Forums

DC Coupled Soundcard DAC Mod

Started by Ken Morrow October 13, 2015
I want to generate a signal in my computer that has a DC offset and
output them to the real world.

Unfortunately, the DC offset does not pass through my soundcard.

So I found this "solution" online.

http://www.whence.com/soundcard-dc-dac/

What the article does not state is that the large chip on the circuit
board gets very hot trying to pump out DC instead of the bipolar
signal it was desgned to produce.

The author suggests a 100R in paralle with the output cap. So I then
tried upping this to 10K which provided less but still acceptable
shift, but still caused heating.

I then tried a 1000Uf cap in parallel, which again caused heating.

However, these are cheap and handy ltittle devices.

My question is, apart from clamping a large heatsink on the
overheating chip, and accepting as little DC shift as possible, is
there any way I could safely use this particular USB soundcard to
output a DC offset signal?

BTW my application requires me to use the computer to produce the
signal, not a function generator, etc.

Ken Morrow

On Wed, 14 Oct 2015 12:42:09 +1100, Ken Morrow <kmorrow@kemptrail.com>
wrote:

> >I want to generate a signal in my computer that has a DC offset and >output them to the real world. > >Unfortunately, the DC offset does not pass through my soundcard. > >So I found this "solution" online. > >http://www.whence.com/soundcard-dc-dac/ > >What the article does not state is that the large chip on the circuit >board gets very hot trying to pump out DC instead of the bipolar >signal it was desgned to produce.
That chip outputs a unipolar signal, always positive.
> >The author suggests a 100R in paralle with the output cap. So I then >tried upping this to 10K which provided less but still acceptable >shift, but still caused heating. > >I then tried a 1000Uf cap in parallel, which again caused heating.
Maybe it always runs hot.
> >However, these are cheap and handy ltittle devices. > >My question is, apart from clamping a large heatsink on the >overheating chip, and accepting as little DC shift as possible, is >there any way I could safely use this particular USB soundcard to >output a DC offset signal? > >BTW my application requires me to use the computer to produce the >signal, not a function generator, etc. > >Ken Morrow
I can't see why the mod would make the chip get hot. What is your load?
On Tue, 13 Oct 2015 19:25:12 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

>On Wed, 14 Oct 2015 12:42:09 +1100, Ken Morrow <kmorrow@kemptrail.com> >wrote: > >> >>I want to generate a signal in my computer that has a DC offset and >>output them to the real world. >> >>Unfortunately, the DC offset does not pass through my soundcard. >> >>So I found this "solution" online. >> >>http://www.whence.com/soundcard-dc-dac/ >> >>What the article does not state is that the large chip on the circuit >>board gets very hot trying to pump out DC instead of the bipolar >>signal it was desgned to produce. > >That chip outputs a unipolar signal, always positive. >
OK. It must be the DC offset that is causing the problem then. IOW the chip is being asked to output more current than it is designed for.
>> >>The author suggests a 100R in paralle with the output cap. So I then >>tried upping this to 10K which provided less but still acceptable >>shift, but still caused heating. >> >>I then tried a 1000Uf cap in parallel, which again caused heating. > >Maybe it always runs hot. >
No it only happens when I make these mods to the board. The more the DC offset, the hotter it runs.
>> >>However, these are cheap and handy ltittle devices. >> >>My question is, apart from clamping a large heatsink on the >>overheating chip, and accepting as little DC shift as possible, is >>there any way I could safely use this particular USB soundcard to >>output a DC offset signal? >> >>BTW my application requires me to use the computer to produce the >>signal, not a function generator, etc. >> >>Ken Morrow > >I can't see why the mod would make the chip get hot. >
See my guess above.
>What is your load? >
Just a scope probe at this point. Ken Morrow
On Wed, 14 Oct 2015 13:49:26 +1100, Ken Morrow <kmorrow@kemptrail.com>
wrote:

>On Tue, 13 Oct 2015 19:25:12 -0700, John Larkin ><jjlarkin@highlandtechnology.com> wrote: > >>On Wed, 14 Oct 2015 12:42:09 +1100, Ken Morrow <kmorrow@kemptrail.com> >>wrote: >> >>> >>>I want to generate a signal in my computer that has a DC offset and >>>output them to the real world. >>> >>>Unfortunately, the DC offset does not pass through my soundcard. >>> >>>So I found this "solution" online. >>> >>>http://www.whence.com/soundcard-dc-dac/ >>> >>>What the article does not state is that the large chip on the circuit >>>board gets very hot trying to pump out DC instead of the bipolar >>>signal it was desgned to produce. >> >>That chip outputs a unipolar signal, always positive. >> > >OK. It must be the DC offset that is causing the problem then. IOW the >chip is being asked to output more current than it is designed for. > >>> >>>The author suggests a 100R in paralle with the output cap. So I then >>>tried upping this to 10K which provided less but still acceptable >>>shift, but still caused heating. >>> >>>I then tried a 1000Uf cap in parallel, which again caused heating. >> >>Maybe it always runs hot. >> > >No it only happens when I make these mods to the board. The more the >DC offset, the hotter it runs. > >>> >>>However, these are cheap and handy ltittle devices. >>> >>>My question is, apart from clamping a large heatsink on the >>>overheating chip, and accepting as little DC shift as possible, is >>>there any way I could safely use this particular USB soundcard to >>>output a DC offset signal? >>> >>>BTW my application requires me to use the computer to produce the >>>signal, not a function generator, etc. >>> >>>Ken Morrow >> >>I can't see why the mod would make the chip get hot. >> > >See my guess above. > >>What is your load? >> > >Just a scope probe at this point. > >Ken Morrow
If you're driving a scope probe, adding 100 ohms, or 10K, across an output coupling cap does nothing to the chip. Makes no sense to me.
On 10/13/2015 10:49 PM, Ken Morrow wrote:
> On Tue, 13 Oct 2015 19:25:12 -0700, John Larkin > <jjlarkin@highlandtechnology.com> wrote: > >> On Wed, 14 Oct 2015 12:42:09 +1100, Ken Morrow <kmorrow@kemptrail.com> >> wrote: >> >>> >>> I want to generate a signal in my computer that has a DC offset and >>> output them to the real world. >>> >>> Unfortunately, the DC offset does not pass through my soundcard. >>> >>> So I found this "solution" online. >>> >>> http://www.whence.com/soundcard-dc-dac/ >>> >>> What the article does not state is that the large chip on the circuit >>> board gets very hot trying to pump out DC instead of the bipolar >>> signal it was desgned to produce. >> >> That chip outputs a unipolar signal, always positive. >> > > OK. It must be the DC offset that is causing the problem then. IOW the > chip is being asked to output more current than it is designed for. > >>> >>> The author suggests a 100R in paralle with the output cap. So I then >>> tried upping this to 10K which provided less but still acceptable >>> shift, but still caused heating. >>> >>> I then tried a 1000Uf cap in parallel, which again caused heating. >> >> Maybe it always runs hot. >> > > No it only happens when I make these mods to the board. The more the > DC offset, the hotter it runs. > >>> >>> However, these are cheap and handy ltittle devices. >>> >>> My question is, apart from clamping a large heatsink on the >>> overheating chip, and accepting as little DC shift as possible, is >>> there any way I could safely use this particular USB soundcard to >>> output a DC offset signal? >>> >>> BTW my application requires me to use the computer to produce the >>> signal, not a function generator, etc. >>> >>> Ken Morrow >> >> I can't see why the mod would make the chip get hot. >> > > See my guess above. > >> What is your load? >> > > Just a scope probe at this point.
When you say "hot", how hot? Does the chip burn your finger if you hold it on the chip? If so, how long does it take? If you can keep your finger on the chip for a couple of seconds, it is likely not too hot for the chip. -- Rick
On 14/10/2015 03:49, Ken Morrow wrote:
> On Tue, 13 Oct 2015 19:25:12 -0700, John Larkin > <jjlarkin@highlandtechnology.com> wrote: > >> On Wed, 14 Oct 2015 12:42:09 +1100, Ken Morrow <kmorrow@kemptrail.com> >> wrote: >> >>> >>> I want to generate a signal in my computer that has a DC offset and >>> output them to the real world. >>> >>> Unfortunately, the DC offset does not pass through my soundcard. >>> >>> So I found this "solution" online. >>> >>> http://www.whence.com/soundcard-dc-dac/ >>> >>> What the article does not state is that the large chip on the circuit >>> board gets very hot trying to pump out DC instead of the bipolar >>> signal it was desgned to produce. >> >> That chip outputs a unipolar signal, always positive. >> > > OK. It must be the DC offset that is causing the problem then. IOW the > chip is being asked to output more current than it is designed for.
