PDA

View Full Version : WIFI for a Stamp


dpeterson3
12-15-2008, 08:24 AM
Hi all. I wanted to try a project here before the build season starts. The theatre at school is in need of a sound system overhaul, and I heard the FCC is limiting the wireless mic band. I was thinking why not have wireless mics that go over VoIP. A?D converters are cheap, and gig switches and cards are becoming cheap. I wanted to use a Stamp to do the interface between the WIFI and converter (PIC for the finial version, but the code for the stamp and pic is so close, I figure that going from one to the other will not be a problem). I can not find a good way to do the WIFI interface, however. I thought maybe a USB stick, but I have no clue how to program it. I tried google, and had no luck. Maybe a grandmaster googler out there could help me or someone has an idea of where to begin. Thanks

EricVanWyk
12-15-2008, 08:56 AM
WiFi is a bit harder than other wireless interfaces, but doable.

I wouldn't recommend attempting to use a USB wifi stick, writing a USB host driver is well beyond the scope of a fun project.

Lantronix - http://www.lantronix.com/device-networking/wireless.html is the bet I've found for an easy embedded WiFi solution.

I would strongly recommend considering other wireless methods, as they will be significantly cheaper/easier. Also, please consider upgrading the Stamp to a more modern option, it will make things easier in the long run.

Have fun!

Andrew Schreiber
12-15-2008, 09:05 AM
http://www.sparkfun.com/commerce/product_info.php?products_id=8869 is another serial - wireless device. I agree with Eric, using a different method might be better. Why not just use FM transmission? Or Zigbee if you really wanted too.

Wifi is going to be expensive.

Russ Beavis
12-15-2008, 09:30 AM
http://www.nordicsemi.com/index.cfm?obj=menu&act=displayMenu&men=73

Another option to consider.

Have fun!

Russ

dpeterson3
12-15-2008, 10:29 AM
The reason I didn't want to use FM is because everything can not be brodcast on the same channel. One of my teachers mentioned that the FCC is reducing the wireless bandwidth for wireless mics. That is where the WIFI idea came from. Everything would use the same channel. Each mic could have its own IP address, so everything can talk on the same network. A web GUI could be loaded on each mic so it can be tuned to the individual using it. A computer in the back running some custom software could handle the mixing. I didn't want to use serial because I wasn't sure I could have a web GUI to tune the mics. Also, serial is shorter range. I need at least 100 feet. Maybe this is just wishful thinking for the time being.

Is there anyway to build those serial converters. I was looking at the sticks because I can get a chip that converts TTL to USB signals. I thought that would do the same thing as the converters you mentioned.
thanks

Daniel

btw, that headset is pretty much what I want to do at a reduced cost. If I could do the interface with the stick, then the total cost for a mic would be around $70, which is still cheaper than most production wireless mics.

EricVanWyk
12-15-2008, 10:44 AM
Is there anyway to build those serial converters. I was looking at the sticks because I can get a chip that converts TTL to USB signals. I thought that would do the same thing as the converters you mentioned.
thanks

What chip are you referring to? Most USB/TTL chips act as a USB device. You would need a USB host. FTDI makes a few nice USB bridges, but they are all devices. USB was designed to make devices simple, but unfortunately this makes hosts harder.


btw, that headset is pretty much what I want to do at a reduced cost. If I could do the interface with the stick, then the total cost for a mic would be around $70, which is still cheaper than most production wireless mics.

Dan, I wish you well. When you get it working, please share with us. I'd be really interested for my own projects.


Yet another option would be the Cypress CyFi. http://www.cypress.com/CyFi

Cypress has the most impressive analog processors I've encountered, I can't say enough good things about the PSoC. I haven't tried their wireless yet, but it may be worth a shot. A rep claimed that they spec 10m, work at 30m, and can hit 1km with a power amp. Reps do lie for a living, but I tend to believe this one.

RyanCahoon
12-15-2008, 02:26 PM
One of my teachers mentioned that the FCC is reducing the wireless bandwidth for wireless mics.
Yes, this is correct, I would just recommend buying future-compatible mics.

