500W Power Amplifier

[chatziva]

If I were you I would get an 400W or 700W amp from hypex.

much simpler for a first project compared to the options above

You will also need a proper softstart circuit. especially for the UcD700

good luck!

[Krisfr]

Your Instructor should have his head examed IF he thinks you COULD do this project.... That much power, and with construction would take a GROUP 6 months or more to do something that sounded decent.... Good Luck, I would be interested IF there was more time to do it RIGHT with digital computer controls and a SMPS WOW

[tiltedhalo]

I thought the hole idea behind "school" was to learn, so basically you want somebody else to design and layout a project for your test, or just copy someone else, right?
I could give you and I am sure a bunch of other people on here also could give you hundred different designs to accomplish your task, but what will you learn?
There are simple bipolar designs you could use in a bridge mode that could allow those power levels, IGBT's, MOS, JFET's, and you can even get 600W out of lm3886's into a 2 ohm load. do some research use some creativity and in the process you may learn something

[Dr.EM]

You could take a look at this amp which a member here built:

1000 Watt Sub Amp: Design / Build

But you'll see it's no easy challenge!

[ostripper]

By krisfr - Your Instructor should have his head examed IF he thinks you COULD do this project.... That much power, and with construction would take a GROUP 6 months or more to do something that sounded decent....

6 months ??? yikes You take any standard class B topology , augment the OP stage and power supply for high current operation and a 300-500 watt unit is not much harder than a simple 60w discreet amp. MJL's big sub-woofer amp and my "supersym" are just the symasym and the bootstrapped standard RCA designs with triple OP stages (or CFP's) and higher rails.

On my design , I hit a "wall" with the double EF (beta droop) , I also wanted to use the cheap NJW21193/4 output devices (bigger die/lower beta). With the double EF , 300w was the max using higher gain NJW0281/0302's as outputs ( very excellent devices for $1.40 each). With a very tame triple ,Self type 3 EF (hybrid triple)... any pointers ??

and any "front end" (LTP /VAS)... symasym - dx amp - standard self amp... and output device(s) for the main driver pair, I found I could parallel as many outputs as I wanted with NO effect on the former stages. The biggest issue in "supersizing" your standard medium power amplifier is 1.a more expensive , massive PS (1kva + toriod + 100,000uf ) ,10/12 gauge wire, 40amp bridges. 2. even being class B , massive heatsinking is required. 3 . with more dangerous voltages available , GREAT care must be observed in amplifier stability , fusing , tools !!! () .

I have my amp in it's final config now (triple powered) , all I need to do is add a few more pairs of the cheap NJW's to reach the magic 1/2 KW. It really was not hard to go from 2 OP pairs to 4 , but to go to 6 + took a little "new thread/LT" time.
OS

[kipman725]

I don't understand how building somone elses design is of education value apart from if the course is some kind of training to learn how to solder. Also I don't see why power output would be important as high power amplifiers are not more electronicly complex just far more dificult to actualy construct due to the larger currents and power disipation.

[ostripper]

I don't understand how building somone elses design is of education value apart from if the course is some kind of training to learn how to solder.

Nearly all the designs here on DIYA are "somebody elses" , the only "non-standard" ones are roenders RMI and syn's VSOP . Even these are at least partially derived from classic designs.
The educational value would best be derived by downloading LT spice http://www.linear.com/designtools/software/ltspice.jsp, seeing the "cause and effect" , loop gain responses , maybe even seeing an error (oscillation , rail sticking)
and then customizing an existing design for high power operation.

Beyond this , even with "somebody elses design" , if you decide to choose your own devices and lay the amp out yourself , there is a lot that one can learn (PCB grounding , component derating , etc.)
OS

[ostripper]

high power amplifiers are not more electronicly complex

Just an observation , going from small to big. Layout , basestopper values , current sharing among OP's, biasing ... all are much more critical for both the safety and final sonic character of the amplifier. There are more factors , but the list would be very long.
OS

[kipman725]

Originally Posted by ostripper

Nearly all the designs here on DIYA are "somebody elses" , the only "non-standard" ones are roenders RMI and syn's VSOP . Even these are at least partially derived from classic designs.
The educational value would best be derived by downloading LT spice http://www.linear.com/designtools/software/ltspice.jsp, seeing the "cause and effect" , loop gain responses , maybe even seeing an error (oscillation , rail sticking)
and then customizing an existing design for high power operation.

Beyond this , even with "somebody elses design" , if you decide to choose your own devices and lay the amp out yourself , there is a lot that one can learn (PCB grounding , component derating , etc.)
OS

Yes I see what you mean but all human achivment is ultimatly dirivative! modifying an existing design would indeed be educational. I can be abit stuffy as my education is very formal in presentation.

[ostripper]

by kipman - human achivment is ultimatly dirivative!

Exactly , I feel no shame in "deriving" the best ideas from this forum , manufacturer's , the internet ... and combining them to come out with something unique and educational that satisfies my needs.

To learn , one just needs to split the subject into (input stage / voltage amp / current stage ), throw it into the "blender" (LTspice) , and hopefully you won't get a "bang" out of the finished product .
OS