Saturday, February 17, 2007

MC-7 keyboard

I found a nice keyboard from flea market yesterday. It was not in very good condition-> the drum section was completely silent. I opened it and found that some components were loose because of bad soldering work. After soldering these properly the drums started working.
The sounds of this keyboard are simple but very nice. All keyboard instruments are quite raw sounding squarewave. Drum sounds are short noise (hi-hat), smooth click (bass drum), and sharper click (snare).

Original features:
2 notes polyphony (monophonic if accompaniments are on)

Cotrol panel from left to right:

-power on/off switch
-accomp volume 4 way switch – this controls only volume of the drums not the accompaniment bass line
-master volume 4 way switch

Rhythm section:
-8 preset rhythms: disco, march, rock, waltz, pops, 16 beat, rhumba, bossanova
-accomp on/off button: the accompaniments are very simple and funny
-tempo buttons: slow and fast
-synchro button
-start button
-stop button

Orchestra section:
-8 preset instruments: clarinet, violin, oboe, piano, elec guitar, xylophone, harpsichord, mandolin
-sustain button
-vibrato button
-demo button

record section:
-record button: 28 notes
-play/stop button

Custom drummer section:
-3 drum pads: bass drum, snare, hi-hat
-program button: 16 steps
-play/space button

After some testing I found out that each of the three drum sounds are triggered from individual pins from the ic. There is also a fourth trigger pulse witch was unused in this model. I soldered wires from all 4 pins. Now I have 4 trigger output lines that follow the preset rhythms and three of them (bass drum, hi-hat and snare trigger signals) are programmable for 16 step sequences. These wires can be connected to various points on the pcb, when 1 or more of them are routed to the pins that read the buttons/keyboard you get many strange rhythmical bleeps and other crazy features like “bouncing” tempo and “random” notes and chords played.
I also found the output pins for keyboard sounds and the accompaniment sounds, these can be connected to various points on the pcb to change and modulate the sounds in many ways.Because the many possibilities of expanding the features of this keyboard I decided to build an extension case to fit the new controls, outputs, patch bay etc. So many ideas: adding a low pass filter circuit, new drum sound circuits to use with the trigger signals, joystick controller for some features, new buttons, switches and a patch bay…. a lot of decisions to make.

Well now I addad few wires (30 and counting) to different points on the pcb. All of these will be going to a patch bay so there is quite many possibilities to change the sounds and functions of the keyboard. Here is a picture.

The new wires from the pcb are soldered to a piece of stripboard so it will be easier to direct them to different controls and patch bay.

After some thinking and going through materials at our studio. I found the perfect extension case. It seems that I´m going for a modular design with this one. :)

Here is mc-7 posing with the extension case. I cut some new panels from red plastic. The alumnium panels were on the case.

Wednesday, February 14, 2007

Bending Yamaha PSR-6

I got my hand on Yamaha PSR-6 and of course I did bend it :)
here is the story, hope you can learn something.

after opening look for the big chips...

here are the two main chips the XE323B0 is CPU and YM2413 is the FM sound chip.
and the PCB side looks like this(note:image flipped around for better viewing):

there is nice information of YM2413
YM2413 FM Operator Type-LL (OPLL) Application Manual
and as far as i understand the CPU sends bits on eight data lines to FM chip and the sound's are made inside the FM chip...

after reading nice instructions on on bending PSS-270
I wanted to know more about the data flow between the CPU and FM chips. I installed LEDs that indicate if the pins are high of low, a kind of dataflow monitor.

some details:

wires attached to YM2413

wires coming out from little screw hole

transistor station

the leds and transistor station

And finally the schematic:

duplicate this 8 times.

here is a video of this thing running

next thing to do is the switches and the banana board.

here is my messy soldering but - everything is fine i checked whit multimeter - no short circuit's :)

and here is the 2mm mini-banana patch points (yellow -from CPU. green- to FM)

and here are switches for getting back to normal and for setting crazy settings

"Each switch now controls the flow of data from the FM chip to the main chip. Turn the yamaha on and select a patch (let's pull up #89 "ghost"), I usually play some keys(don't know if you have to do this or not), then turn some switches off , now select a different patch, and turn the switches back on. Now try playing the new patch. If it went well some of the data from the all patch got held up, and then inserted into the new patch making a totally different sound. Sometimes leaving the connections off will change the sounds also. I've noticed them some of the bent up patches will only work when multiple keys are played. I've heard that this will effect the drums as well, but I haven't gotten it to do much with them as of yet"

after setting the bended tone - i messed around whit the banana patch -
it's sooo nice -
listen .mp3 sound samples