Each mic could have its own IP address, so everything can talk on the same network.
The problem with using IP based networks is they weren't designed for real-time applications. You could probably get sound through them but it is probable that it would end up sounding like a poorly dubbed movie. Also, something to keep in mind is if you've ever used VoIP, you know the frequency range is not that great. You might be able to tune it like you mention so that it sounds ok for plays, but musicals would almost certainly noticeably bad. The art of sending quality musical audio over a wireless communication link is the subject of much research and in any case ends up needing a lot of bandwidth: as stated here (http://books.google.com/books?id=e-IeVousmzoC&pg=PA116&lpg=PA116&dq=minimum+carrier+wave+frequency&source=web&ots=XCcsVFT5eZ&sig=M9fOvfdv7ns1X6_t-y7Qhbs13Og#PPA116,M1) you realistically need a sampling/transmission frequency (the sampling frequency on A/D follows approximately the same rules as the carrier wave of a modulated radio system) of 10x that of the highest frequency you want to accurately reproduce. It's hard to get that much data running over an IP network in real time without getting packet congestion.

A web GUI could be loaded on each mic so it can be tuned to the individual using it. [...] I didn't want to use serial because I wasn't sure I could have a web GUI to tune the mics.
I assume by tuning you mean some sort of frequency equalization (EQ). In order to have this on-chip, you're really going to have to look at a microcontroller that has DSP functionality like Analog Devices Blackfin (TI also makes a good line of DSPs, I'm just not familiar enough with them). Also, a web interface means you're goingg to have to run a webserver on the device, which could probably be done on a STAMP or PIC, but in order to have that running at the same as you're piping through the audio data, you're going to have to have some type of real time scheduling. I'm not sure if FreeRTOS has a PIC or STAMP port, but that's where I'd start looking. If there isn't one, FreeRTOS is supposedly fairly easy to port.

A computer in the back running some custom software could handle the mixing.
Supposing you figure out the real-time communication, try looking at a realtime audio processing OS. If you're familiar with OS X programming, that could be a way to go, or take a look at Ubuntu Studio. I'm not sure if either of these have a real-time IP stack though, which would almost probably be necessary.

Is there anyway to build those serial converters. I was looking at the sticks because I can get a chip that converts TTL to USB signals. I thought that would do the same thing as the converters you mentioned.
The chips you're looking at probably present a USB device interface proxy for TTL. In order to interface with a USB stick, you need a USB host interface. (EDIT: oops sorry ErikVanWyk looks like you beat me to it).

btw, that headset is pretty much what I want to do at a reduced cost. If I could do the interface with the stick, then the total cost for a mic would be around $70, which is still cheaper than most production wireless mics.
See above comments and I think you'll be hard pressed to get it under $70. You might still be able to get it under the price of a commercial wireless mic, though, but then you have to factor in Ben Franklin's wise maxim that Time is Money.


One final note is to look at the new AVR32 chips from Atmel. The UC3 family has some DSP capability and the AVR32s have an onboard Ethernet interface that you could maybe couple with a Ethernet-Wifi bridge. There's also a FreeRTOS port for the UC3 series (http://www.freertos.org/portAVR32.html) already released, and it appears to include webserver code already written.

This post is not necessarily meant to discourage; I think it's a fascinating idea, you just have some development challenges to work through.

--Ryan

Al Skierkiewicz
12-16-2008, 07:27 AM
OK, here is a subject that has a lot of confusion.
The FCC is not limiting wireless mic bandwidth. There is a discussion taking place that affects wireless mic frquencies in light of the new digital changeover and the desire by the FCC to auction off some of the previously used frequencies being abandoned by the analog transmitters. The FCC has made a statement that wireless mics come under the "unlicensed transmitter" jurisdiction and some of these mic may exceed the specifications for that rule. If that is the case then there a lot of theaters and churches that may be in violation. Wireless communications, i.e. mics, intercom, and data have all used frequencies that are regulated by the FCC in this country and their counterparts in other countries. No one is sure where all this discussion will lead. Currently new wireless mics are using UHF frequencies or the 900MHz band. Mic designs use both digital and FM modes. There are some popular mics that are available now with USB outputs for use with laptop applications.
When digitizing audio, it is common to use a sampling frequency of at least twice the highest frequency of interest. CDs for instance are 44.1kHz. This pushes the envelope of the 20Hz-22kHz accepted hearing range including the needed harmonic energy for music reproduction. If you think about it, a 22kHz sine wave would have just over two samples per cycle at 44.1kHz. Without some audio black magic, the distortion on this signal when decoded back to analog would be unlistenable. Common digital recording rates are 48KHz or multiples of this frequency (96kHz and 192Khz) to achieve the most accurate reproduction. For a comparison, TV stereo audio is bandwidth limited to 15kHz so as to not interfere with the 15734Hz pilot. Various manufacturers even limit the receivers to only 12kHz to make the decoders less complex.