Originally Posted by
nocash
Glad that some new info has showed up!
Originally Posted by
Joe
The two ROMs are not directly interchangeable, since the '61 board rearranges some of the address and data pins relative to a standard EPROM pinout. I descrambled the 2.03 ROM (thanks, Farid) to compare it to 1.10, and the two are clearly based on the same code.
Good to know that there are least two versions confirmed to exist. Probably there are even more. With the data being stored in PROM/EPROM, they might have released thousands of different builts without raising the ROM production costs.
Are you sure that the older board didn't have the pins swapped in same fashion? (I haven't checked if the PCB photos are showing wiring differences).
Btw. here are the PCB photos from Jazz-disassemblies,
http://jazz-disassemblies.blogspot.d...ng-system.html
and here he dumped the rom, and found the hidden message
http://jazz-disassemblies.blogspot.d...rs-hotels.html
having never heard about those hotels, I was quite scared for a moment: Don't stay where? ; - )
Anybody desoldered the data ROM, and dumped that one, too?
Here I am. I was looking at some of my blog hits and found you guys.
I have been reviewing the public schematics and am interested in the sound portion and would like some information on some of the pins such as "OUT1, OUT2, OUT3 and OUT4" of the J004 IC as well as the "PWM" pin of the AS0012.
Envelope Generator
If my assumptions are correct, then the PWM signal is being sent through U4A which is wired as both an integrator and a comparator. I am almost certain that this pin is controlling the envelope of the main Sound. Lets say that the rate of oscillation is constant. I do not know the actual frequency, so we'll just say 1kHz for now.
1. When the pulse width is very wide (High for the majority of each period) then the capacitor, C26, will continue to charge at a rate set by R59 and C26.
2. When the pulse width is very narrow (low for the majority of each period) then the capacitor, C26, will continue to discharge at a rate set by R59 and C26.
3. When the pulse width is exactly 50%, the charge on the capacitor will essentially remain constant.
So by manipulating this idea, then attack, decay, sustain and release may be created which is then fed to the LM13600 OTA IC.
This is the first time seeing anything like it and I am rather impressed. The idea is pretty damn cool, but I could be entirely wrong about the purpose of PWM!
:
VCA - Voltage Controlled Amplifier
The LM13600 is obviously wired up as a voltage controlled amplifier. The basic idea is well described in the datasheet for the LM13600/13700 IC.
1. The Analog signal from the envelope generator described above is fed into the "Ibias" pin of both OTA's as for stereo control. This amplified the incoming signal depending on the current going into the Ibias pin.
Oscillator(s)
Here is what I am confused about; pins OUT1 through OUT4 from the J004 IC. Pins OUT1 through OUT3 must be the Digital to Analog conversions of the sound samples. At what amplitude, I do not know. It could be from 0v to 5v seeing as the samples are probably stored as data from $00 to $FF, but I will have to wire up my scope to be certain.
In any case, There are three pins that are mixed together into the two stereo outputs. OUT1 goes 'mostly' to the Left output, OUT2 goes equally to both left and right and OUT3 goes 'mostly' to the Right output. I say mostly because there will be some crosstalk where left and right leach into each other.
What is OUT4 used for? This pin skips the mixing stage and goes directly into the LM13600 at the noninverting input, whereas the other three signals go into the inverting input. If I am not mistaken, this will be added to the inverted input prior to amplification. This may act as an offset or something else altogether such as tremolo.
Mixing and Filtering
As I just mentioned, the channels are mixed together with different currents so that when they are converted to voltages with the resistor loads prior to the op amps, they will be at different volumes on the left and right output terminals.
This section; consisting of many passive components, the IC U10 and part of U4; also filters the sounds. A few active low pass filters and some passive low pass filters are used to filter out both noise and harmonics from the Tones.
Depending on the values of each component, this may simply be used to filter out noise from the power supply or to shape the tones of the actual sound samples. Some filtered sounds would look like this, where the first photo is the actual filter section of the piano and the second photo is filtered even more. These are called Tone or Wave shapers and I will be commenting on this below.
What I want to do
I want to remove the filter section so that I may add my own filter controls to the front of the keyboard as shown here:
There is room enough for several knobs, switches, LEDs and maybe even a 3.5mm input or output. Two knobs could be used to control "R" in the filter section. I could add both a high pass filter and low pass filter which could be switched on or off to produce Low Pass, High Pass, Band pass and even Band reject filters! A little more interesting would be if I could modulate the filter control with an analog signal found inside of the keyboard.
I suppose I could design an LFO controlled by the PMW output used in the VCA section. If I used an integrator with different values for C and R, then it would act much differently than the ADSR signal.
The other idea is to have several 3.5mm outputs that I could use to connect to my Euro rack modules.
1. Gate can probably be found pretty easily by looking at the keyboard controller IC.
Alternatively, I could use a comparator with hysteresis so that the signal rises when the attack sequence begins and drops when the release section begins. This would require some measurements of the actual envelope.
2. Trigger can be done with some diode logic. When Gate is triggered while still high, then trigger would jump high.
3. Envelope obviously! This can be directly taken from U4A.
4. CV will be more difficult since the frequency is determined internally to one of the proprietary IC's. Maybe I could use a LM13700 based Tachometer to convert the frequency to a voltage which could be scaled for my own purposes. I have nto built one before but the theory is pretty straight forward.
So does anyone have any input for the pins I am uncertain about? Out1-4 and PWM? The above information is only speculation and theory. I still have to probe around with my scope.
Also, does U12 need dumped? It is not a uv eprom, btu could it be a eeprom, or otprom?
︻╦╤─
old school
Posting Permissions
