power i can give u schematic with ir2110 as a driver, this thing can produce more power and u need huge power supply and capacitor bank unless u re using smps then u will save the weight. i`ve 4 pcs amp with ir2110 and one amp i put with my computer as sound system and its working alright all the time. I tested with mosfet irf4710 u can drive as low as 1 ohm .

regards,
Azmi.

 

[zeus_threat]

Hi noramzi thanks for the reply very useful info. Well i did have a look at deadtime and there are several ways to implement it most of what i have seen relies on gates to introduce deadtime though it can be done with discrete circuit but would be more complex. I did not use deadtime in my proto PWM amp and the low side mosfet was heating but not the high side as well as my bench psu and this was worse due to the switching frequency and number of cross conduction it had to handle. My drivers however remained cool. Did you try to mod this circuit to use only n-channel in the high side? One reason in my opinion why p channel fet are used is that the drive signal can be referrenced to the positive rail which throws HV floating supplies and bootstrap supplies out of the way. I will see how far i can go with discrete gate drivers but i have seen implementations using LM311+ LTP & level shifter with IR2110 with can go near or above thousands of watts. In case you have any circuit i would be glad to have a look in case you don't mind sharing.thanks.

 

[norazmi]

Good zeus, nice to hear that, i do before sharing idea and making pwm class D discrete with other person on diyaudio.com. we work together but the advantage is hes got scope and i dont so i just can simulate it with ltspice and we finish that amp can deliver 100wrms at 8 ohm this one have no heating at all for discrete component, its long time project i think about a year a go. That project is a gift to my student for his final year project at another college in here. i`ll try to find the schematic since i dont keep it but i can ask my student for it. But this also run with nch and pch mosfet. its good if you can implement with both Nchannel mosfet .

If you want i can follow your experimental in class D including to build it and test but remind that i dont have scope . hope u have it lol. well we can go with your thread abd i`ll join there.

 

[stewin]

hi nozrami thanks for sharing your project . is your amp clear in the mid section and hf section i am thinking of building this amp . BTW can the amp you suggested using ir 2110 gate driver can it use irp240,250 or 260

 

[zeus_threat]

Hi noramzi ok no prob but be careful with what i will post there. I don't plan to get anything into a final pcb version in short time but rather learn start into a learning curve until i reach a good balance between stability/performance/efficiency.

 

[norazmi]

hi nozrami thanks for sharing your project . is your amp clear in the mid section and hf section i am thinking of building this amp . BTW can the amp you suggested using ir 2110 gate driver can it use irp240,250 or 260

IR2110 class D is scale able.

yes, it can drive irfp240 250 and 260, but u need to adjust gate resistor for suiteable with mosfet rds-on, in class D u need to know that low rds-on is better and take a look at datasheet also, watch at Qrr and Trr for better performance, good efficiency and low thd. for the chips mosfet u can go with irf540 and if u need more power bridge it with irf540, if u want run with low impedance and good efficiency go with irf3710, btw u can always go with more than 50vdc supply and you just need to choose the correct output mosfet to drive it or you will have high distortion. but still irfp240-260 can survive and the quality of sound is different with diff mosfet u used.

mid clear, high clear and bass of course your ears cant hold it lol btw i`ve little problem with my ears before when drive ir2110 + irf540 in bridge mode and now i`m wear some protection when im testing sound system for safety reason.