Soundcard from Scratch

[Xer0]

Hi Guys

im currently reading the CMedia Datasheets and i am curiious about the Idea, to build yourself a Card. with only the Features "WE" need; 6-8ch analog output, and maybe line/mic in for some analytical use, fe analog'ed feedback from the galvo loop signals to build a scansafety in software on the host pc; or other status information of the scanner...
maybe the chip can be even set to provide higher voltage than usual and do differential output with 0V as zeropoint, so you dont even need correction. or write a custom firmware that installs it as standalone device, so it wont mess up on other computers or when you use software DSP's.

The Question is, would that be possible? CM106 seems fairly easy to use for me, all DACs in it. While CM6220 seems to need external ones to output analog.

For what? Size! Or havent anyone here dreamed of a tiny DAC build completely into a DB25 before...

[ElektroFreak]

You'd still need amplification. There's no way a CM chip is going to drive galvos directly, unless scan size is irrelevant to you.. For most of us, the wider the better. An image the size of a sheet of paper at 2m is pointless. If you design the system with the ability to provide a large scan angle, you can always turn down the volume on the sound card if you need it to be smaller. Also, as weartronics showed you, the sound chip doesn't need to provide differential output. It can be added in the amplification stage.

I do like your ideas about scansafety, but I'm not sure it would be possible. I've never read the datasheets for the c-media chipsets, but there may be a way to provide that functionality.

The idea of a tiny DAC is a good one, and you may be on to something with this idea. Miniaturizing the amplifier shouldn't be too hard with SMD components.

[mixedgas]

hate to burst your bubble, but....

The problem is the driver, both DrLAva and I have tried to talk to Cmedia about driver code or documentation access. Answer, from C-Media, NO.. It pretty much does exactly what is in the existing 25$ sound card and that is it.... We'd love to get the aux TTl working, or bypass the surround sound processing stuff, but nope...

If you'd like something open source and hackable, there is

http://elm-chan.org/works/vlp/report_e.html

http://www.linux-laser.org/bilda/bilda-hw.php

Norm's should fit in here someplace..

And a few projects on laserfreak.de



Steve

[Xer0]

Oh no. my bubble!

K we cant get the SDK from CMedia. so it must stay a normal Soundcard. and you have to amplify, thats ok, 2x quad opa for single ended (or 4x for differential) are tiny in SMD. But we rip off SPDIF and so on and it will be a 1-3 Chip solution + correction stage. And by safety i meant make it software-side and use the Mic or Line input of the soundchip to recieve signals from a circuit in the projector which watches the Galvo's...

I just had this Idea tonight that wont let me sleep:
An DAC build completely into the Case of a DB25 Plug. ony a USB-B/Mini/Micro Plug is on the small side. You can plug it directly in the ILDA port of your Scanner and enjoy. it's single ended to save space and cause its so near to the scanner, no need for symmetrical signals. or if you need long cables, use a small "Extender" Box, which is build in a case between D-SUB m & f port, gets 5V power on unused ILDA pins, and inside it features a SE-to-Differential Conversion so you can run it looong away...

[drlava]

A year or two ago when this was fresh, I thought for a while about all kinds of ways to make the sound card DAC into a product with ttl, dmx, etc but the biggest thing holding it back is the C-media driver. It basically needs to be re-written to create a commercial success. While that's doable, it's probably not worthwhile, you might as well write firmware code for a ATMega DAC and use their prebuilt windows driver library. Then you'd have a real DAC with extensibility for a similar effort.

[mixedgas]

I just had this Idea tonight that wont let me sleep:
An DAC build completely into the Case of a DB25 Plug. ony a USB-B/Mini/Micro Plug is on the small side. You can plug it directly in the ILDA port of your Scanner and enjoy. it's single ended to save space and cause its so near to the scanner, no need for symmetrical signals. or if you need long cables, use a small "Extender" Box, which is build in a case between D-SUB m & f port, gets 5V power on unused ILDA pins, and inside it features a SE-to-Differential Conversion so you can run it looong away...[/QUOTE]
\

Have you seen a RIYA micro? I own one, its only 35% bigger then what you describe. Very useful little widget.... Has its limitations, but serious bang for the amount spent. Get a RIYA micro or larger unit and spend your time writing useful code or learning show programming.

Steve

[sugeek]

I just had this Idea tonight that wont let me sleep:
An DAC build completely into the Case of a DB25 Plug. ony a USB-B/Mini/Micro Plug is on the small side. You can plug it directly in the ILDA port of your Scanner and enjoy. it's single ended to save space and cause its so near to the scanner, no need for symmetrical signals. or if you need long cables, use a small "Extender" Box, which is build in a case between D-SUB m & f port, gets 5V power on unused ILDA pins, and inside it features a SE-to-Differential Conversion so you can run it looong away...

