That's what i thought myself when i saw someone's work here, i believe it was luka's. The answer is simple. You cut a gap, fill it with non-magnetic material and glue (i used pieces of old sandpaper as they tend to soak superglue nicely), you wait for it to dry, you cut another... and so on. Done right the core will be just as strong as it was before gapping. I accidentally dropped mine while winding and it didn't fall to pieces.
Nope, always been too lazy to wind one. The transformer was already wound but had only one 45v AC winding. I took it apart, removed half the wire, brought a tap out, then wound the wire back on. So i have two 22.5v AC windings.
I already have 100nF next to the MOSFETs and 47uF very close. After beefing up the wiring there are no more voltage imbalance problems, but again, i'm not using this transformer at even half what it can do. Since the SMPS will be unregulated, we'll see what happens there.
Of course it does matter. But only resistance matters hence how much current (amps) it handles, there's no such thing as smooth or rough current flow.
hi guys i use T130-2 in my class d power amp running at 450Khz LM311+74AC00+IR2110
(74AC00=deadtime and protections) efficiency is 92% @ 800W the sound is very clean
here is my earlier amp 06062010005.flv video by ionutzxpo - Photobucket selfoscilating @250Khz back then i did not have micrometals inductor TXX-2 i was using inductor from pc power supply and got very very hot even with 2mm gap (the T106-2,T130-2 and T200-2 wounded with litzwire is the most suitable for high power class d amps)
(74AC00=deadtime and protections) efficiency is 92% @ 800W the sound is very clean
here is my earlier amp 06062010005.flv video by ionutzxpo - Photobucket selfoscilating @250Khz back then i did not have micrometals inductor TXX-2 i was using inductor from pc power supply and got very very hot even with 2mm gap (the T106-2,T130-2 and T200-2 wounded with litzwire is the most suitable for high power class d amps)
@ norazmi - I i were supposed to get an email from you it isn't here. Hit me up with a PM.
That's because 1 gap isn't enough. I have 7 in my current one... I do hope the T106 needs only 6.
I'm curious whether this driving scheme can be implemented for full bridge... i'll give it a shot. I don't want to use any specialized ICs. Reason being a single rail power supply is easier to protect.
That's because 1 gap isn't enough. I have 7 in my current one... I do hope the T106 needs only 6.
I'm curious whether this driving scheme can be implemented for full bridge... i'll give it a shot. I don't want to use any specialized ICs. Reason being a single rail power supply is easier to protect.
I`m sorry late reply today i was busy, today i`ve spent little time about 30 mins before im login here, i just take out my blue inductor from ir2110 ucd and mount it to this cheap ucd, and i was impressed (with modified circuit) the sound is very low thd, no noise, big bass, middle and treble also high, sound balance to me for full range. on simu posted by Th3 uN1Qu3 , i was able to lowser thd as low as 0.04% 4 ohm load, i did test with 2 ohm load and notice no over heating at mosfet and its tolerate with that load. speaker coil warm and i can feel the cone heat when play bass i love u. play hard and big sound but my supply insufficient current and having voltage sag each rails, supply +-45 vdc 6 amp, 8x3300uf caps, filter is ferrite toroid blue one that i post image here, od is about 1.3 inch 16 awg coil wire, i would like to bridge it too, for testing purpose might try at low voltage first, let say 24 0 24 vac.
later i`ll post something here, waiting for Th3 uN1Qu3 maybe he has good idea with this amp, i just do experiment by read some class d information and other schematics so far non of my mosfet burn yet 😀. happy with it.
regards,
Azmi
later i`ll post something here, waiting for Th3 uN1Qu3 maybe he has good idea with this amp, i just do experiment by read some class d information and other schematics so far non of my mosfet burn yet 😀. happy with it.
regards,
Azmi
I got the email and will play with the schematic a little. I'll also tidy it up so it looks neater. 🙂 Replacing the compensation components (C6 and R13 as figured) is something i have also done myself, but lower C means higher oscillation frequency, it looks better in sim but has to be tested real-life to ensure that the MOSFETs switch properly. I also don't think that those D-S diodes are really needed, but will see.
I also have been playing mine on a full range speaker box today (well, it needs a bit of sub help but for normal listening it's alright), and it sounds great. I have to build the SMPS, need more power. 😀
I also have been playing mine on a full range speaker box today (well, it needs a bit of sub help but for normal listening it's alright), and it sounds great. I have to build the SMPS, need more power. 😀
Okay, this should be close to what norazmi has on his bench atm. Check attachment.
I've also finished the specs on my final version.
Coming soon. 😀
I've also finished the specs on my final version.
- It will deliver at least 350W/4ohm, 2ohm capability is not a design objective.
- Prototype will use 2x IRF9640 for top MOSFET and 1x IRF640 for bottom MOSFET. Reason being they're cheap and readily available.
- PCB uses all thru-hole parts and is designed around a heatsink scavenged from an ATX power supply.
- Board size is 120x80mm and is double sided, the top side being nothing but a big ground plane.
- It will use an unregulated SMPS, with overcurrent and overtemperature protection. Most likely the SMPS will be made using good ol' TL494 as controller and IRF740 as switching devices.
Coming soon. 😀
cool, i can see that on this one, will do modified today in real life with this value. and i will post some picture and video while its playing. and i cant wait for final schematics with 2 p channel and 1 n channel from Th3 uN1Qu3. with 350wrms/4 ohm is loud enough its per channel, and if u have stereo cant imagine that psu for this amp 😀 . Thanks to Th3 uN1Qu3 for his continue contributing mod on this amp, its really improve it with more efficiency and lower the thd and credit for modified version goes to him.
regards,
Azmi.
regards,
Azmi.
Put R3 back to 47 ohms btw, it eats up a little efficiency. At least that's what the sim says...
I'll buy plenty of close value resistors so i can test in real life. 🙂 And 350W @ 87% efficiency = 402W, so for two channels an 800W supply isn't THAT hard to make. Heck, i'm going to build one from junk... I only have one subwoofer so i will only need one channel, but i'll be making the PSU for two just in case. Or that's just an excuse to have some extra watts around. 😀 Bridging anyone?
Btw, the reason why we can't get crazy high efficiency with this amp is because there is zero dead time. It helps with sound quality but it sure doesn't help with heat. I'll try to figure out a way to insert some. I know that my crappy old stripboard actually had too much dead time with the same parts, due to high parasitics, and the MOSFETs were stone cold, so i know it's possible... just need to figure out how to do it with actual parts.
I'll buy plenty of close value resistors so i can test in real life. 🙂 And 350W @ 87% efficiency = 402W, so for two channels an 800W supply isn't THAT hard to make. Heck, i'm going to build one from junk... I only have one subwoofer so i will only need one channel, but i'll be making the PSU for two just in case. Or that's just an excuse to have some extra watts around. 😀 Bridging anyone?
Btw, the reason why we can't get crazy high efficiency with this amp is because there is zero dead time. It helps with sound quality but it sure doesn't help with heat. I'll try to figure out a way to insert some. I know that my crappy old stripboard actually had too much dead time with the same parts, due to high parasitics, and the MOSFETs were stone cold, so i know it's possible... just need to figure out how to do it with actual parts.
GOOD result, today tested with 20uh t106-2 feed with 1uf cap, speaker load 4 ohm, sound is powerfull only with small trafo 24 0 24 3a with 8x3300uf caps bank, i cannot test with 40 0 40 trafo yet before that one already mount on finish ucd case. will try to get new trafo or use my smps and trying to push 47 0 47 vdc 10 amp. i`ll try to load as minimum as i can but need to make sure this amp stable with every load. my target only mimimum is 2 ohm. 4 ohm no overheat at max vol only little warm, will post the video after finish at youtube the file is on the way now. ATTENTION there is no noise hehehe 😀.
Th3 uN1Qu3 dude 😀 can u post schematic for 2 pchannel vs 1 nchannel. i would like to test it.
regards,
azmi
Th3 uN1Qu3 dude 😀 can u post schematic for 2 pchannel vs 1 nchannel. i would like to test it.
regards,
azmi
here is the video YouTube - Video026
its low thd, using calibre pre-amp, no incoming noise, on/off supply rail smooth only little pop very little. dc offset when idle is about 20 mv, but my apologies i cannot measure it with osc for the frequency i dont have it yet, will order soon.
regards,
Azmi
its low thd, using calibre pre-amp, no incoming noise, on/off supply rail smooth only little pop very little. dc offset when idle is about 20 mv, but my apologies i cannot measure it with osc for the frequency i dont have it yet, will order soon.
regards,
Azmi
This video show up close to speaker mohawk 240wrms 8 ohm, i`ll test later with 10 inch sub audiobahn double coil double magnet at 2 ohm, first i need to power up my smps with rails +47 0 -47 vdc. NOTICE no overheat at mosfet, little warm.
here is the video:
YouTube - Video028
regards,
Azmi.
here is the video:
YouTube - Video028
regards,
Azmi.
Your new PCB looks a lot like mine. 😀 Yes, i will give new schem, just have to complete my power supply and give it a go myself. I'm very busy so progress is a little slow but i hope to have the power supply done by the end of the week.
I have the first power transformer done, it seems i chose the wiring size perfectly as it fit in the EI33 core but just barely. Primary is 62 turns with 8x 0.2mm wire twisted together (hey, litz wire is expensive), and secondary is 37 turns with 4x 0.35mm tapped at the 29th turn for a 18v supply for the linear regs. Remember this will be an unregulated supply so we need more turns than a regulated one would, because the transformer will see maximum duty cycle at nominal line voltage. Idle rails are expected to be +/-88v, we'll see what happens when i test this first transformer. I remember i also had a GDT lying around so all i have to do is wire everything together and see how it runs. The supply will work at 100kHz.
I have used the common technique of winding half the primary, winding the secondary, then winding the other half primary. I took some pics and will post them soon. 😀
I have the first power transformer done, it seems i chose the wiring size perfectly as it fit in the EI33 core but just barely. Primary is 62 turns with 8x 0.2mm wire twisted together (hey, litz wire is expensive), and secondary is 37 turns with 4x 0.35mm tapped at the 29th turn for a 18v supply for the linear regs. Remember this will be an unregulated supply so we need more turns than a regulated one would, because the transformer will see maximum duty cycle at nominal line voltage. Idle rails are expected to be +/-88v, we'll see what happens when i test this first transformer. I remember i also had a GDT lying around so all i have to do is wire everything together and see how it runs. The supply will work at 100kHz.
I have used the common technique of winding half the primary, winding the secondary, then winding the other half primary. I took some pics and will post them soon. 😀
aight, next time i`ll use N82 to record, seems 3110c picture is not clear and sound also low quality, becoz N82 is playing music while 3110c record it. Th3 uN1Qu3 i have 2 version of pcb and the another one small than this one. about half. its ok if you`re busy take time dude, we do this only for hobby, i`ve many power amp so far at my home, some at my friend house and at karaoke shop hehe 😀 so lets make em for collection our own creation of the pwm and of course DIYAUDIO.COM.
regards,
Azmi
regards,
Azmi
Nice videos norazmi.
two questions :
1- How can i get the THD value in ltspice?
2- How many watts are you getting with +-47 rails ?
regards!
two questions :
1- How can i get the THD value in ltspice?
2- How many watts are you getting with +-47 rails ?
regards!
Hey pat on back for Th3 uN1Qu3 and azmi for the work. I feel like talking a lot here... but..work at my end don't permit me this moment. dont worry at night i ll try to talk to u here.
- Status
- Not open for further replies.