The chip outputs the same voltage as it ever did it is just that the poor thing is now seeing whatever external load you are offering it. Either that or you have splashed solder across a land somewhere and it is now overloaded. I expect the chip would also run hot (although about half as hot) if you forced it to output a maximum amplitude square wave into the same load. You need a buffer or unity gain voltage follower to protect the chip from abusive external loads.
> >>> >>> The author suggests a 100R in paralle with the output cap. So I then >>> tried upping this to 10K which provided less but still acceptable >>> shift, but still caused heating. >>> >>> I then tried a 1000Uf cap in parallel, which again caused heating. >> >> Maybe it always runs hot. >> > > No it only happens when I make these mods to the board. The more the > DC offset, the hotter it runs. > >>> >>> However, these are cheap and handy ltittle devices. >>> >>> My question is, apart from clamping a large heatsink on the >>> overheating chip, and accepting as little DC shift as possible, is >>> there any way I could safely use this particular USB soundcard to >>> output a DC offset signal? >>> >>> BTW my application requires me to use the computer to produce the >>> signal, not a function generator, etc. >>> >>> Ken Morrow >> >> I can't see why the mod would make the chip get hot. >> > > See my guess above. > >> What is your load? >> > > Just a scope probe at this point. > > Ken Morrow
-- Regards, Martin Brown
On 14/10/2015 02:42, Ken Morrow wrote:
> > I want to generate a signal in my computer that has a DC offset and > output them to the real world. > > Unfortunately, the DC offset does not pass through my soundcard.
One way is to use a second channel output (or other half of stereo) to generate a PWM signal which when averaged provides your variable offset. See for example: <http://electronicdesign.com/embedded/sound-card-signal-generator-interface-adds-variable-offset-control> piglet
On Tuesday, October 13, 2015 at 9:42:25 PM UTC-4, Ken Morrow wrote:
> I want to generate a signal in my computer that has a DC offset and > output them to the real world. > > Unfortunately, the DC offset does not pass through my soundcard. > > So I found this "solution" online. > > http://www.whence.com/soundcard-dc-dac/ > > What the article does not state is that the large chip on the circuit > board gets very hot trying to pump out DC instead of the bipolar > signal it was desgned to produce. > > The author suggests a 100R in paralle with the output cap. So I then > tried upping this to 10K which provided less but still acceptable > shift, but still caused heating. > > I then tried a 1000Uf cap in parallel, which again caused heating. > > However, these are cheap and handy ltittle devices. > > My question is, apart from clamping a large heatsink on the > overheating chip, and accepting as little DC shift as possible, is > there any way I could safely use this particular USB soundcard to > output a DC offset signal? > > BTW my application requires me to use the computer to produce the > signal, not a function generator, etc. > > Ken Morrow
As others said that's weird, with a 10k load how can it draw too much current. But maybe there is some weird feedback path. What about if you reduce the amplitude below 3.3 (3.0) V or so? George H.
That is because when Mr. Kamal Mostofa copied a common laser show  modification  he decided OP-Amp buffer amplifiers were not needed, and his graphics quality showed it. Worse yet he did not give credit to those who developed this modification..

http://www.google.com/imgres?imgurl=http://laserpointerforums.com/attachments/f47/26535d1271607455-sound-card-dac-tutorial-correction-amp.png&imgrefurl=http://laserpointerforums.com/f47/sound-card-dac-tutorial-40569-3.html&h=480&w=800&tbnid=cjf3HX0dqRxQKM:&docid=MVwSs-gzxEwOkM&ei=XakeVrWHOcScgwSeu6TQBw&tbm=isch

Steve
Ken Morrow <kmorrow@kemptrail.com> wrote:
> > I want to generate a signal in my computer that has a DC offset and > output them to the real world. > > Unfortunately, the DC offset does not pass through my soundcard. > > So I found this "solution" online. > > http://www.whence.com/soundcard-dc-dac/ > > What the article does not state is that the large chip on the circuit > board gets very hot trying to pump out DC instead of the bipolar > signal it was desgned to produce.
You are not supposed to pump out large currents from such a chip. When you have a low resistance DC load, use a buffer amplifier.