Author Topic: Someone needs to make a DC-coupled soundcard crowd-funded project  (Read 6444 times)

0 Members and 1 Guest are viewing this topic.

Offline Ben321

  • Frequent Contributor
  • **
  • Posts: 506
This is just a suggestion. I myself don't know the first thing about starting a crowd funded project. Hopefully, somebody will eventually make this.

My idea is for a sound card that is capable of digitizing any signal down to (and including) 0Hz, which is DC. It would work just fine for audio applications, but would also be able to work as an oscilloscope for electronic hobby applications. The real problem with most sound cards these days, is that they depend on a single voltage level, so the mid-point voltage has to be derived with a voltage divider, and the signal has to be AC-coupled. To be DC-coupled, it needs to use 2 supply voltages (+5 and -5, or +12 and -12) to power the signal amplifier and to provide the reference voltages for the ADC (analog to digital converter). It may seem that you would need an external power supply to provide the negative voltage level, but the fact is that ATX power supplies in most PCs have both positive and negative voltage levels. I think that you will find that most ATX power supplies have a -12V pin, and I think a -5V pin as well (If not, -5V can be generated from -12V using a -5V voltage regulator). So I believe that a PCI or PCI-e soundcard could easily be built to use DC coupling (assuming that the PCI and PCI-e buses have a line for negative voltage, which I believe they do).

I may be mistaken, but I believe that even though there are hundreds of soundcards available on the market, not a single one of them actually has a DC-coupled Line In port. All sound card manufacturers have taken the quick&easy option of making sound cards with a single supply voltage and AC-coupling.

In light of this fact, I think that a soundcard with a DC-coupled Line In port would be a great product to build via crowd funding. And even better would be if it also had a DC-coupled Line Out port.


Below are a couple schematics. The upper one shows a design for a typical Line In circuit on a sound card, with AC-coupling. The lower one shows a design for my proposed soundcard with a DC-coupled Line In port.


The term "data bus" in this diagram, can refer to any of a number of busses that computers can use for communicating I/O digital data with the CPU (such as ISA, PCI, PCI-e, Firewire, USB, etc).

Note that in my diagrams, I don't use any model numbers for any of the chips, and instead use textual descriptions. It will be up to the actual person who decides to go forward with my idea and makes the crowd funding project, to determine what actual components are needed.
 

Online blueskull

  • Supporter
  • ****
  • Posts: 13531
  • Country: cn
  • Power Electronics Guy
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #1 on: May 20, 2017, 03:15:38 am »
There won't be a market. Most EEs can afford a proper scope, and most non-EEs don't want DC component in their audio system, unless you want to blow up your headphone.
There may be some professional cards with pure DC signal path for some reason (say, better low frequency response), designed for those knowing what they are doing (professional recording engineers), but those won't be cheap.
 
The following users thanked this post: Kilrah

Offline amspire

  • Super Contributor
  • ***
  • Posts: 3787
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #2 on: May 20, 2017, 03:38:19 am »
You can get 8 bit oscilloscope kits from about $10 and proper Hantek 48MS/s 8 bit usb scopes for about $50. An 8 channel 12 bit 2.4MS/s Hantek 1008B is well under $100. That has a high enough sampling rate for audio and should be able to get down to about 0.05% distortion.

The idea of using audio channels or plug in cards in PC's for a cheap oscilloscope has been around about as long as PCs have been around and it still hasn't caught on. Can't see anything changing that.
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1417
  • Country: us
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #3 on: May 20, 2017, 06:41:35 pm »
I also don't think there is much merit in designing this for use as an oscilloscope, but it could have merit as a higher quality sound card. In either case, though, it might be best implemented with a USB interface, rather than PCI-e.

And no, the PCI-e slot does not supply a negative voltage rail - just 3.3V and 12V - so a dc-dc converter will be needed.

It might also be worthwhile to implement the line input with a fully floating isolation amplifier - extra robust and lower noise/risk of ground loops.