\

Have you seen a RIYA micro? I own one, its only 35% bigger then what you describe. Very useful little widget.... Has its limitations, but serious bang for the amount spent. Get a RIYA micro or larger unit and spend your time writing useful code or learning show programming.

Steve[/QUOTE]

I was about to say that, if you want a RIYA micro, I was about to sell mine...

-Adam

[riyalasers]

...An DAC build completely into the Case of a DB25 Plug. ony a USB-B/Mini/Micro Plug is on the small side. You can plug it directly in the ILDA port of your Scanner and enjoy.
...

My five cents
lasershow devices in most situation are used not in laboratory, but in extreme situations (dance club... disco... dark rooms).
to use USB-B/Mini/Micro in similar devices is bad idea: they have very low reliability... and can be damaged...
For example... where is general problems for photo cameras? IN CONNECTORS!!!!

We use in our Multibus Micro controllers cable, which is soldered to PC board directly:






About size of controller: some days ago I traced new version of controller (Multibus Lite)
It have 12 bit X&Y and 8 bit R, G, B, Int; USB interface.
Their size is 34*37.5 mm
It is world record in controller size I think

[dave]

hate to burst your bubble, but....

The problem is the driver, both DrLAva and I have tried to talk to Cmedia about driver code or documentation access. Answer, from C-Media, NO.. It pretty much does exactly what is in the existing 25$ sound card and that is it.... We'd love to get the aux TTl working, or bypass the surround sound processing stuff, but nope...

If you'd like something open source and hackable, there is

http://elm-chan.org/works/vlp/report_e.html

http://www.linux-laser.org/bilda/bilda-hw.php

Norm's should fit in here someplace..

And a few projects on laserfreak.de



Steve

Combine the BILDA with the new wiznet IMCU7100?? The new wiznet chip looks pretty sweet 8051 core with a hardware IP core and dual port ram to link the 2 together.

Just a thought

[weartronics]

I just had this Idea tonight that wont let me sleep: An DAC build completely into the Case of a DB25 Plug. ony a USB-B/Mini/Micro Plug is on the small side. You can plug it directly in the ILDA port of your Scanner and enjoy.

The Question is, would that be possible?

Yes, it would!



I have investigated at least half a dozen of the popular USB ILDA DACs, and none of them have working Linux drivers; evidently there is no demand for this. Rather than design an entirely new DAC for this purpose, I wrote a Linux driver for the C-Media chips so that a) audio applications don't see them as sound cards and accidentally cause galvo movement or laser emission, and b) a simple laser frame based API can be used, with the PCM generation and buffering abstracted inside the driver.

At the same time, I used the GPIO pins to get the ILDA return input signal (pin 12) and set the ILDA shutter output signal (pin 13), and I used a shift register with the micro-controller interface to provide eight TTL outputs for beam or effect actuators on the user-defined signals (pins 9-11, 21-24). It would be possible to use the same micro-controller interface for more comprehensive DMX or TTL input/output, but eight TTL outputs is all that fits nicely on the ILDA connector.

This approach seems to work well. In fact, because there is no dead-time between frames, you actually get higher average brightness and faster animation rates than any commercially available USB ILDA DACs. At the moment I'm still using the blue box prototype that many of you have seen, but I will be producing a few more for other projectors using the board above, which integrates the C-Media circuitry along with the differential amplifiers inside a DB-25 enclosure.

The problem is the driver, both DrLAva and I have tried to talk to Cmedia about driver code or documentation access. Answer, from C-Media, NO.. It pretty much does exactly what is in the existing 25$ sound card and that is it.... We'd love to get the aux TTl working, or bypass the surround sound processing stuff, but nope...

The C-Media chips implement the USB audio device class interface. That's why they work using generic drivers included with your operating system, rather than having to install any specific drivers from C-Media. The USB audio device class interface specification contains all the necessary information to write your own driver if you so desire, and the C-Media data sheet lists a few HID class requests that are used to control the proprietary features. I'm not aware of any SDK or other documentation available from C-Media.

I think DrLava has already mastered the HID class requests to disable the interpolation on CM106. The same requests are used with different parameters to disable any surround processing and to control the GPIOs. The parameters are all documented in the C-Media data sheet. If you have any questions about this, I can probably help you.

Last year, I sent a copy of my Linux driver to James Lehman in case he wanted to add real-time output to LaserBoy without having to go through the sound subsystem. However, he hasn't pursued it because the vast majority of his development and his user-base are on Windows.

For various reasons that I won't explain here, developing a similar driver for Windows is much more difficult. In fact, for the volumes involved, it's simply not practical to write this driver. If the generic driver included with Windows is not suitable, it would be simpler to forget about the C-Media chips altogether and use a general purpose micro-controller, where a variety of higher-level libraries are available to write the driver.