2 kwts discreet class d need help

[stewin]

hi nozrami i told you this amp is very powerfull .BTW if you think the power is little just bridge it will drive your load comfortably. thanks thanks for your your simulated schema . I've been testing it today with my smps due to the noise reduction it is loud and very powerfull clear even in the mid section . yes there is small amount of noise if you put your ears very close i.e next to the speakers i'll post pictures as soon as i get a camera . i did as you suggested and put a variable resistor (1k) in series with a 100 ohm resistor if the sum total resistance is 330ohms it works well at low volumes but there is some distortions if driven hard the fets are cool (IRFP250) . but if i set the resistance at 310ohms the fets(IRFP250) warm get hot and require a bigger heatsink but the sound is clear and high fidelity at high output levels . here is the pcb and schema i used. boards sizes are 194.5mm x 49.5mm

 

[norazmi]

yep i have an idea how to bridge it, i`ve hypex bridge schema, topology is almost same, but halfbridge is loud enough for me.

that resistor is for duty cycle. As u can read on phillips apps note, they already state there, to get harmonic sound is to lower that value but the cost is mosfet will heat fast and risk of blown mosfet if play hard. so better do it at a middle. But stewin my amp does not have any noise, and mosfet cool all the time, Stewin what you need to do is watching aux supply is enough to drive Hi and LO side to avoid distort what to do is you can play with gate resistor to get more clear sound until found suiteable gate resistor for each mosfet. For now my mohawk speaker rated 240wrms 8 ohm can hold the power of +/- 50vdc by using IRFP140N

regards.

 

[irvynejay]

very cool design and easy to get parts i hope i can also make this one to test,,can i ask whats the information /details of the amp pls

 

[norazmi]

Yes irvy u can make this amp, i need to ask stewin for the part correction first. the pdf schema and board have several wrong for the component, its need to be change for better performance and eliminated noise to 0%. I can guaranteed with my adjustment the noise will be totally dissapear and the quality of the amp? Its actually depends on which mosfet u use, at least irf540 or IRFP250N or better IRF540N and IRFP140N and more better mosfet still available such as IRFB4710, IRFB4020 and so on. Efficiency is almost 92% and duty cycle will be => 50%. Im really satisfied with this amp. Bass tight and mid and high is really nice . its produce very clear sound.

There is several thing we need to prepare to power up this amp, i`m using 50va 24 0 24 vac for the startup this amp and for testing to avoid mosfet blown. and set duty cycle at 1k ohm then turning it down until this amp oscilate. It will produce sound. But at the minimal duty cycle this amp still provide distortion at high volume, so u need to measure resistance for the duty cycle and slowing turning down and watch if the mosfet heat. For example, your trimmer is set to 440 ohm to get this amp oscilate and produce sound, and maybe at 400 ohm is the middle top duty cycle this amp will provide crystal clear sound and mosfet does not overheat, too low for the trimmer will cause the mosfet heat fast and risk of blown ur mosfet. So i hope u get what i mean.

Irvy if u want to build this amp i can guide u, u can use Stewin pcb which good enough for the gain and include op-amp and i`ll explain for the part change.

This amp really pawn, with +/- 50 vdc by using harman kardon linear transformer , my 240 wrms 8 ohm mohawk coil warm up at 3/4 volume within 3 mins of playing, so i decrease the volume to avoid my speaker burn lol i dont want this hi quality speaker burn, but tested with 15 inch 500 wrms 8 ohm is fine even with full volume, i have up the gain by using external op-amp i`ve made before, its enoug to get full gain. Im using +/- 50 vdc with IRFP140N and IRF540 and IRF540N, IRFB4710 all is ok with this voltage. And i already tested with IRFB23N15D with +/-65 VDC still this amp can manage without part change, but aux resistor need to be change R2 will be 6.8K 3 watt or 2 watt. good heatsink is really need to the aux transistor at this supply to avoid burn. but for the mosfet they still steady and run smooth, no overheat, burn or etc... Will make new board to have it stereo and mount it on new box hehehehe


regards.

 

[capsighter]

Wow another class d project for irvynejay.. good luck... nice explanation @norazmi hope you can post some bridge version for this amp...

Regards...

 

[michelle]

discrete ucd

Hi Norazmi
greetings is it possible to share pcb of discrete amp you showed on you tube lol to many IR ics
gone looking forward to making it

warm regards
michelle

 

[norazmi]

Yep, Full Bridge is possible for this amp, but i didnt test it yet, if anyone wanna try it i can post full birdge schema and for Michelle, dont use my pcb because its ugly hehehehe , Stewin PCB is better and have additional op-amp for the gain. I need to talk with Stewin when he s free, now i think hes busy with student exam. I may include OVP, OCP and startup delay for this amp by using 555 timer and pic817 opto, this protection come from classdfromru and i`ve tested its working allright but the cost is pcb will be more big to have additional space for the protection.
So we will wait until stewin come back then i can ask him to change component value and post working pcb with new schema.

regards.

 

[capsighter]

Thanks for your reply norazmi can you share your Schematic i try to make a PCB layout of your Full Bridge amp....

Regards...

 

[norazmi]

Yep sure capsighter, are u good of making pcbs? , i`m too lazy to make nicely pcb layout, my bad. i`ll draw it tonite for full bridge ucd, starting voltage will be +/- 30vdc but remember in full bridge it will double that power and u really need good psu lol to supply enough current.

Btw i`m gonna to upload some video discrete amp vs. half bridge smps, to prove and evidence that my design have 0% noise hehe, u may watch it by urself i`m taking 2 video for it.

regards

 

[capsighter]

Thanks a lot norazmi... im not really good in PCB designing but ill try my best to make a good one... just need more time in making some pending projects now because im facing some problems of my computer the mainboard is smoking...