 

Offline Ben321

  • Frequent Contributor
  • **
  • Posts: 506
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #4 on: May 21, 2017, 11:13:26 pm »
I was thinking of it as a way of inputting analog sensor data (such as from a photo-cell) into a computer, rather than being used for testing circuits.  Both very slow and very fast changes of brightness from a photo-cell could be digitized, if there was no DC-blocking capacitor. With a DC-blocking capacitor only changes in light that are fast enough to produce frequency components above about 20Hz can be detected. I'm considering making my own scanning infrared detector (with spinning mirror taken out of a laser printer or barcode scanner), and using a single photosensor made out of exotic material (InAs, InGaAs, InSb, etc) for various exotic wavelengths of IR. Both SWIR and MWIR (longer than NIR that could be detected with a silicon CCD or CMOS sensor, but shorter than LWIR that can be detected with VOx) are wavelengths that require even more exotic materials to detect, and typical imagers in this wavelength range cost much more than LWIR imagers. Typical InGaAs 320x240 SWIR imager costs about $20000. A single InGaAs photodiode though only costs about $100, and with a spinning mirror scanner for the x-coordinate scan (and manually moving the device at a near constant speed with your hand for the y-coordinate scan) one can create a 2D image in this wavelength (or other exotic wavelengths, depending on the detector material) for very cheap prices.

And this is where the DC-coupled soundcard requirement comes in. I need a way to get the signal from such a scanner into my computer. I need a way to digitize the signal. AC-coupling is unacceptable, because there is always a DC offset in image data (pixels aren't supposed to ever have negative brightness). With AC-coupling, the digitized signal ALWAYS has an average signal level of 0 volts, meaning the signal is forced to dip into the negative range at times. There's no way to recover the missing DC component (as well as other nearby very low frequency components). This will ruin the the image that gets recorded.

I hope, for the sake of my scanning imager project being able to move forward, that somebody here at EEVBlog will decide to move forward with my crowd funded project idea for a DC-coupled sound card. When they do, I can then buy it from them. A DC-coupled sound card is probably the hardest part to get for my entire project, so rather than wasting my time hoping to find one for sale from normal sources, and rather than possibly ruining my own good sound card by trying to modify the circuit for DC-coupling, I've decided instead to try to convince EEVBlob members to start a crowd funding project for a DC-coupled sound card. If such a crowd funded project is in fact started, I can guaranty I will be one of the first people to buy the product.
« Last Edit: May 21, 2017, 11:15:56 pm by Ben321 »
 

Offline Ben321

  • Frequent Contributor
  • **
  • Posts: 506
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #5 on: May 22, 2017, 02:31:36 am »
Here's why it's important.

Lets say you have a scanning mirror with a rotational speed of 3000RPM, this is 50RPS. If the mirror that's being spun has 6 sides, this then gives you 300 scan-lines per second. For a 48000Hz sample rate sound card, that gives you 160 samples (pixels) per scan line. An image with a 4:3 aspect ratio would then have dimensions 160x120, and to scan 120 lines (at 300 lines per second) would take 0.4 seconds. This is only slightly faster than 2FPS. That's quite slow, and as a result low frequencies must not get blocked out, or you will have a bad image.

I've approximated the frequency response of my current AC-coupled soundcard, with the following equation:
wave(n+1)-wave(n) + wave(n-1)*0.99
wave is the signal value
n is the current sample number
wave(n) is the current signal value
wave(n+1) is the next signal value
wave(n-1) is the previous signal value, and it is dependent on the output of the filter operation from the previous sample

This is a simple IIR filter, and it closely approximates the frequency response of my sound card, when the sample rate is 48000Hz.


To test how this would effect an image. I loaded a Windows7 sample image (tulips.jpg) into Gimp, converted it to grayscale, and reduced it in size to 160x120 (using bilinear interpolation).
I then exported the raw data (no header, just the raw 8-bit grayscale values) from Gimp. I loaded it into GoldWave audio editor software (which supports loading and saving raw audio data). I then used the above IIR filter equation in GoldWave's equation evaluator feature (which allows almost any equation you can think of to be used to process audio data, just by typing it into the text field and pressing the Ok button). I then saved the resulting filtered data back to raw data, and imported it back into Gimp. I then saved both the original and filtered copies of the image as PNG files from Gimp, and uploaded them to Imgur so I could embed them in this post.

Here's the original image.


For comparison, here's the filtered image.



