12v 20 amps for a switching computer power supply?
I'm lost here
Are you wanting to create the power supply for the computer itself (hence the need for a 12v and 5v rail) or are you powering a 12v device that runs a computer?
Either way you probably wouldn't want to linear regulate it. You can parallel nationals simple switching regulators for 20 amps with a fair amount of ease.
Otherwise, in the 78xx datasheets theres a schematic that allows for a greater current capability using discrete transistors. This will make alot more heat than a switching regulator.
cheers
Matt
I'm lost here
Are you wanting to create the power supply for the computer itself (hence the need for a 12v and 5v rail) or are you powering a 12v device that runs a computer?
Either way you probably wouldn't want to linear regulate it. You can parallel nationals simple switching regulators for 20 amps with a fair amount of ease.
Otherwise, in the 78xx datasheets theres a schematic that allows for a greater current capability using discrete transistors. This will make alot more heat than a switching regulator.
cheers
Matt
Hmm, if you want a fanless regulated powersupply giving 20A,
you shouldn't go without switchmodetopology. But switchmode
doesnt use any topology like 7812.
Let's say you feed the powersupply with 15volt unregulated DC,
you have to dissipate 3v*20A = 60watts. Unless you use a seperate
case just for the supply, you need a big fan to get rid of that much heat.
And i'm not sure if 15volts is enough.
you shouldn't go without switchmodetopology. But switchmode
doesnt use any topology like 7812.
Let's say you feed the powersupply with 15volt unregulated DC,
you have to dissipate 3v*20A = 60watts. Unless you use a seperate
case just for the supply, you need a big fan to get rid of that much heat.
And i'm not sure if 15volts is enough.
fr0st said:12v 20 amps for a switching computer power supply?
I'm lost here
Are you wanting to create the power supply for the computer itself (hence the need for a 12v and 5v rail) or are you powering a 12v device that runs a computer?
Either way you probably wouldn't want to linear regulate it. You can parallel nationals simple switching regulators for 20 amps with a fair amount of ease.
------------------------------------------------------------------------------
The PS for the computer is a 12/5/3/3V compact unit that sits on the power supply socket. It is dc-dc convertor requiring 12V input. Hence 20A needed for peak 200W spec of the convertor. Might as well over engineer.
Other people seem to use 60W swtching units to popwer a 200W dc convertor!
------------------------------------------------------------------------MikeB said:
Yes separate case and big heat sink; I already have these.
I have silenced the dc-dc convertor that the PS will supply; don't want another switch mode unit to complicate matters.
My experience is that computer audio sound can be imp[roved considerably by doctoring and improving the PS. Getting rid ofn fans is another issue.
That computer SMPS powered from 12V may not require a 12V regulated input and thus a simple transformer plus diodes and capacitor would do the job
Anyway, never forget that the 12V powered SMPS uses higher peak currents, uses switching MOSFETs instead of bipolar transistors and uses more EMI-likely topologies so it will surely generate more EMI than any standard 120V-220V powered ATX PSU
Also I think 12V powered equipment does not have to pass as serious EMI tests as mains powered equipment, so your sound card is likely to be getting more supply line ripple
The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits
The insides of any computer is an EMI hell!!
And I'm not talking about vague audiophile stuff, I talk about listening to your hard disk/CR-ROM/DVD operations, video card refresh pulses, bus activity, fan ripple, even keystrokes... from the analog audio output of the sound card
Anyway, never forget that the 12V powered SMPS uses higher peak currents, uses switching MOSFETs instead of bipolar transistors and uses more EMI-likely topologies so it will surely generate more EMI than any standard 120V-220V powered ATX PSU
Also I think 12V powered equipment does not have to pass as serious EMI tests as mains powered equipment, so your sound card is likely to be getting more supply line ripple
The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits
The insides of any computer is an EMI hell!!
And I'm not talking about vague audiophile stuff, I talk about listening to your hard disk/CR-ROM/DVD operations, video card refresh pulses, bus activity, fan ripple, even keystrokes... from the analog audio output of the sound card
And I'm not talking about vague audiophile stuff, I talk about listening to your hard disk/CR-ROM/DVD operations, video card refresh pulses, bus activity, fan ripple, even keystrokes... from the analog audio output of the sound card [/B][/QUOTE]
--------------------------------------------------------------------------
I know. Taming the PS rails and shoielding ribbons bring great sonic benefits.