 

[norazmi]

here is test with smps, proven no noise at all

[video=youtube;H2OYV6yNqIE]http://www.youtube.com/watch?v=H2OYV6yNqIE[/video]

 

[norazmi]

Full Bridge

here u go, full bridge with discrete ucd amp, anyone willing to make it one, you should use my suggestion voltage, after success i`ll try to help as i can, we can go with more higher voltage, but i think full bridge with +/- 40 to +/- 45 will be loud enough, estimated psu current must supply at least 25 amp.

 

[norazmi]

Full Bridge Correction

Sorry this is correction for the schematic,

 

[norazmi]

Second video running with smps filterless still sound crystal clear.

[video=youtube;2SrT-G1L9Ls]http://www.youtube.com/watch?v=2SrT-G1L9Ls[/video]

 

[stewin]

hi nozrami and others sorry internet issues but i,m back. is your fullbridge silent ? i was thinking overkill like using your fullbridge schema with irfp250/260 at +/-70vlts . BTW did the plan work of 2aux hi and low side aux 12v transistors , and do double aux transistors heat as much as when using one. did distortion reduce and power increase as a result of using dual aux transistors? the amp i last posted is still working loud and clear +/-42vlts irfp250 at 6ohms(two 8inch full range cabinets wired in series)

 

[norazmi]

Hi Stewin,

in simulation it works, like i said, i dont build it yet, but someone asking me for full bridge, and dont go +/- 70vdc with IRFP250N which cannot take the stress at that supply. because in fullbridge will draw much current, event i`ve tested with +/- 65 vdc and aux transistor hot like crazy and need big heat sink for it or using fan, so thats why i include 2 aux transistor and its work in LTSpice, distortion is depend on what mosfet u use and enough supply for aux also will give effect on it.

Stewin last schema ur using totem pole at the mosfet gate? Because on simulation those bjt transistor bds getting hot and provive about 4-6% THD at full power and without them distortion only 0.08-0.1% at full power. for the last schema u posted, u can increase the rails voltage at +/- 68 vdc and aux transistor cannot be tip41c and must be higher voltage too, R2 must be 6.8k 2-3 watts.

Last stewin please some picture of ur amp with last schema. make it clear picture and also please include pcb art work and silk screen.

 

[stewin]

yeah i noticed about the gate drive bjt s do increase distortion . but the reason i like them is because when i use them with irfp250 they run cooler than not using them and i think i can parallel them, the 250's to go to +/-70 in bridge mode

 

[norazmi]

do not think lol, i will simulate it, lets see what the result with parallel mosfet.

 

[norazmi]

Stewin, its work, totem pole must be 2n5401 and 2n5551 so i refer on ur first schema , R30, R15 = 10 ohm, R24, R14 = 10 ohm, R23,R16 = 15k, now simulate is not finish yet, i`ll post the result here

 

[norazmi]

Starting source stepping with srcstepmethod=0
Source Step = 3.0303%
Source Step = 9.84848%
Source Step = 17.4242%
Source Step = 25%
Source Step = 32.5758%
Source Step = 40.1515%
vernier = 0.375
Source Step = 50.7576%
Source Step = 62.1212%
Source Step = 73.4848%
Source Step = 84.8485%
Source Step = 96.2121%
vernier = 0.5625
Source stepping succeeded in finding the operating point.

Fourier components of V(out)
DC component:0.140376

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+02 2.870e+01 1.000e+00 -148.20° 0.00°
2 2.000e+02 6.768e-02 2.358e-03 9.06° 157.25°
3 3.000e+02 1.748e-01 6.088e-03 96.04° 244.24°
4 4.000e+02 3.912e-02 1.363e-03 21.80° 169.99°
5 5.000e+02 4.178e-02 1.456e-03 -6.31° 141.89°
6 6.000e+02 1.671e-02 5.822e-04 -9.27° 138.93°
7 7.000e+02 1.031e-02 3.591e-04 -18.21° 129.99°
8 8.000e+02 1.402e-02 4.884e-04 -0.48° 147.72°
9 9.000e+02 1.055e-02 3.674e-04 -1.77° 146.43°
Total Harmonic Distortion: 0.688817%


Fourier components of V(in)
DC component:-1.40235e-009

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+02 2.000e+00 1.000e+00 -0.00° 0.00°
2 2.000e+02 2.382e-09 1.191e-09 141.37° 141.37°
3 3.000e+02 2.473e-09 1.237e-09 75.91° 75.91°
4 4.000e+02 1.030e-09 5.151e-10 14.57° 14.57°
5 5.000e+02 8.334e-10 4.167e-10 -57.82° -57.82°
6 6.000e+02 1.650e-10 8.250e-11 -87.30° -87.30°
7 7.000e+02 1.048e-10 5.239e-11 -111.48° -111.48°
8 8.000e+02 1.517e-10 7.583e-11 -78.77° -78.77°
9 9.000e+02 3.148e-10 1.574e-10 159.04° 159.04°
Total Harmonic Distortion: 0.000000%



Date: Wed Mar 07 06:49:54 2012
Total elapsed time: 697.338 seconds.

tnom = 27
temp = 27
method = modified trap
totiter = 7326494
traniter = 7323905
tranpoints = 2019822
accept = 1431765
rejected = 588057
matrix size = 116
fillins = 232
solver = Normal
Matrix Compiler1: 27.3 KB object code size 10.6/6.9/[2.9]
Matrix Compiler2: off [3.7]/4.7/20.5