The simulated AC-Coupling filter obviously messes up the image. Now you know why I will need a DC coupled soundcard before I can start my experiment at building a mechanical scanning infrared imager that interfaces to my computer via the Line Lin port on the soundcard.
« Last Edit: May 22, 2017, 02:34:12 am by Ben321 »
 

Online blueskull

  • Supporter
  • ****
  • Posts: 13531
  • Country: cn
  • Power Electronics Guy
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #6 on: May 22, 2017, 02:40:10 am »
Then you are supposed to modulate the signal in order to remove and reconstruct DC component before putting it through a DC blocking medium and demodulate it.
Sound cards are designed for audio, which doesn't have DC component.
For your application, what you need exists, and it's called a DAQ card.
 

Offline daveatol

  • Regular Contributor
  • *
  • Posts: 134
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #7 on: May 22, 2017, 02:43:44 am »
As stated above, there is next to no need for a DC soundcard. I can't imagine anyone starting up a crowdfunding campaign to make a product just for you to buy a unit.

That said, you can get analog data into your PC in a number of ways; some simple methods include using a microcontroller (e.g. arduino) to sample its ADC and then send the data to the PC through a USB port. You could use a dsPIC for up to 4Msps acq. rate. Or, if you still want to use a sound card, you can modulate the signal to have something that will pass through the DC-blocking caps: AM/FM modulation, frequency shifting, etc. will all allow you to transfer a DC signal to your PC software. There's actually a project that does this on the net somewhere already.
 

Online xrunner

  • Super Contributor
  • ***
  • Posts: 4674
  • Country: us
  • hp>Agilent>Keysight>?
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #8 on: May 22, 2017, 02:47:45 am »
For your application, what you need exists, and it's called a DAQ card.

Yep -

National Instruments
I am a Test Equipment Addict (TEA) - by virtue of this forum signature, I have now faced my addiction
 

Offline Ben321

  • Frequent Contributor
  • **
  • Posts: 506
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #9 on: May 22, 2017, 02:51:24 am »
As stated above, there is next to no need for a DC soundcard. I can't imagine anyone starting up a crowdfunding campaign to make a product just for you to buy a unit.

That said, you can get analog data into your PC in a number of ways; some simple methods include using a microcontroller (e.g. arduino) to sample its ADC and then send the data to the PC through a USB port. You could use a dsPIC for up to 4Msps acq. rate. Or, if you still want to use a sound card, you can modulate the signal to have something that will pass through the DC-blocking caps: AM/FM modulation, frequency shifting, etc. will all allow you to transfer a DC signal to your PC software. There's actually a project that does this on the net somewhere already.

Where can I buy this dsPIC? I see there's a starter kit for it shown on the company's website, but no "Purchase" button. I assume they sell this through distributors? Or maybe you have to call them by phone and ask them for a secret link to get to their online store? The dsPIC just from the fact that it can be used for acquiring analog signals at up to 4Msps makes it sound like a VERY good product.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 8456
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #10 on: May 22, 2017, 03:12:35 am »
IIRC EPE magazine published a USB soundcard based DC coupled scope and signal generator project a few years back.  It used OPAMPs to add/remove the DC offset required for the USB audio chip.
 

Offline daveatol

  • Regular Contributor
  • *
  • Posts: 134
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #11 on: May 22, 2017, 03:36:59 am »
Actually, the PIC32 has some ADC up to 12bit 18 Msps. Find the part you want here http://www.microchip.com/maps/main.aspx . You should be able to buy from element14, mouser or digikey
 

Offline mmagin

  • Frequent Contributor
  • **
  • Posts: 613
  • Country: us
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #12 on: May 22, 2017, 03:55:32 am »
There certainly are audio interfaces which are DC coupled.  http://www.expert-sleepers.co.uk/siwacompatibility.html
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 405
  • Country: us
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #13 on: May 22, 2017, 04:07:40 am »
Such an audio interface won't be too hard to build when using a USB codec (eg: PCM2906), these would be a lot simpler to implement than a PCIe card. Could also use a micro with a decent ADC and USB interface, and implement the USB audio endpoints.

If it's just for a bit of hacking, buy a cheap <$10 USB sound card from any of the usual sources, and replace the DC blocking caps with a 0 ohm resistor. Add a suitable front-end with a PGA and probe interface.
 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2043
  • Country: 00
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #14 on: May 22, 2017, 04:12:11 am »
You can buy the "sound card you want"

Have another name and another price but follow your specs

https://www.mccdaq.com/PDFs/specs/DT9837-Datasheet.pdf

 

Offline daveatol

  • Regular Contributor
  • *
  • Posts: 134
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #15 on: May 22, 2017, 04:12:27 am »
Such an audio interface won't be too hard to build when using a USB codec (eg: PCM2906), these would be a lot simpler to implement than a PCIe card. Could also use a micro with a decent ADC and USB interface, and implement the USB audio endpoints.

