Uploads from linux-works, tagged arduino

my table setup at 'burning amp' this year (sercona audio remote controlled preamp)

nobody@flickr.com (linux-works) — Mon, 29 Oct 2012 10:49:10 -0700

I ended up bringing more gear than most. took well over an hour to setup. too many cables; why can't we go all wireless? lol

the main component of this setup that I'm demo'ing is the silver 1U rackmount style device in the center, with the red (bigfonts) lcd display. its an 8-channel analog (consumer/pro line level) preamp that is controlled digitally but varies volume purely in the analog domain.

the new feature I just added was a 'link up' between the computer's notion of a volume control (eg, a volume slider on linux ALSA sound system) and my own preamp's volume value. turn the knob on my real-world preamp and the value inside the pc changes. change the pc's volume slider via any of the usual means (eg,android or alsamixer) and my volume mirror that value and your volume will go up or down that amount, to match.

any app that has a volume slider and uses alsa sound engine in linux (almost all do) can make use of my preamp. the connection is even RF wireless between the pc and the preamp (small xbee rf modem sitting on the white wifi router, right behind the silver sercona preamp box). this means your pc music server could be located somewhere else, and the preamp and pc can still 'stay in touch' and keep the mirrored volume slider working. and, of course, you can walk around with your android phone, change music using mpd and now change volume, too, without harming the sound in the process.

and btw, if your source is natively analog (turntables, etc) this volume control avoids any conversion to digital. its purely an analog volume control that happens to be digitally controlled.

it runs the open-source LCDuino command code and a 2nd arduino processor runs embedded on the cirrus 3318 (vol control engine chip, itself) board. the 2 arduinos are linked via serial-to-softserial. the main arduino is 'reachable' via its hardware uart serial port and this is what the webserver or mpd server uses to talk over.

a linux daemon process runs in the background keeping the mixer volume slider on ALSA linked to the serial stream on the sercona preamp. if the sercona changes, alsa updates. if alsa changes, the sercona updates. its just that simple ;)

top view of the preamp, showing the hand-wired proto boards: www.flickr.com/photos/linux-works/8126599842/in/photostream

DIY: sercona preamp getting ready for Burning Amp (this sunday)

nobody@flickr.com (linux-works) — Fri, 26 Oct 2012 19:54:22 -0700

linux-works posted a photo:

fort mason, san francisco is where this guy is headed, this coming sunday. the annual 'burning amp' DIY audio show comes right around the last week of oct.

today, I cleaned up (really, lol) the wiring and reoriented the internal boards so they fit better and are a bit neater.

the input board is new (far left with green caps) but its passive, at this point (no op-amps yet). that will come later and there's board space reserved for it.

the output board (far right) does have sockets installed for output buffers but I am using shorting bypass plugins for the time being.

the center chip is what this is all about and that's the cirrus cs3318 8ch volume control chip.

to the front, center, is my own LCDuino.

in this build, there are 2 arduinos and they are linked over a short serial connection.

all that hand wiring and proto board soldering. so much work! ;)

(sercona = "serial controlled audio")

DIY: start of a new embedded project

nobody@flickr.com (linux-works) — Wed, 03 Oct 2012 10:46:20 -0700

linux-works posted a photo:

digital section is mostly done and on the left side of this board. it has its own ground plane (you can see a line going up/down where I cut the ground plane, creating 2 separate ones).

power can be from the arduino 'download/host' cable or via the onboard 5v regulator (red jumper selects). if reset is needed, use blue jumper; if serial i/o between host then remove blue jumper.

the opto chip, below, has 4 leds and 4 transistors inside. I'll use this for an SPI level converter from 5v down to 3.3v for the 'right hand side of the board'.

rasp_pi_and_cd4050_level_converter_2

nobody@flickr.com (linux-works) — Tue, 11 Sep 2012 17:41:36 -0700

linux-works posted a photo:

rasp_pi_and_cd4050_level_converter_3

nobody@flickr.com (linux-works) — Tue, 11 Sep 2012 17:41:25 -0700

linux-works posted a photo:

rasp_pi_and_cd4050_level_converter_1

nobody@flickr.com (linux-works) — Tue, 11 Sep 2012 17:41:49 -0700

linux-works posted a photo:

using a cmos cd4050 buffer chip to convert 3.3v to 5v.

the raspberry pi has its UART run at 3.3v and the arduino needs 5v. so, this little DIY perfboard 'shield' is needed.

note, how the chip is specifically (yes, very much on purpose) resting on the unused white riser connector. there is no other support (or screw mounts!) on this version-1 raspberry pi board.

schematic link1: picasaweb.google.com/lh/photo/j-l64p2I_4QVANbyiK8JFdcYewr...

schematic link2: www.andremiller.net/content/raspberry-pi-and-arduino-via-...

discussion link (AMB diy forum thread): www.amb.org/forum/raspberry-pi-lcduino-system-t1970.html

this board was so simple, I actually didn't run out of wire-wrap wire colors ;)

DIY: wm8805 spdif-rx and LCDuino controller

nobody@flickr.com (linux-works) — Sun, 19 Aug 2012 10:44:29 -0700

linux-works posted a photo:

an ebay wm8805 board (blue, rear) and a custom perf board (center), plus the LCDuino-1 (front) and finally the sigma25 (left) powering it all.

this is an input selector for digital audio, and its goal is also to clean up the timing (jitter) if its too far out of whack.

there are 3 inputs on the rear (usb, opto/toslink and coax/white). there are 2 outputs (coax/red and balanced TRS).

the front has an LCD display (2x16 char), a learn-mode IR receiver and 3 direct access push buttons.

hardware-based (arduino) SPI is now working. (mcp4822)

nobody@flickr.com (linux-works) — Mon, 28 May 2012 12:28:14 -0700

linux-works posted a photo:

MUCH better, now! 5.8microseconds to set a 12 bit value to the dac over hardware-spi. (arduino 0022 has a bug in its SPI.cpp lib, sigh).

for demo, I set the step-size to 1024 so that you only get 3 levels on the wave. the hardware frequency counter on this scope reads 35khz. that means I was able to synthesize a crude 35khz wave (could be a nice sine after shaping) using a serial-based dac chip. or, that I can get data samples out 172,000 times per second! that was the real improvement when I enabled hardware SPI.

see:

www.flickr.com/photos/linux-works/7275918648/

for the software-only timing (264us vs 5.8us !)

DIY: spi-based 12-bit DAC chip (mcp4822)

nobody@flickr.com (linux-works) — Sat, 26 May 2012 14:09:12 -0700

linux-works posted a photo:

arduino driving an spi-based 12 bit DAC chip.

(I know.. all white wires on a white background make this really easy to follow, lol)

DIY: voltage programmable 'active load'

nobody@flickr.com (linux-works) — Mon, 23 Apr 2012 23:04:23 -0700

linux-works posted a photo:

inspired by dave jones (eevblog). here's an updated schematic that helped me get mine working: www.sleepyrobot.com/?p=95

lm358 dual single-supply op-amp controlling a 'logic level' mtp-3055 mosfet. mosfet is in negative feedback loop and the drain-src flow is where you put your power supply (or amp) under test. the mofset will present a load based on how much voltage you supply to the op-amp's non-inverting input. essentially, that's it.

I used whatever I had on hand for a power resistor (src on the mosfet, to gnd). the DJ circuit and many others use a power of 10 value of R (1ohm or .1 ohm) so that you can put a voltmeter across it and read, directly, the current being consumed.

in my version, I plan to have an arduino drive this (via a digital pot or even DAC chip) and I can calibrate it via a lookup table.

you do need a stable voltage source and your accuracy will depend on that, to a large degree.

on my build, I included a stud mounted thermistor (2 small orange wires) and it was a neat little device that comes pre-mounted in a round hole mount metal washer. the resistance varies from about 10k at room temp to less than 1k when I touch my soldering iron to it (lol). when I push 1amp thru, its 2400 ohms. I plan to pwm control the fan (this is an old CPU heatsink fan!) via the arduino if the temperature gets too hot. I'm hoping I can get a continuous 3amps from this, maybe even more.

link to dave jones blog:

www.eevblog.com/2010/08/01/eevblog-102-diy-constant-curre...

link to dave's video: www.eevblog.org/video/EEVblog102-480x272.m4v

DIY: automated linux-assisted hardware testing

nobody@flickr.com (linux-works) — Mon, 02 Apr 2012 21:48:23 -0700

linux-works posted a photo:

a new milestone for me: I threw together some C code, shell scripts, cables and test gear in order to see how linear my relay attenuator (for audio) is.

link to my write-up on how I did all this:

www.amb.org/forum/automated-testing-of-delta1-steps-t1752...

the fluke shows 17.44dB measured and my LCDuino says it should be 17.5dB. that's .06dB error on this step. not bad, huh?

with 50v put into the attenuator, the fluke reads 6.728v out. I have the fluke setup to read both relative dB and absolute volts in its two displays (and also on its serial port).

a typical fluke line would look like:

-17.44 +6.7275

or

-125.42 +0.027E-3
-126.02 +0.025E-3
-126.33 +0.024E-3
-127.00 +0.022E-3
-127.18 +0.022E-3

as we get closer and closer to 127.5dB attenuation (the most this relay board can handle).

(not all the equip shown was used; but if it had its display on, it was part of this test) ;)

DIY: adding s/pdif LED display to an existing DAC

nobody@flickr.com (linux-works) — Mon, 27 Feb 2012 13:10:29 -0800

linux-works posted a photo:

for more info on this project, see: www.amb.org/forum/lcduino-code-that-displays-spdif-sample...

user interface detail: the left group of leds show the digital audio frequency (sample rate). only 1 will be on at a time (or none) and I only had 3 colors to work with; but from a distance you can tell which of the leds is on by the color and whether its toward the left side or the right side. no 'digits' display is really needed, and in fact, would be less readable at a distance than this scheme.

the right 3 leds will show which one of the rear inputs is selected; only 1 of the leds will be on just as only 1 of the inputs could be selected at a time.

very minimal drilling (or drilling skills, lol) needed for this user-interface style.

I plan to label the leds, but after a while, you get to know what they mean and labels only clutter up the front.

DIY: arduino s/pdif samplerate LED display

nobody@flickr.com (linux-works) — Sun, 26 Feb 2012 08:30:58 -0800

linux-works posted a photo:

a fully home-made arduino. an ancient 168 version part, at that.

it runs a software frequency counter program and displays the nearest standard spdif 'data rate' on a colored led. for this build, I chose 6 led's in 3 colors. from a distance, you can easily see which of the 3 colors it is, and you can also see which side of the display it is. that's enough to know which item is displayed, even with lower eye vision.

some standard rates that you could assign to leds: 32k, 44.1k, 48k, 64k, 88.2k, 96k, 128k, 176.4k, 192k. there are 6 leds on this board but there are more arduino digital pins still available. I chose 6 for visual simplicity and I can pick the 6 most common spdif sample rates and have those be the ones that the leds will show.

this board is meant to be mounted in an existing stereo system. only small round holes (your choice how many you want to use) are needed. you can then add sample-rate display to your DAC, switch, cd/dvd/bd player, satellite receiver or anything that uses spdif digital audio.

for solder-side, see:
www.flickr.com/photos/linux-works/6931855049/

for the frequency counter arduino code, see:
interface.khm.de/index.php/lab/experiments/arduino-freque...

I will upload my own code for this complete solution, shortly.

(crude video to show the leds coming on during valid spdif frequencies: www.netstuff.org/spdif/spdif_led_meter_1.mp4 )

with complete wiring to a usb 'debug' port as well as the 2 wires needed for frequency sense that are schmitt-triggered by a 74hc14 chip (air-wired on the back of the board):
www.flickr.com/photos/linux-works/6788052496/

DIY: calibrating an arduino frequency counter

nobody@flickr.com (linux-works) — Thu, 23 Feb 2012 15:13:10 -0800

linux-works posted a photo:

after adding a software calibration constant, the readings are much closer and within acceptable tolerance, now.

DIY: calibrating an arduino frequency counter

nobody@flickr.com (linux-works) — Thu, 23 Feb 2012 15:13:10 -0800

linux-works posted a photo:

after adding a software calibration constant, the readings are much closer and within acceptable tolerance, now.

mcp4551 digital pot on MSOP carrier

nobody@flickr.com (linux-works) — Thu, 26 Apr 2012 16:59:11 -0700

linux-works posted a photo:

I think I positioned this 180degrees the wrong way.

when you get down this small the index hole is not easy to see. at least with this vendor, they didn't do me any favors with a more obvious pin-1 marking.

sigh. I'll have to label this module with some sticky tape or something so that I don't plug it in the wrong way. I'm not about to remove and redo it, either!