After making this I received some questions. Here are the questions and my answers:

Quote, mister.
"I have recently begun bending a Yamaha PSR-6. The only info I have found on bending for this model of keyboard was your's. I had a few questions if you don not mind. First I wanted to know how your particular approach came out (hoe pleased are you with the results). Second, I wanted some more info on your LED wiring. I have just begun learning more details on electronics and from what I saw on the forum I'm farely certain I know what you did, but more details would be nice. For example, what type of NPN did you use exactly? Also where the hell did you get those pin jacks? I have found some probes that look similar, but I don't know if there are any sockets for them. I have oredered RCA and will use those, but I like the pin jacks a lot and even if I don't use them on this project it would be cool for future work. Thank you for your time and much repect for your work from me. I hope you can afford the time to help me out. Any details you can give would kick ass."

Quote, sponge.
"hi there, i saw that you have successfully bend the yamaha psr-6. i picked one up the other day from a charity shop. im a little confused about what you were trying to do with it... sounded very cool though! would you be willing to tell me where the bends you found were, and what the LED / transistor circuits were for? i've found a couple of bends, but i've only bent one thing before, and that was a cheap effects pedal."

My answer:

Lets start whit the schematics and pics:

here are the main chips from the "solder side".
red dots are points D0 - D7 of the FM chip YM2413
and the blue dots are points D0 - D7 of the CPU chip XE323B0
yellow lines indicate where you need to cut the traces.

here is the YM2413 chip Pin Assignment:

here is the schematic:

you need to ad this circuit to all eight points (D0 to D0, D1 to D1 ...)

the transistors i used are C2389
- but you can use any general type NPN transistor. experiment whit the ones you have.

the bends i made are
1. 8 switches to control data flow
2. 8LEDs to monitor the data flow
3. data-patch bay to patch data

Haven't had much time to play whit my Yamaha but i must say that I'm pleased whit the results.

The method described in Kevin Rees site is good start.
here are some of my own that you can hear in the sample:
start whit sound 40, turn D0 off, change to sound 50 and turn the D0 on - start playing.
start whit sound 40, turn D7 off, change to sound 50 and turn the D7 on - start playing.
start whit sound 40, turn D0 off, change to sound 89 and turn the D0 on - start playing.

As you can now imagine there are MANY combinations .
after i have set a sound. i start patching - but that's not that easy - the monitor helps in this - but it crashes a lot - by trial and error you will learn :)

The D4 is used to trigger the sounds so if you have that turned off, nothing happens

The pin jacks i got from YEinternational but they only serve you in Finland, Russia and the Baltics. The pin jacks are made by Hirschmann. Maybe if you ask your local supplier if they have those 2mm pin jacks? they can arrange something.

About the LED data monitor.
Every time one of the points D0 -D7 goes high, LED D0-D7 goes on.
By looking the LED' s blink you can see the data flow and its easier to learn patching when you actually see whats going on in the inputs of the chip.

hope this helps

for more questions don't hesitate to post it here.


Tuesday, February 13, 2007

Using TV to display audio signals.

you may have heard about: Wave Vessel, Wobblevision, Analog Visualization Unit, AVU, Psychoscope, o-scope, b&w tv into a large screen oscilloscope kits...etc...

This is it all in one.

Before we begin, I have to give everyone a warning.

Before you begin read these safety instructions:
Safe Discharging of Capacitors in TVs and Video Monitors
Cathode-ray tube Safety Issues
Deflection System Safety

There are two areas which have particularly nasty electrical dangers:
the non-isolated line power supply and the CRT high voltage.
Stay away from the red cable with the suction cup end.
After you have opened the TV unit wear rubber gloves!
(longer sleeves better)

The principal of this audio signal displaying is that when TV is normally operated the Deflection Yoke
generates a magnetic field and use it to direct the electron beam in the cathode-ray tube.
Deflection Yoke gets signals from deflection circuit.
Two of the wires control the horizontal motion, and the other two
control the vertical motion. You just replace the Deflection Circuit
signals whit amplified audio signals and then controll the beam by the audio signal.

I have made some schematics to clear how to route audio signals to
Deflection Yoke- these
schematics are based on information from Wave Vessel page and my own experiments.

first you need to find out the two horizontal and two vertical wires.
these instructions on Censtron page will help you
"The diagram above shows all six possible wire combinations. The diagram is drawn from the angle of looking directly at the back of the tube. The colors I used are just for reference,and may be different depending on the model of the television."

"First clip and strip all four of the wires going to the tube, then connect them all back up using alligator clips.
Then, remove one of the clips, and turn the television on. If you get a vertical line, then the disconnected wire is one of the horizontal drive wires, if you get a horizontal line, then you disconnected a vertical wire. Turn the television back off, then follow the same process for the other
three wires to determine which direction each controls."

Here you see the wires going to yoke ->

Hg =horizontal ground
Vg=vertical ground

Here is the unmoded version:
Whit this you can see normal TV programs.

Here is the Censtron version (a
Whit this setup you get a bar that differens in length
a bit like panel of KITT [the car from Knight Rider]

Here is the Censtron version (b
Whit this setup you get vertical line that vibrates.

Here is the stereo version.
Whit this setup you get pulsing circles

After reading this post on
i decided to ad second yoke...

quote: DBerning on
"Magnetic deflection TVs of this era needed the deflection yoke for energy storage for the flyback pulse
that was used to provide high voltage for the CRT. Because I was using the yoke for audio
deflection, I mounted a second deflection yoke somewhere in the TV cabinet so that the inductance
from the winding could be used in the flyback-based high-voltage circuit."

Here is my Philips TX:

One month later:

Okay i did find the extra yoke but, it was not the right type.
This is what happened:
The picture was extremely bright! i mean really really bright.
I was very happy for a while, it looked so cool. After some 15 -30mins both of the TV sets
fuses were burn and the unit was extremely warm. Then i noticed there was a X-ray warning on the tube :" do not drive the CRT whit too much current/voltage or it will start produce X-rays.
So what i assume happened is that i drived the CRT whit too much voltage/current and it started to produce X-rays!

What did we learn?
If you want to ad the extra yoke to extra brightness you need to
measure the yokes inductance first and then find another yoke that has
the same properties.
(or if you want a cheapo X-ray machine just replace the fuses whit thick wire);)

Day later:

I'm feeling lot better now

i called to The Radiation and Nuclear Safety Authority of Finland and asked
if i have been exposed to any radiation danger by moding my TV. They said that there is very
small possibility to get enough radiation from CRTs - in their test
the maximum X-ray radiation from television has been 100mGy/h (hundred milliGrays per hour)
and after that the tube has collapsed.

To get symptoms from X-rays you need to get at least 100Gy.
The symptoms are then the same type of skin burns you get from sun.

The X-ray radiation on tv-set is stronger in the back of the tube than
in the front. The X-ray born's when electrode hits the phosphor wall.
X-ray bounces back to the inner walls of tube. it keeps bouncing and gets weaker and dies.
The glass is thicker in the back so some x-rays get through but we are talking of
really small amounts of radiation.

After replacing the fuses i tryed it again, so i connected the coil
to the points of horizontal and horizontal ground of the Deflection
Circuit(Board) - and wow it was cosmic again. - the beam was so bright - it was just like looking to sun.
well i realized that the unit heated a lot - so i attached two ventilators . . . and i took it to "koneisto" (
The Festival For Electronic Music in Helsinki) it was on stage whit Aavikko. It lasted one hour,
almost the whole gig and died of over heating. I have to say that no TV has ever be so bright!
all the time it looked that it will explode any minute :)

this is the setup what i did

so i have been looking for the next TV (victim) to my experiments. I got now a correct size coil from the broken Philips TX.

Meanwhile Censtron has made their first color vessel:

"We just finished making the first Color Wave Vessel, with it you can now make the visualizations any color from invisible to white."