With fan noise, taming computer audio is a tedious business. But I like to have as hi a fi as possible when keyboarding!
--------------------------------------------------------------------------
I know. Taming the PS rails and shoielding ribbons bring great sonic benefits.
With fan noise, taming computer audio is a tedious business. But I like to have as hi a fi as possible when keyboarding!
fmak said:fr0st said:12v 20 amps for a switching computer power supply?
I'm lost here
Are you wanting to create the power supply for the computer itself (hence the need for a 12v and 5v rail) or are you powering a 12v device that runs a computer?
Either way you probably wouldn't want to linear regulate it. You can parallel nationals simple switching regulators for 20 amps with a fair amount of ease.
------------------------------------------------------------------------------
The PS for the computer is a 12/5/3/3V compact unit that sits on the power supply socket. It is dc-dc convertor requiring 12V input. Hence 20A needed for peak 200W spec of the convertor. Might as well over engineer.
Other people seem to use 60W swtching units to popwer a 200W dc convertor!
It's cheaper to use a gate-driver and a pair of MOSFETs than to use multiple "simple-switchers", or use one simple-switcher as a gate-driver. As you stack simple-switchers you are adding inductance and capacitance -- it may not seem to be a lot until you consider that di/dt is a pretty big number. it's something I asked Nat Semi about, they said it was OK, but not great, not something you would want to get passed UL.
the type of cooling (fanless?) depends on a lot of factors -- if you design the thing correctly it won't have to operate in a wind tunnel !
the type of cooling (fanless?) depends on a lot of factors -- if you design the thing correctly it won't have to operate in a wind tunnel ! [/B][/QUOTE]
---------------------------------------------------------------------------------
Whatever you do with fans, you can HEAR it/them. Fanless is best.
---------------------------------------------------------------------------------
Whatever you do with fans, you can HEAR it/them. Fanless is best.
design your project so that the heat sink is in position to take advantage of convection currents. (hot air rises through the fins, drawing in cooler air from below the heat sink.) Fin spacing should be about 1/4 in. apart. (just less than 1/2 cm)
BTW: From an engineering standpoint, I wish the standard system of measurement would just dissapear.
BTW: From an engineering standpoint, I wish the standard system of measurement would just dissapear.
missing the point
So, you are improving the powersupply, you still have a myriad of other sources of noise. The only solution is to have all your analog circuitry external to the computer case and on a separate supply.
I have to agree with Eva 'The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits'
I can't imagine that any improvement you will experience will come near the improvement that a similar investment in an external converter would bring.
So, you are improving the powersupply, you still have a myriad of other sources of noise. The only solution is to have all your analog circuitry external to the computer case and on a separate supply.
I have to agree with Eva 'The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits'
I can't imagine that any improvement you will experience will come near the improvement that a similar investment in an external converter would bring.
cunningham said:design your project so that the heat sink is in position to take advantage of convection currents. (hot air rises through the fins, drawing in cooler air from below the heat sink.) Fin spacing should be about 1/4 in. apart. (just less than 1/2 cm)
-----------------------------------------------------------------------------------
I agree. The most stupid thing in computers is the way that the processor isn't upside down and thermally bonded to the case.
As engineers, we in UK used to say that Electronics engineers find it much more difficult to pick up and use Mechanical engineering whereas Mechanicals tend to find it easier to use electronics.
Re: missing the point
I have to agree with Eva 'The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits'
I can't imagine that any improvement you will experience will come near the improvement that a similar investment in an external converter would bring. [/B][/QUOTE]
----------------------------------------------------------------------------------
I am trying to close the gap!. External routing isn't possible with the pci bus. If we want high quality integration od CD, DVDV audio, and the ability to playv 2496 wav and aiff files with upsampling on the fly, then you have little choice.
Internal upsampling and DSP spectrum correction can make computer audio sound good.