If it's just for a bit of hacking, buy a cheap <$10 USB sound card from any of the usual sources, and replace the DC blocking caps with a 0 ohm resistor. Add a suitable front-end with a PGA and probe interface.
It looks like the PCM2906 (and other codecs) has digital DC-blocking filters built in, which I don't think are bypassable.
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 405
  • Country: us
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #16 on: May 22, 2017, 04:25:13 am »
It looks like the PCM2906 (and other codecs) has digital DC-blocking filters built in, which I don't think are bypassable.
Yep, on a second reading of the datasheet, it does include a digital HPF, with no apparent means of disabling it.

https://www.mccdaq.com/PDFs/specs/DT9837-Datasheet.pdf
One could also buy an Analog Discovery 2 for less money (and precision/channels).
 

Online rs20

  • Super Contributor
  • ***
  • Posts: 2188
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #17 on: May 22, 2017, 04:59:21 am »
https://www.mccdaq.com/PDFs/specs/DT9837-Datasheet.pdf
One could also buy an Analog Discovery 2 for less money (and precision/channels).

Or, for even less money, a Picoscope 2000 or 3000 series.
« Last Edit: May 22, 2017, 05:36:59 am by rs20 »
 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2043
  • Country: 00
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #18 on: May 22, 2017, 05:01:09 am »
You can't stream on Picoscope at 48000 Khz
 

Online rs20

  • Super Contributor
  • ***
  • Posts: 2188
  • Country: au
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #19 on: May 22, 2017, 05:44:18 am »
You can't stream on Picoscope at 48000 Khz

I assume you mean 48 kHz? Because 48 MHz is a ridiculous requirement in a thread where the OP specced a "sound card".

Even so, the Picoscope 3000D can stream at 125MS/s, Picoscope 2000 models going up to 31.25 MS/s. Almost every Picoscope model can stream 48 kS/s; although one would want to check with respect to the exact model one was proposing to buy. I suspect the 1kS/s models are now obsolete, there's a lot of obsolete permutations in the list I've linked.

And with at least 64 MS of memory built in (up to the 512 MS I have in my model), there'd be absolutely no need to stream for the OP's proposed laser imaging system anyway. Could all easily fit in a single capture, given that 64 MS corresponds to over 20 minutes of recording at 48 kS/s.
 

Offline IanMacdonald

  • Frequent Contributor
  • **
  • Posts: 944
  • Country: gb
    • IWR Consultancy
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #20 on: July 07, 2017, 05:28:18 pm »
I'm not convinced that DC coupling offers any advantage to sound quality. Usually the reason DC coupled power amps sound better is that in order to be DC coupled they have to have better bias point stability in order not to put DC through the speakers and wreck the voice coils. Thus it's the better design, not the actual DC coupling, that makes the difference.

A big problem with total DC coupling is that any tiny voltage on the input can damage the speakers, or can offset an ADC so that it has reduced headroom. Remember that a condenser mike can have an output in the sub-millivolt range. Try to DC couple that, and even thermocouple effects on connectors can create a significant offset voltage.
 

Offline NANDBlog

  • Super Contributor
  • ***
  • Posts: 4631
  • Country: nl
  • Current job: ATEX certified product design
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #21 on: July 07, 2017, 09:36:03 pm »
Actually, the PIC32 has some ADC up to 12bit 18 Msps. Find the part you want here http://www.microchip.com/maps/main.aspx . You should be able to buy from element14, mouser or digikey
And probably an Errata stating that the ADC does not work at those speeds and resolution. PIC32 is known for that.
 
The following users thanked this post: Kean

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 2624
  • Country: it
Re: Someone needs to make a DC-coupled soundcard crowd-funded project
« Reply #22 on: July 08, 2017, 06:34:27 am »
yep, TURBO MODE IS NOT FUNCTIONAL
or CH0 GIVES BACK INCORRECT RESULTS (or the same as CHx) IN TURBO MODE :--

anyway, there are already DC coupled soundcards. A number of MOTUs are known for that, also some from RME. They don't cost 50€ but they also have damn good pres and outputs
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf