Single 300volt rail Class D amplifier?

[ThyDntWntMusic]

would it be feasible, to run a class D amplifier off a single 300volt or so rail.

the maths suggests that this would yield about 2500watts of possible output power into 4 ohms, obviously half that into 8 ohms.

is there any class D amplifier IC out there, which could be used to control some high voltage mosfets?

the IC would obviously be run off lower voltage.

National Semiconductor have an IC, except I believe it uses split rails, which wont work in this situation.

thanks for any input.

 

[y8s]

out of my league but maybe here's some starter information:

http://www.tripath.com/downloads/TK2350.pdf
http://www.tripath.com/downloads/RB-TK2350.pdf

http://www.tripath.com/downloads/TA3020.pdf
http://www.tripath.com/downloads/RB-TA3020.pdf
http://www.tripath.com/downloads/BRBTA3020.pdf

data taken from
http://www.tripath.com/data.htm
and
http://www.41hz.com/

Matt

 

[MikeHunt79]

Why do you need so much power?

Surely 60v rails should provide plenty of power to almost any speaker...

 

[luka]

Hi

http://www.tripath.com/downloads/RB-TA0105A.pdf

Note that none of them run of single rail, amp would have to be bridge to do that, but when going into bridge more, you have 4 times as much power as you do in single channel, same load.

You are way better off buying one, trust me on this

 

[Tim__x]

They already told you in the solidstate forum that this is a bad idea.
Isolation is extremely important.

 

[Dave Z]

Sooo... you are going to directly rectify the 220v AC wall outlet (I'm assuming you have 220V in Australia and the math works for it) and power an amplifier from it? If you do please make sure the fire department and paramedics are at your house when you turn it on for the first time

Oh... video, don't forget to take video of it. Hire a professional cameraman to video when you turn it on. And make sure he continues to film them as they resuscitate you

Dave

 

[ThyDntWntMusic]

Hi

http://www.tripath.com/downloads/RB-TA0105A.pdf

Note that none of them run of single rail, amp would have to be bridge to do that, but when going into bridge more, you have 4 times as much power as you do in single channel, same load.

You are way better off buying one, trust me on this

does it HAVE to be bridged?

and who said anything about mains voltage?



an amplifier to give the kinda power I'm looking at may have +-120 - 150 volt rails anyway...

if it can be done with a single rail, it makes things a lot easier.

the aim, is to get about 2500watts rms, into 4 ohms.

this is more an "in theory" thing, than something I'll actually do.

 

[luka]

hi

if you want single rail yes, if dual/split like +/- then that chip is ok for you...but not for 2.5kw...I am sure you know that, since you won't get that good fets...or if you do...well don't know how it would all work...even higher voltages were used by someone I know, not one but two...

 

[MikeHunt79]

the aim, is to get about 2500watts rms, into 4 ohms.

this is more an "in theory" thing, than something I'll actually do.

IMO, you would be better off using 5 x 500W RMS amps... Unless you have a single driver that will handle 2000W+...

 

[Dave Z]

OK so now that I know this is just "theory", I would use a single 200V (approx) supply and bridge the output into 8 ohms. You can use one of the high voltage, high and low side gate drive ICs from International Rectifier for each half bridge. Open loop (other than the normal issues associated with open loop) would be OK in this configuration but if you add feedback the high DC voltage from the 50% duty cycle on each half bridge is where you may see some problems. Of course this is theory, in practice you will have tons of issues with layout, parasitics etc with high current and high switching freq. Laugh if you want, but you could even use one of those little 5V 1W-2W filterless class D amps from TI as the modulator and just add the feedback but keep the DC in mind.

Dave

 

[ThyDntWntMusic]

Laugh if you want, but you could even use one of those little 5V 1W-2W filterless class D amps from TI as the modulator and just add the feedback but keep the DC in mind.

Dave

I was considering using one of these as the controller circuit:

http://cgi.ebay.com.au/2watt-D-Clas...8503410QQihZ013QQcategoryZ64570QQcmdZViewItem

would that be doable?

would it be possible to use one of these to run some fets, off a higher rail voltage? or, would I still need split rails?

 

[Ouroboros]

There is a real market requirement out there for a class-D amplifier design that will provide an output voltage of 350V pk-pk or so. Commercial amplifiers available to feed 100V line systems for PA and voice-alarm ceiling speakers are constrained in size by the step-up output transformers required to feed the 100V rms signal to the speaker line. (I know 100V rms is only 283V pk-pk but headroom needs to be allowed for real audio waveforms).

Of course, for this type of application the effective load for a 200W amp will only be 50 Ohms, so the current levels are quite low. A high-voltage amp to feed 8 Ohms would be a very different matter!

 

[Eva]

I'm currently doing 400V class D. Check this thread out:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=103826&highlight=

The prototype shown in that thread is more or less working (there are some faults to fix), but development is currently in standby because I have to complete other projects first.