diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   SWPS (http://www.diyaudio.com/forums/solid-state/1144-swps.html)

Mkulikow 14th November 2001 08:38 PM

Question,

If you pull a power supply out of a 200W car amplifier to use what is the difference between that and a 1kW amplifier? Can you change the values and make a new coil to carry more current or what, I am sure it depends on the frequency and the pwm scheme but can;t I just rape a shitty amp and kick it up a notch?

- Mike

Nelson Pass 14th November 2001 09:39 PM

1) 800 watts :)

2) Forget it. The difference between 200 and 1000
watts is not a notch.

Mkulikow 14th November 2001 09:43 PM

OK Then
 
I need some ideas to get started then, I know I am gonna need tons of current unless I can impliment soft start circuitry and PSPICE is sick of me running boost converter simulations all day, where are the equations for these power supplies or is this why amps cost so much $$$?

- Mike

haldor 14th November 2001 10:54 PM

Designing a reliable, high power switching supply is not a trivial exercise even for EE's. Probably the most difficult part is specifing the inductor correctly. This is a very arcane field with a steep learning curve that doesn't leave much room for trial and error. If you get this part even slightly wrong your power supply will normally self destruct, often very spectacularly. Most switching supplies use inductors that custom designed for that particular supply.

If you want to see how this is done go to the QSC website and look at the schematics for the PLX1602 amp (click on the tab marked PDF's). This design is an 800 Watts per channel stereo amp that uses a switching supply (AC powered rather than 12 VDC), but this will give you an idea of the complexity doing something like this entails.

http://www.qscaudio.com/

Phil Ouellette

hifiZen 16th November 2001 09:29 AM

I can vouch for that! I've tried building switch mode power supplies twice, and all I can say is "yuk!" - and I'm an EE (though i was still a student at the time). I eventually got mine working, and the second time around was much easier, but i find the most difficult part is getting the EM and electrical noise down to acceptable levels. Getting the outputs filtered nicely isn't so easy, especially if you're shooting for efficiency and economy (and hence don't want too many lossy or heavy components like chokes in there).

Another amusing anecdote: we use switching supplies for development stuff at work. One of the company engineers (one rather more experienced than I) designed a nice SMPS for use in our dev kits. Anyway, we're now refitting our kits with COTS supplies from a third party... ;-)

One problem with SMPS units is that they don't become economical until you're well into the mass production range, at which point they become dirt cheap. Building them in small quantities can cost more than a much nicer linear supply. Admittedly, with DC-DC conversion, you've not got many options...

These days, I just avoid switch-mode supplies as much as i can. My advice: leave these bad boys to the pros... the pros who specialize in SMPS design.

Mkulikow 16th November 2001 12:27 PM

Well
 
I may of completed this exercise, I too am an EE undergrad working at IBM full time and I hace figured out the solution. At least from a simulation standpoint I have my 12V battery going into a center tapped xfrmr to give me +/-50V including coil resistance and capacitance. The current right now is limited to about 20A RMS per phase so 40A total. I have the switching FETs controlled by a NSC LM3488. Here's the trick, the transformer is more or less linear with respect each side of the center tap, I use the PWM module to control one half and the other half is driven by the same chip so the outputs are VERY well regulated, my question is, am I getting enough current yet?

AudioFreak 19th November 2001 10:43 AM

short answer ....... NO.

Mkulikow 19th November 2001 08:47 PM

Quote:

Originally posted by AudioFreak
short answer ....... NO.
OK then, I have gotten a lot of nowhere with this project and its killing me. I have a design that I could get upwards of 100A out of and people are telling me that's not enough, there is no way that these car amps draw that much current through only 8awg wire. The circuit I have has soft start circuits built in so there is no current surge, I need to know how much current I need to be able to maximize this design, the LM3488 frm NSC has no current limit since its driving an external FET. Can someone please tell me how much current I would need, anything, throw me a frikin bone here :) I want to put 300W rms into a 4 ohm load, bridged even, by my calculations with EXTREMELY BAD efficiency I would only need less than 25A worst case, where am I missing some critical information.

- Mike

Nelson Pass 19th November 2001 09:30 PM

With a switching supply, you pretty much have to
assume that the current rating will equal the peak
draw of the load, which for 300 watts rms will be
600 watts. Using I^2 R = 600, we get 12.24 amps
peak (per channel). Your +-50 volts will give you
just about the right voltage, and so there you are.
:)

Mkulikow 19th November 2001 09:33 PM

Quote:

Originally posted by Nelson Pass
With a switching supply, you pretty much have to
assume that the current rating will equal the peak
draw of the load, which for 300 watts rms will be
600 watts. Using I^2 R = 600, we get 12.24 amps
peak (per channel). Your +-50 volts will give you
just about the right voltage, and so there you are.
:)

Thank you god, I was starting to get the feeling from all of these people that to get 300W was going to take 100A or more, I knew something had to be wrong. I now officially have a SWPS for a 12V battery that can be variable voltage and current by changing 2 resistors. I am now trying to get this circuit on copper, I may have to place 2 boards on top of each other like a multilayer board built cheap. Thanx!!

- Mike


All times are GMT. The time now is 06:06 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2