Two months later:

My Friend gave me a old soviet travel-TV that was not working but, it could display the usual noise.
I did the things described in this article (whitout the extra coil)- whit a 2mm pinjack -patch panel and speakon connector - so i can connect straight to my amp.
here are some pictures:

not much space inside so i removed the speaker

The patch panel made of perspex:

the whole unit:

This video shows the beauty of electron beam bouncing whit music

I tryed whit a color unit and gues what, it's really easy, the scope mod its the same as whit B/W tv's and the color adjuts mod is done by soldering of three potentiometers from the back of the PCB that's in the back of the CRT. and mounting new potentiometers to the case. here is a pic of the pots that needs to be taken of and replaced by a same value.

here is the color unit in action, Photo by Tomi F

Some links to DIY Oscilloscope's

instructions to make GBDSO (GameBoy Digital Sampling Oscilloscope)

LED-matrix oscilloscope

there is a neat how to on geek technique. org
how to make Mac se/30 audio visualizers

Take care

Monday, February 12, 2007

From Joystick to noisestick

Noistick- create a noise making device from a joystick with autofire


This is experimental electronics project to turn your old joysticks into nice sound devices with no external sound source needed.

What you need to build this project:


1.Potentiometer (commonly pc or mac) joystick with autofire circuit. (C64, Atari and some other old joystics are different and there are no potentiometers inside them.)

Here in Finland these you can find these very cheap from fleamarkets. You recognice them from smooth and quite long movements of the stick and of course a switch that says autofire or turbofire.

The autofire circuit will act as oscillator and joysticks horizontal and vertical movement will be wired to control it.

2. soldering iron and a screw driver

3. some wire (or use the wire from joystick)

4. 9v battery and battery clip or a battery holder for 2-4 AA batteries

You might have problems fitting these inside the joystick. Make an external battery compartment.

5. some capacitors, resistors and diodes (recycled components from old broken electronic devices will do fine)

6. amplifier.

Small active computer speakers are probably the cheapest test amps you can find.

7. wires with alligatorclips (very useful when doing experiments)

8. output jack.

Open your joystick to see what is inside.

1. transistor based (No example here, but read through this article and start experimenting with these too. )

2. 555 based

3. cmos based

In this project we use 555 or cmos based circuit because they use only few components (In most cases 1 ic, 1 capacitor an 1 resistor) and are easy to do. When adding new components to circuit try different values with alligatorclips.

555 timer

step1. power it up

To power this circuit battery + needs to be connected to pin 8 and ground (-) to pin 1 . In some cases the connections for + and ground are marked on the pcb. If not you can easily follow the copper traces under the board. At this point use alligatorclips to power your circuit.

step2. how to hear if it works

When the circuit is powered correctly you can probe the circuit with alligator clips connected to input of your test amplifier. Pin 3 is output in the 555 circuit, just follow the copper traces on the pcb. You should be able to find ticking or motor running type sound. The oscillator is running quite slow because of the original timing network (resistors and capacitor).

diagram1. 555 oscillator inside the joystick

step3. make it adjustable

To make the sound adjustable remove resistor R1 and replace it with either x or y potentiometer inside the joystick (or both in series to get wider range). You can also change the capacitor to change the scale of the oscillator (larger values to get lower and smaller values to higher frequencies. Experiment different values)

After the resistor is removed solder two wires to the points where resistor was connected. Then solder other ends to one of the two potentiometers. One wire to middle pin and one to side pin. You might want to solder a resistor in series with the potentiometer so that the oscillators scale stays in audible frequency (try different values to find which value you need). Now moving joystick will change the frequency of the sound.

diagram 2. x potentiometer wired to make the oscillators frequency adjustable

step4. output

Solder a capacitor between the output wire from circuit and output jack. The signal is quite strong so you might want to add volume control -> a resistor, trimmer or potentiometer to output line.