I have to agree with Eva 'The only way to go with computers is galvanically isolated SPDIF outputs [transformer or optical coupling] and using external DACs. Computer expansion cards were never intended for precision analog circuits, only for digital circuits'
I can't imagine that any improvement you will experience will come near the improvement that a similar investment in an external converter would bring. [/B][/QUOTE]
----------------------------------------------------------------------------------
I am trying to close the gap!. External routing isn't possible with the pci bus. If we want high quality integration od CD, DVDV audio, and the ability to playv 2496 wav and aiff files with upsampling on the fly, then you have little choice.
Internal upsampling and DSP spectrum correction can make computer audio sound good.
Pardon, internal up..what?, dsp spect... what?
The result of any digital processing is just a bitstream that is used to feed a DAC. This bitstream may even contain real-time configuration info for the DAC and should be transmitted out of the computer by means of SPDIF or any other digital audio bus with galvanic isolation
Sound cards aiming to the professional market use external modules containing multichannel DACs/ADCs and the rest of the analog circuits, while the digital stuff is left inside the computer in an expansion card
However, some of the lamest manufacturers [ie: Creative Labs] place all the DACs/ADCs and all the analog stuff inside the expansion card and use cheap noisy mixed-analog-digital volume controls, eqs and mixer, all integrated in a single IC, but sell external huge plastic boxes filled with just air triying to fool the customer into thinking that this is a serious sound card package due to its gold-plated connectors and its attractive name
Does anybody remember the last versions of Sound Blaster 16 and Sound Blaster AWE/32/64 (de)Value(d) cards? The entire sound card [mic.amp+mixer+ADC+DAC+DSP+bus.interface] was implemented in a single IC!!!
Anyway, main-board integrated sound chipsets are just fine for listening to MP3/WMA files and even the MPEG-audio used in DVD pictures. These chipsets have a conveniently high noise floor in order to mask all the noise, quantization and distortion
The result of any digital processing is just a bitstream that is used to feed a DAC. This bitstream may even contain real-time configuration info for the DAC and should be transmitted out of the computer by means of SPDIF or any other digital audio bus with galvanic isolation
Sound cards aiming to the professional market use external modules containing multichannel DACs/ADCs and the rest of the analog circuits, while the digital stuff is left inside the computer in an expansion card
However, some of the lamest manufacturers [ie: Creative Labs] place all the DACs/ADCs and all the analog stuff inside the expansion card and use cheap noisy mixed-analog-digital volume controls, eqs and mixer, all integrated in a single IC, but sell external huge plastic boxes filled with just air triying to fool the customer into thinking that this is a serious sound card package due to its gold-plated connectors and its attractive name
Does anybody remember the last versions of Sound Blaster 16 and Sound Blaster AWE/32/64 (de)Value(d) cards? The entire sound card [mic.amp+mixer+ADC+DAC+DSP+bus.interface] was implemented in a single IC!!!
Anyway, main-board integrated sound chipsets are just fine for listening to MP3/WMA files and even the MPEG-audio used in DVD pictures. These chipsets have a conveniently high noise floor in order to mask all the noise, quantization and distortion
"However, some of the lamest manufacturers [ie: Creative Labs]"
YEP !!
In a single word: All audio signals generated inside a computer can't
be high quality. (that's 11 words or so)
That's why professional soundcards always use external DACs.
No dsp or other thing can rescue the signals generated inside the "EMI hell"
The easiest way to have external DAC is using SPDIF.
But even this is bad with normal soundcards. As described in another
threat, the 44.1khz signal from the mp3/cd/wma is upsampled to 48khz,
then downsampled again to 44.1khz. What a nonsense...
And as long as computers don't get quiet again (like amiga & co),
it really does not make much sense to generate HQ audiosignals with them,
unless the pc is in another room.
Mike
"PC = HQ-Audio" is like "X^2 = -1"
YEP !!
In a single word: All audio signals generated inside a computer can't
be high quality. (that's 11 words or so)
That's why professional soundcards always use external DACs.
No dsp or other thing can rescue the signals generated inside the "EMI hell"
The easiest way to have external DAC is using SPDIF.
But even this is bad with normal soundcards. As described in another
threat, the 44.1khz signal from the mp3/cd/wma is upsampled to 48khz,
then downsampled again to 44.1khz. What a nonsense...
And as long as computers don't get quiet again (like amiga & co),
it really does not make much sense to generate HQ audiosignals with them,
unless the pc is in another room.
Mike
"PC = HQ-Audio" is like "X^2 = -1"
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.