WARNING.I do not recommend that you to plug this or any other experimental sound devices to your most expensive hi-fi audio equipment. You might end up having broken expensive hi-fi audio equipment.

step5. modulation input

You can feed external modulation signal to the 555 ic. Simply solder a wire between a input jack and pin 5 on the ic. Modulation signal needs to be quite strong. If you build two noisticks you can use one to modulate the other. As always-> experiment to get the best results.

step6. buttons

There are many possibilities to use the buttons. You should think what functions you need. Here is an example: Run the + wire through a button to the pcb. The circuit will be powered only when the button is pushed. Then add a electrolytic capacitor between the button connections (caps + connection to battery side). Now when the button released sound keeps on going for awhile (big capacitor values produce longer sound )

step7. Make some noise

After all wires, components etc. are soldered properly and everything seems to be working. Close the screws and your noistick is ready to annoy people around you with its horrible sounds.

cmos hex inverter

This type is the most fun because you can easily build more than one oscillator around the chip. So you can have an individual oscillator for x and y movements of the joystick with adding just few components. Only two different circuitry is presented here. You may (and probably will) find various others. Same principles can be used to many other joysticks with hex inverter based circuitry.

Both two joysticks use cmos hex inverter. First circuit uses 4069 second uses 74hc04 and the following instructions apply to these particular joysticks (the ic:s have same pinouts). If your joystick uses different ic than 4069 or 74hc04 you should check the pin connections and other information from datasheets (you can find them on the web)

step1. power it up

To power this circuit battery + needs to be connected to pin 14 and ground (-) to pin 7. Sometimes the connections for + and ground are marked on the pcb. ATTENTION! Some joysticks use 74hc…

marked chip (like the second example: 74hc04 hex inverter) It will fry if powered with 9v! With these use two to four AA batteries instead. Follow the copper traces on pcb to see witch wires go to these pins or solder new wires for battery. Make sure you connect the power correctly because wrog polarity can kill the cmos ic.

step2. how to hear if it works

When the power is connected correctly you can probe the circuit with alligator clips connected to input of your test amplifier. You should be able to find ticking type slow squarewave sound.

diagram 1. 74hc04 oscillator inside joystick 2. c1 and r1 create the timing network. pin 7 is connected to ground. pin 14 to battery +

diagram 2. 4069 oscillator inside joystick 1. c1 and r1 create the timing network. pins 1, 3, and 7 are connected to ground. pin 14 to battery +

step3. make it adjustable

To make this sound adjustable just remove the timing resistor (R1) and replace it with either x or y potentiometer inside the joystick. You can also change the capacitor (C1) to change the scale of the oscillator (larger values to get lower and smaller values to higher frequencies).

After the resistor is removed solder two wires to the points where resistor was connected. Then solder other ends to one of the two potentiometers. One wire to middle pin and one to side pin. You might want to solder a resistor in series with the potentiometer so that the oscillators scale stays in audible frequency (try different values). Now moving joystick sideways or up and down will change the frequency of the sound, depending on witch potentiometer is wired .

step4. second oscillator

4069 circuit

Cut the traces from inverter to ground from pin 1 and pin 3. Also cut the trace between pins 8 and 5. Now the output of your first oscillator will be from pin 8 and you have ½ of the chip free. Now solder a capacitor between pins 1 and 4, connect pins 2 and 3 with a drop of solder, connect pins 4 and 5 with a drop of solder, wires from the unused potentiometer are connected to pins 1 and 6

74hc04 circuit

The original circuit uses ½ of the chip to create the oscillator. The three remaining inverters will be used to create the other. You can copy the first oscillators circuit: a capacitor between pin 13 and 10, connect pins 12 and 11 with a drop of solder, connect pins 10 and 9 with a drop of solder and the remaining potentiometer between pins 13 and 8. When completed the circuit is the same as on the 4069 joystick. And here is the diagram.

diagram 3. two oscillators wired to joysticks x and y potentiometers

step5. output and mixing

diagram 4. The mixing network either with resistors (left) or with diodes (right)

You can mix the signals to a single output from the two oscillators with resistors or diodes. Try both and choose the mixing network you like. Or you can use stereo output connecting one oscillator to right channel and one to the left channel. Solder a capacitor between the output wire from mixing network and output jack. The signal is quite strong so you might want to add volume control -> a resistor, trimmer or potentiometer to output line.

step6. buttons.

In most joystick there are two buttons. You can run the output from each oscillator before the mixing network through individual button. Now by pushing one button you can hear one osc and pushing both buttons you can hear them both. (See also 555 timer joystick step6. buttons)

step7. make some noise.

After all wires, components etc. are soldered properly and everything seems to be working. Close the screws and your noistick project is completed.