Hohner KS 49 midi Casio HZ-600 SD synth with analogue filter & realtime programmable accompaniment RAM CARD

Hohner KS 49 midi

This keyboard from 1987 (embossed case date stamp 1986..1987) is the Hohner version of the Casio HT-700 synthesizer. Casio named its synthesis engine "Spectrum Dynamics" (SD) synthesis, but basically it is nothing else than the user programmable version of Casio's historical "Consonant Vowel" synth engine, which mixes 2 stair shaped waveforms with independent volume envelopes and which is famous for its warm and very sonorous droning multipulse squarewave bass organ timbres.
The HT-700 was likely designed as a successor to the great Casio MT-400V (see CT-410V) and thus also employs a real analogue VCF (low pass synth filter with cutoff/ resonance control) and stereo chorus. But additionally it has programmable sample based rhythm & accompaniment - and this all controllable through MIDI with 2 part timbrale sound. It sounds too great to be true - isn't it?

Unfortunately the thing is far less exciting than the feature list suggest, since for a serious synth the entire instrument sounds disappointingly cold, thin and digitallic; it resembles rather a Bontempi GT-759 than the much warmer sound of Casio's earlier "consonant vowel" keyboards. The problem may be that the HT-700 keeps its digitally controlled VCF always in the signal path (even when not in use), while older keyboards either lacked them or could completely bypass them through a mechanical switch. (Late 1980th ghettoblasters often had similar sound problems with their fancy looking and cheaply designed equalizer or reverb units.) But this gives the sounds also a cheesy lo-fi appeal that reminds to typical C64 sounds, and the thing can do a lot of wonderful multipulse squarewave organ timbres those remind to classic Atari POKEY sounds (but without the complex noise effects and special tone scale), and some waveforms even include that small dose of distortion that is necessary to make polyphonic squarewave musics sound really good.

There are 40 preset sounds, of those 20 can be overwritten by the user. The preset sound and rhythm set differs between Casio HT-700 and Hohner KS 49 midi; at least with my KS 49 most preset sounds are quite unrealistic, but realistic timbres are anyway no great talent of this hardware. A special RAM Card (Casio RA-100, now hard to find) can be inserted to store 20 additional user presets and user rhythms etc., but there is no MIDI sysex feature to backup data elsewhere. The percussion is made from medium resolution samples; they resemble the Casio MT-520 drum section but sound thinner, brighter and knock less woody. The bongos e.g. sound here more like drumming with fingers on the floppy aluminium foil lid of a yoghurt cup. The preset accompaniments of the KS 49 sound quite cheesy and many are over- orchestrated; their chord timbres overuse filter wahwah effects, which makes them rather resemble cheap no-name FM tablehooter accompaniments than realistic instrument timbres. Fortunately the fingered accompaniment at least accepts also non- standard key combinations beyond establishment chords, and the rhythm programming mode can be abused as a tekkno drum computer (although the percussion sounds rather boring for this and can not be routed through the VCF).

There is only a single common VCF used for all main voice polyphony channels, but a 2nd VCF exists for the also editable accompaniment voice on the left keyboard half. All synth parameters and tempo are selected and set through 2 buttons and a dial knob. Unlike the endless rotary encoder of other digital synths, the dial here turns a normal potentiometer, thus the value jumps to the current pot position when touched, and you have to turn it all way back and tweak the previous value when you only want to change a parameter a little; a normal number keypad would have been far less awkward here. Also the LCD only shows parameter numbers (0..95) instead of names. At least you can nicely abuse this dial as a realtime sound control for a previously selected parameter (e.g. filter cutoff), which often makes grainy zipper noise and can be also used for granular timbre change effects. But it takes many seconds to assign a different parameter to it, which disturbs live play. There is also a pitchbend-, but no modulation wheel, and synth parameters can not be accessed through MIDI.

main features:

modifications:

  • polarity protection diode added, power supply jack polarity corrected.

    • notes:

    Strange is why Casio released their HZ- and HT-series SD synthesizers at all, since their own phase distortion synthesizers could do all the same sound effects and many additional ones with a much higher timbre quality. I doubt that SD synths were built for a lower manufacturing cost, since at least my Hohner KS 49 midi (almost identical with Casio HT-700) contains a very complex hardware with multiple stacked PCBs and many discrete components, that doesn't look in any way cheaper or easier to manufacture than the PCBs of my Casio CZ-230S phase distortion synth. Adding a similar simple parameter editor to the latter would not have increased cost much, and also the additional accompaniment would have been only a matter of ROM software. Nowadays the HT synths have their own fandom by their electronic squarewave sound, but when they came out in 1987, Casio certainly did not design them as a dedicated special purpose instrument for such customers, since there was no market for such sounds anymore upon a time where most people rather threw their outdated Commodore C64 computers out in favour for a new Amiga.

    All SD synthesizers are based on the the sound IC "NEC D935G" with analogue 4-pole VCF "JRC 2090D" controlled by a "NEC D78C10G" CPU with external RAM and ROM. While regular versions employ only 1 sound IC and 2 VCF (1 for chord and melody section), the high-end model Casio HT-6000 contains even 4 sound ICs with 8 VCF (1 per polyphony channel) and additionally an MCS-48 CPU for velocity(?) and 2 gate array SMD ICs for bus handling. I expect that the D935G gets all waveform data from the CPU, since it has no software number and the count of available waveforms differs between HT-6000 and other models, but it may be that HT-6000 only layers the same standard waveforms of multiple ICs.

    The sound engine of the HT-700 was far inferior to the CZ synths; e.g. while the CZ synths can do a lot of different hiss timbres by "noise modulation", the noise generator of the HT can not change its fixed hiss timbres (except through the monophonic VCF), and its noise pattern has always a tremolo with about 6Hz in it, which is likely caused by a too short shift register loop pattern (like with certain Atari POKEY noises). The main voice oscillator of the HT-700 can generate 32 different basic "DCO" waveforms, those all sound like based on filtered multipulse squarewave. Any of the 32 basic waveforms seems to be made from 2 layered suboscillators those each generate a stair shaped(?) waveform with independent envelope, but the internal parameters of these suboscillators can not be set directly, but only the 32 resulting basic waveforms. While most of these are simply static squarewave tones with different multipulses run through different filter capacitors, there are also special ones. E.g. number 6 is pure noise, some others mix that noise with a squarewave tone and 23 even automatically slowly grows brighter by crossfading between 2 suboscillators in the classic "consonant vowel" manner. Also some others contain quick timbre changes during their attack phase to approximate mallet, flute or picked string instruments, but they always keep their multipulse squarewave appeal. Just disable the VCF (set cutoff=31, resonance=0), set chorus to 0 or 1, select a squarewave with simple envelope and voila - you get the nicest historical videogame music timbre; in some waveforms even the typical mild distortion is there, that gives polyphonic squarewave music the classic Atari VCS2600 appeal. The programmability of the accompaniment voice (left keyboard section) is far more restricted than the main voice (only 16 basic waveforms, less parameters), and annoying is also that the octave range of the oscillators can not be changed, because their pitch is too high to create really low bass sounds. The bass voice of the accompaniment can not be edited at all.

    Like with CT-410V, the sound generator itself is mono and only post- processed by the stereo chorus, which adds to the main and accompaniment voice a cold sounding detuned chorus that reminds to My Music Center timbres; unlike the stepless slider of the CT-410V it can be set only to 3 panning/ vibrato speeds or turned off (but apparently always stays in the signal path), and the chorus setting is here part of each preset sound. Strange is that my Casio MT-800 and MT-520 sound much warmer despite active stereo chorus, thus possibly it is only the VCF or even a poorly designed power amplifier that ruins the sound of my KS 49. Despite the data entry dial also the resolution of some other parameters is quite coarse, which appears especially unnecessary with the rhythm tempo setting. Instead of the awkward tempo/ entry dial, Casio should have better used the semi- OBS buttons as a keypad to type in parameters. Likely at least the analogue filter and chorus section of the the HT-700/ KS 49 can be easily upgraded with external potentiometers for better realtime controls, and I expect that like with my CT-410V there are many other things changeable through circuit bending. (I haven't analyzed the hardware closer yet.) Very unusual for Casio hardware is that the buttons of my KS 49 employ instead of rubber contacts metal cap switches, despite these tend to be less reliable (well known from the infamous original Atari VCS2600 or Spectravideo Quickshot I joysticks, those contacts often cracked apart or slid away under their adhesive film cover on the PCB). A bit confusing is also that the accompaniment control buttons {single finger, fingered, off} don't work so long any keys are held (likely a bug related to semaphores because a double press on "fingered" or "single finger" toggles the key split point).

    On the internet is a lot of info available about the Casio HT-700, thus I won't describe its operation and features in detail here.

    While most preset sound names on Casio HT-700 correspond to many of the Casio CZ-230S presets, on the Hohner KS 49 midi several sounds are different; some of them differ even so much (e.g. "plunk extend" vs. "psycho", "pearl drop" vs. "harmonium") that it is unlikely that only the control panel names were changed, although some of them may be only in different order. With Casio HT-700 the presets are:

    The Casio HT-700 has a black case and also some control panel names differ among Casio and Hohner ("tone"/ "melody tone" = "poly voices", "chord tone" = "arr. voices", "tone edit" = "voice edit", "control" = "total control", "rhythm pattern" = "rhythm", "bass pattern" = "bass", "chord pattern" = "chord", "casio chord"/ "chord"/ "accomp" = "arrangeur", "chord/ operation memory" = "sequencer", "main volume" = "master volume").

    Generally the preset sounds are not really realistic; in the bass range many sounds turn into a more or less buzzy, sonorous purring drone, which is a characteristic style element of squarewave based instruments. When a key is trilled with sustaining sounds, each new note occupies a new sound channel, which produces a great phasing sound and volume increase effect although this eats up polyphony. The semi- OBS preset sound buttons have each a LED and respond nicely fast; pressing them retriggers the notes of held keys with the selected preset sound, which can be used for live play tricks. The "piano" has to much presence with a thin bass and slightly dull treble range. The "harpsichord" is a multipulse timbre with a nicely sonorous (although unrealistic) bass range; "clavinet" (Hohner invented that thing) is the same without sonorous multipulse. The "jazz organ" (resembles Bontempi GT-770 "circus organ") has a percussive attack and strong fast 6Hz square vibrato and chorus vibrato, but it fades thinner and generally sounds ridiculously thin and creaky despite sonorous multipulse bass range. The "brass" is a cold tooting electronic chorus tone with semi- percussive attack, which resembles the famous Casio SA-series "brass ens". "strings" employs the same timbre with slow attack, long sustain and 4Hz tremolo; unlike the latter it sounds astonishingly realistic. "wow brass" is a "brass" variant with filter meow effect and -tremolo; it turns bright and dull again and reminds to typical cheap FM synth brass sounds. "synth celesta" is bright squarewave musicbox timbre with nicely sonorous multipulse bass range and 4Hz tremolo. "celesta" sounds more massive and resembles more an unfiltered squarewave pseudo- piano. "magic flute" is a dull flute (or wooden organ pipe) with a small dose of chorus; the dull bass range resembles a grainy very low resolution sine wave sample. "harmonium" has an additional 2 octave higher overtone and thus sounds brighter. "harmonica" is a sonorous reed organ timbre (resembling sawtooth wave) that fades slightly thinner. "vibes" is a bright vibraphone timbre with long sustain made from filtered multipulse squarewave; "soft bells" sounds slightly brighter and shorter. "e-piano" is a duller "vibraphone" variant with only short sustain. "wah-brass" is a quickly dull fading dry filter sweep timbre with fast 6Hz tremolo; it resembles much rather a funky e-bass than a brass timbre. "crescendo" resembles a harp or dull xylophone layered with a louder and brighter fading synth strings timbre with fast 6Hz tremolo and some sustain (nice pad sound). "fantasy" has here nothing to do with any Casio sounds of that name, but sounds like a water drop in a cave layered with a short dull flute or xylophone tone (multipulse timbre) with sustain (that is slightly shorter with held notes). "psycho" is a sort of upward howling and brighter growing US police car siren, that with held key stops after 4 howls (about 3s), but slowly fades duller and fades away (while it keeps howling) when the key is released earlier. "storm" is an eerie wind noise that hisses brighter until key release (or 2s) and then duller using a filter sweep while it fades away with very long 10s sustain. There is also some chorus and a continuous 4Hz tremolo in it (caused by the poor shift register noise generator), which sounds very much like a typical classic analogue 1960th or 70th SF movie sound effect. Its fixed pitch unfortunately ignores the played note, but pressing multiple keys causes a dramatic volume increase and opens the (monophonic) VCF again each time.

    Additionally there are 20 overwriteable default user preset sounds, those original names I don't know because I only have the downloaded Casio HT-700 version of the manual. Thus the following names were chosen by me where the Casio names don't match. "piano 2" is a bit brighter than 1. "marimba" is a bit too dull. "pipe organ" is the well known multipulse timbre that simulates a metal pipe organ rank; this one has long sustain and sounds a bit thin. "strings 2" has a faster attack and more mids than 1. "wow brass 2" resembles 1 but grows only dull. "harp" sounds ok with some chorus. "koto" is a bright picked string that fades duller during attack (like intended) with some sustain/ chorus. "double reed" is a dull rather thin reed organ timbre with some chorus. The "clarinet" is ok with some chorus. "fantasy 2" sounds like a thin harp layered with a howling flute that slowly fades silent and has short sustain. "honkytonk piano" sounds like "piano 2" with howling detuned chorus vibrato. "vibraphone 2" sounds duller than 1. "violin" sounds fairly realistic with some chorus. "synth-strings" sounds like "strings" with a small dose of detuned chorus vibrato. "synth trumpet" is a buzzy squarewave trumpet timbre with percussive attack and sustain. "mandolin" doesn't ring and has a percussive metallic treble range like a toy piano; the mids rather resemble a mandolin or small harp. "sitar" is a bright picked string with chorus and sustain; the buzzy mids are rather a decaying synth brass timbre (made from sawtooth waves?). "slash reed" is a thin and squawky reed organ(?) timbre, which scratchy attack phase somewhat resembles a saxophone. "synth-guitar" is a thin picked string timbre, which fades duller by a filter sweep (like an e-guitar wahwah effect); it has no sustain and immediately stops by key release. "storm 2" resembles 1 but starts bright instead of dull and sounds thinner and less spectacular.

    The preset accompaniments of the KS 49 sound quite cheesy and many are over- orchestrated; especially their chord timbres overuse filter wahwah effects, those resemble 1970th funk "wokachika" sounds. The accompaniments and rhythms are programmable and there is an additional bass track with a fixed monophonic e-bass sound. Nice is that the rhythm programming mode can be also abused as a tekkno drum computer, although the percussion sounds boring. Like with a Yamaha "custom drummer" mode, a template pattern repeats in a loop and you can add individual percussion in realtime through keyboard keys (press simultaneously leftmost ">" key for loud accent notes). Holding the "record/ delete" button simultaneously with a particular percussion key deletes the corresponding percussion track from the user pattern only for that duration the key is pressed, which nicely permits small changes in the running pattern. Pressing "record/ delete" with stopped pattern clears the entire pattern. Unfortunately the thin drums can not be routed through the VCF (the old Casio CT-410V still could do this). The accompaniment sound depends on the currently selected preset rhythm (also with rhythm off). Annoying is that despite in fingered chord mode the left keyboard section can be expanded to the entire keyboard half, it can not be used as a real 2nd main voice, because it transposes any played notes into a single octave and always attempts to hold chords so long any key in that section stays pressed. Also the maximum chord volume is a bit too low as 2nd main voice. Running rhythm always starts the automatic accompaniment and can not be combined with manual chord mode (e.g. as a 2nd main voice) - outsider keyboards like Antonelli 2495 were much more versatile here. The programmable accompaniment also only accepts 8 standard chords, those are entered in a strange way through the rightmost 8 white keys.
     

    hardware details

    The Hohner KS 49 midi is a rebranded variant of Casio HT-700, that is member of the HT-3000 hardware family. The complex multi-chip hardware is built around the CPU NEC D78C10G that controls the SD sound IC NEC D935G and 2 percussion ICs OKI M6294-03 and -04. The melody and chord voice is routed from the sound IC through each a VCF JRC 2090D, followed by the stereo chorus. The sound ic also controls the LCD.

    This text is based on the service manuals of Casio HT-3000, HT-700, HZ-600, MT-600 and CT-630.
     
    The rom "NEC D23C256EAC" of my Hohner KS 49 midi has software number 059, while Casio HT-3000 and HT-700 schematics shows 39, and the actual HT-700 parts list even has 019. Thus Casio and Hohner software definitely differ. Unlike Hohner PK60 the CPU itself is romless, hence a rom swap should be sufficient to replace the preset sounds. The Casio CT-630 (non-synth variant of HT-3000) in service manual part list has software number 023.
     
    software number model notes & features
    019 HT-3000, HT-700 original version (synth, uses percussion ICs M6294-03 and -04)
    023 CT-630 no synth, HT-3000 shape
    39 HT-3000, HT-700 (in schematics)
    044 HZ-600 no rhythm/chord/accompaniment, no speakers
    059 Hohner KS49 midi like 019, HT-700 with changed preset sounds
    65 HT-600 no synth, HT-700 shape, only 1 percussion IC M6294-06, 2 gate arrays for keyboard matrix
    ? Hohner PK100 like 023, CT-630 with changed preset sounds

    Exciting is that beside MT-600 the wiring of all instruments is very similar and the CPU is controlled by only one external rom; custom chips (beside percussion ICs) have no software numbers, which makes the system open for experiments with rom swaps or even custom software by replacing the rom with a socketed 32KB eprom (27C256) to change preset sounds or (more interesting) implement proper midi sysex dump support to get rid of the rare ram cards and define synth parameters through midi.

    There is plenty of address decoding going on. The CPU e.g. selects ram, sound IC, percussion IC 1 & 2 and LED controller 1 & 2 through the binary number on address lines A14, A13, A12 while A15=hi. A15=lo selects ROM. Lower Address bits A0..A7 are latched from data bus during a pulse on ALE. If I understand the address decoder tables correctly, this is the memory map.
     
    0000h external ROM (32KB)
    8000h RAM (8KB)
    A000h Music LSI
    B000h percussion IC1
    C000h percussion IC2
    E000h LED controller
    FF00h ? CPU internal RAM (256 byte)

    The RAM Card RA-100 is not in address space but accessed through I/O pins C0..C7. Lower address is latched during a pulse on C11. Then upper address is latched during a pulse on C12. /CE is generated by C5=lo during rising edge of C12. RAM Card insertion is detected when C15=lo.

    The sound IC D935G generates melody and chord channel. CPU pin 21 (N) switches the routing, to send 4 of the polyphony channels through a different VCF when in chord or keysplit mode. The VCF IC JRC 2090D contains a 4-stage low-pass filter. It is voltage controlled by the sound IC. The melody voice VCF cutoff is pins AO5 (increases from 2.5V to 0.8V), envelope is AO4 (0.5V to 3.5V) , resonance is AO6. The chord voice VCF cutoff is pins AO2, envelope AO0, resonance AO2.

    The VCF adjustment is complicated because the original text is useless without schematics, so I try to summarize it naming the IC pins.

    For VCF cutoff adjustment, ground the VCF IC (JRC 2090D) pins 3 and 9. Connect a signal generator through a capacitor to pin 18 and inject 0.8V Vrms +/-1% 590Hz +/-1% signal. Check with a level meter at pin 12 is 0.8V +/-2%. Apply 2.0V +/-1% to pin 3 (of course after removing your grounding test lead). Adjust the trimmer until the voltage at pin 12 is 0.2 Vrms +/-5%. Reduce the signal generator output to 0 Vrms. Now the pin 12 voltage has to be between 1.8 and 2.3V. This procedure needs to be performed for both VCF. The trimmer for Melody VCF is VR2, for Chord VCF is VR4.

    For VCF bias adjustment, set on the keyboard the entry dial in mode = 10 (Melody VCF Cutoff Frequency), data = 0. Connect a voltmeter to pin 3 of the IC and adjust trimmer VR5 until the voltage is 2.5V +/-20mV. Set the entry dial in mode = 10, data = 31. Verify that the voltmeter shows 0.8V +/-20mV.

    The mute circuit is in a strange way resusing the BAS (bass squarewave tone) pin of the sound IC in what looks like a kind of handshake.
     
    The HT-3000 service manual claims:
    1. When Music LSI does not generate melody sounds, it falls signal /MUT LOW.
    2. CPU then sends melody mute command via the data bus.
    3. Upon receipt of the command, Music LSI falls signal P0 to turn FET T1 on.
    4. Noises which may output from Melody VCF is drained to ground.
    5. If Music LSI does not create chord and bass, it falls signal P3.
    6. Transistor T11 turns on sending MUT2 signal to CPU.
    7. CPU transmits chord/bass mute command through the data bus,
    8. Music LSI falls signals P1 and P2.
    9. FET T2 turns on cutting Chord VCF noises.
    10. By the signal P2, Bass circuit noises are eliminates.
    (FET T1 pulls melody VCF intput to GND. FET T2 pulls chord VCF intput to GND. FET T11 pulls BAS (bass waveform ar sound IC and MUT2) to GND.)

    The stereo chorus is a phase shifting delay circuit with plenty of analogue stuff, using a 1024 steps BBD IC MN3207 clocked by IC MN3202. It stays off when both CPU pins 19 (CH0) and 20 (CH1) = lo, shallow when CH0 = hi and deep with both hi.

    keyboard matrix

    The matrix outputs KC1..KC11 are inverted from pins C0..C10 of the CPU. The keyboard inputs KI1..KI6 are latched inverted to the data bus during the inverted pulse at 1Y0 from the LED controller. The same happens with panel inputs SI1..SI6 and pulse at 1Y1.

    I haven't analyzed this matrix by myself. This is based on the Casio HT-3000 and HT-700 service manuals. Matrix places those are not used in HT-700 I  have marked as eastereggs. So the HT-3000 has an own group of preset sound buttons for lower (chord section) keys and a longer keyboard. Some preset sounds of Hohner KS 49 midi differ from HT-700 due to different rom.
     
    KC1
    KC2
    KC3
    KC4
    KC5
    KC6
    KC7
    KC8
    KC9
    KC10
    KC11
     
    pin
    out 1
    out 2
    out 3
    out 4
    out 5
    out 6
    out 7
    out 8
    out 9
    out 10
    out 11
    out / in
     
    o
    C1
    o
    F#1
    o
    C2
    o
    F#2
    o
    C3
    o
    F#3
    o
    C4
    o
    F#4
    o
    C5
    o
    F#5
    o
    C6
    in 1
    KI1
    o
    C#1
    o
    G1
    o
    C#2
    o
    G2
    o
    C#3
    o
    G3
    o
    C#4
    o
    G4
    o
    C#5
    o
    G5
     
    in 2
    KI2
    o
    D1
    o
    G#1
    o
    D2
    o
    G#2
    o
    D3
    o
    G#3
    o
    D4
    o
    G#4
    o
    D5
    o
    G#5
     
    in 3
    KI3
    o
    D#1
    o
    A1
    o
    D#2
    o
    A2
    o
    D#3
    o
    A3
    o
    D#4
    o
    A4
    o
    D#5
    o
    A5
     
    in 4
    KI4
    o
    E1
    o
    A#1
    o
    E2
    o
    A#2
    o
    E3
    o
    A#3
    o
    E4
    o
    A#4
    o
    E5
    o
    A#5
     
    in 5
    KI5
    o
    F1
    o
    B1
    o
    F2
    o
    B2
    o
    F3
    o
    B3
    o
    F4
    o
    B4
    o
    F5
    o
    B5
     
    in 6
    KI6
    O.
    piano
    O.
    synth. bells
    OL.
    brass ens.
    R.
    rock 1
    R.
    swing 1
    O.
    preset
    OL.
    preset
    R.
    preset
    R.
    fill-in
    		    
    in 7
    SI1
    O.
    elec. piano
    O.
    magical wind
    OL.
    jazz organ
    R.
    8 beat
    R.
    slow rock
    O.
    internal
    OL.
    internal
    R.
    internal/ card
    R.
    synchro
    		    
    in 8
    SI2
    O.
    jazz organ
    O.
    blues harmonica
    OL.
    elec. piano
    R.
    16 beat 1
    R.
    samba
    O.
    card
    OL.
    card
    C.
    normal
    R.
    start/ stop
    		    
    in 9
    SI3
    O.
    brass ens.
    O.
    light harp
    OL.
    clarinet
    R.
    disco 1
    R.
    bossa nova
    mode
    C.
    auto harmonize
    C.
    split
    R.
    intro/ ending
    		    
    in 10
    SI4
    O.
    symphonic ens.
    O.
    plunk extend
    OL.
    synth. ens.
    R.
    pops 1
    R.
    waltz
    data/ entry 
    modulation
    on/ off
    C.
    fingered
    S.
    select
    		    
    in 11
    SI5
     
    O.
    select
    OL.
    select
     
    R.
    select
     
     
     C.
    single finger
    S.
    record
    		    
    in 12
    SI6

    The input lines are active-low, i.e. react on GND. Any functions can be triggered by a non- locking switch in series to a diode from one "in" to one "out" pin.
     

    legend:
    "o"
    		 = keyboard key
    underlined
    		 = function needs locking switch (i.e. stays active only so long the switch is closed)
    R.
    		 = preset rhythm
    O.
    		 = preset sound upper keys ('orchestra')
    OL.
    		 = preset sound lower keys
    C.
    		 = chord
    S.
    		 = sequencer
    orange
    background     		 = easteregg (unconnected feature)
    grey
    background    		 = unconnected doublet

    The pitch wheel is at CPU pin 34 (AN0). The additional modulation wheel of HT-3000 is wired to CPU pin 35 (AN1). (Adding it will likely need an additional "Modulation on/ off" button at SI5->KC7.) Theoretically it might be possible to upgrade non-synth models also with synth buttons and data entry dial (CPU pin 35 (AN1)), but without LCD (and RAM card slot) it would be pointless.

    The CT-630 matrix in service manual has same layout but different preset sounds and lacks synth features. It is unknown how much of it may be reviveable as matrix eastereggs, since the hardware lacks the 2 upper address latch ICs (TC74HC374P and TC74HC174P) and the rom has a different software number 023.

    The MT-600 is even more crippled. In its service manual it has only 1 percussion IC M6294-06 and 2 additional gate arrays UPD65005GF-151 (for panel matrix and LEDs) and -152 (for keyboard matrix) containing latches, buffers and decoder. It is hard to understand why; apparently they translate the matrix layout for a case with wrong control panel that was not designed for it (using 10 buttons to select 40 preset sounds and 10 buttons for 20 rhythms), so I doubt that it has eastereggs common. The rom is software number 065.

    The HZ-600 matrix layout otherwise is clearly related to HT-3000. In service manual it lacks all parts related to chord and rhythm (it only has keysplit), and some preset sounds differ. Interesting is that at matrix place SI6->KC6 it has a "sustain" button.

    pinout JRC 2090D

    The "JRC 2090D" (18 pin DIL, in schematics named "NJM2090") is a voltage controlled filter (VCF) used in Casio keyboards. It contains a 4-stage low-pass filter. Cutoff frequency reduces when pin 7 voltage (0.8 .. 2.5V) increases. Resonance is proportional to voltage of pin 9 (0.5 .. 2.2V).
     
    pin name purpose
    1 V+ op-amp (+) in
    2 V- op-amp (-) in
    3 OP-OUT op-amp out
    4 GND ground 0V
    5 Adj. (wiper of VRef trimmer)
    6 VRef (trimmer to GND)
    7 FC cutoff frequency cv in
    8 FT (resistor to GND)
    9 QC resonance cv in
    10 QT1 (resistor to GND)
    11 QT2 (resistor to GND)
    12 OUT audio out
    13 +Vcc supply voltage
    14 CF4 (capacitor to GND)
    15 CF3 (capacitor to GND)
    16 CF2 (capacitor to GND)
    17 CF1 (capacitor to GND)
    18 IN audio in

    pinout D78C10G

    The CPU "NEC D78C10G" (64 pin zigzag DIL) was used in Casio SD synthesizers. It is a documented romless 8-bit microcontroller, which behaviour depends on software in external rom. It has  256 byte internal RAM.

    This is the pin meaning when used in instruments of the Casio HT-3000 hardware family, that includes almost all Spectrum Dynamics instruments like HT-700 (with variant Hohner KS 49 midi), HT-3000, HZ-600, but also MT-600 and CT-630. Surprisingly they use their CPU in a very similar way despite changes in rom contents. Regard that the pinout in the much more complex Casio HT-6000 strongly differs.
     
    pin name CPU pin purpose
    1 C0 PA0 key matrix in, led drive, ram card bus
    2 C1 PA1 key matrix in, led drive, ram card bus
    3 C2 PA2 key matrix in, led drive, ram card bus
    4 C3 PA3 key matrix in, led drive, ram card bus
    5 C4 PA4 key matrix in, led drive, ram card bus
    6 C5 PA5 key matrix in, led drive, ram card bus
    7 C6 PA6 key matrix in, led drive, ram card bus
    8 C7 PA7 key matrix in, led drive, ram card bus
    9 C8 PB0 key matrix in, led drive, ram card bus
    10 C9 PB1 key matrix in, led drive, ram card bus
    11 C10 PB2 key matrix in, led drive, ram card bus
    12 C11 PB3 key matrix in, led drive, ram card bus
    13 C12 PB4 key matrix in, led drive, ram card bus
    14 C13 PB5 ram card read /enable
    15 C14 PB6 ram card write /enable
    16 C15 PB7 ram card detect /in
    17 PC0 PC0/TXD midi in
    18 PC1 PC1/RXD midi out
    19 CH0 PC2/SCK stereo chorus enable1 out
    20 CH1 PC3/ /INT2 stereo chorus enable2 out
    21 N PC4/TO chord/melody mode changeover out
    22 MUT1 PC5/CI melody vcf mute request in
    23 MUT2 PC6/CO0 chord vcf mute request in
    24 APO PC7/CO1 auto-power-off /out
    25 OFF /NMI power-off /in
    26 DWN INT1 low battery voltage detect in
    27 MODE1   mode 1 in | memory cycle 1 out
    28 /RESET   reset
    29 MODE0   (wired through 100k to VDD)
    30 X2   crystal out
    31 X1   crystal in | clock in (12 MHz)
    32 VSS   ground 0V
    pin name CPU pin purpose
    33 AVSS   port T threshold voltage in (ADC ground)
    34 AN0   pitch wheel in
    35 AN1   modulation wheel in (HT-700: not used)
    36 AN2   entry dial in
    37 AN3   low battery voltage sense in
    38 AN4   analogue in (not used)
    39 AN5   analogue in (not used)
    40 AN6   analogue in (not used)
    41 AN7   analogue in (not used)
    42 VAREF   ADC reference voltage
    43 AVDD AVCC ADC supply voltage +5V
    44 /RD   read strobe out
    45 /WR   write strobe out
    46 ALE   address latch out
    47 A8 PF0  port F I/O | address bus bit 8
    48 A9 PF1  port F I/O | address bus bit 9
    49 A10 PF2  port F I/O | address bus bit 10
    50 A11 PF3  port F I/O | address bus bit 11
    51 A12 PF4  port F I/O | address bus bit 12
    52 A13 PF5  port F I/O | address bus bit 13
    53 A14 PF6  port F I/O | address bus bit 14
    54 A15 PF7  port F I/O | address bus bit 15
    55 D0 PD0  port D I/O | address / data bus
    56 D1 PD1  port D I/O | address / data bus
    57 D2 PD2  port D I/O | address / data bus
    58 D3 PD3 port D I/O | address / data bus
    59 D4 PD4  port D I/O | address / data bus
    60 D5 PD5  port D I/O | address / data bus
    61 D6 PD6  port D I/O | address / data bus
    62 D7 PD7  port D I/O | address / data bus
    63 VDD   RAM backup power supply | /stop (wired to VCC)
    64 VCC   supply voltage +5V

    Like pitch and modulation wheel, the data entry dial is simply a potentiometer with 0..5V output to an ADC. Because in some modes it directly changes sound, an analogue modulation or envelope circuit may be wired to its input.

    pinout D935G

    The Music LSI "NEC D935G" (80 pin SMD) is the Spectrum Dynamics sound IC of Casio HT/HZ series keyboards. It is 8 note polyphonic with digital envelope and can route 4 polyphony channels to a separate output to play melody and chord (or keysplit) voice. Additionally it has a squarewave bass tone & envelope output, and control voltage outputs for 2 VCF (each cutoff, resonance, envelope). In total it contains 12 DACs. Those for melody & chord are each 7-bit using separate DAC outputs for upper and lower bits those need to be combined through an external voltage divider (820+130 ohm resistor for ratio 1:158, add MSB (highest bit) inverted or uninverted with trimmer from 2 separate pins). So the resolution is likely 15-bit. The DCO waveforms include white and metallic digital noise. The DCA envelopes have ADSR parameters and can crossfade between waveforms. The LFO has waveform, depth, delay and speed settings. The VCF port supports cutoff and resonance and has ADSR envelope with depth too. In HT-6000 it can do ring modulation. Many parameters have 5-bit (32 steps) resolution; possibly they got artificially limited only by the ADC resolution of the data entry knob. It is hard to identify how much of all this is software defined or hardcoded inside the chip. As a real sound IC the D935G is controlled by an external CPU.

    The D935G was basically the last and most sophisticated version of Casio's classic Consonant-Vowel-Synthesis engine. As successor of the D931C it handles 2 VCF instead of switching between fixed filters and was likely more optimized for use in synthesizers. And like the old D77xG it even has an LCD port.
     
    pin name purpose
    1 /WR write (read cpu data)
    2 /I1 command read in (from cpu pin A10 | HT-6000: from gate array B)
    3 /I2 data read in (from cpu pin A11 | HT-6000: from gate array B)
    4 /CS chip select
    5 D0 data bus
    6 D1 data bus
    7 D2 data bus
    8 D3 data bus
    9 D4 data bus
    10 D5 data bus
    11 D6 data bus
    12 D7 data bus
    13 /RST reset
    14 SA4 lcd out
    15 SB4 lcd out
    16 SC4 lcd out
    17 SA3 lcd out
    18 SB3 lcd out
    19 SC3 lcd out
    20 SA2 lcd out
    21 SB2 lcd out
    22 SC2 lcd out
    23 SA1 lcd out
    24 SB1 lcd out
    25 SC1 lcd out
    26 CMX lcd out
    27 CMY lcd out
    28 CMZ lcd out
    29 VDD supply voltage +5V
    30 CLK clock/2 out (when I3 is wired lo)
    31 PG2 clock in (19 MHz)
    32 PG1 clock out
    33 GND ground 0V
    34 I3 CLK pin mode (wired lo = out)
    35 IO0 (not used)
    36 IO1 (not used)
    37 IO2 (not used)
    38 IO3 (not used)
    39 IO4 (not used)
    40 IO5 (not used)
    pin name purpose
    41 IO6 (not used)
    42 IO7 chord & melody A audio out (MSB)
    43 IO8 chord & melody A audio out (/MSB)
    44 NC (not used)
    45 GND3 ground 0V
    46 O3 chord & melody A audio out  (lower bits)
    47 GND2 ground 0V
    48 O2 chord & melody A audio out (upper bits)
    49 NC (not used)
    50 O1 melody B audio out (upper bits)
    51 GND1 ground 0V
    52 O4 melody B audio out (lower bits)
    53 GND4 ground 0V
    54 VDAC1 analogue supply voltage
    55 IO9 melody B audio out (MSB)
    56 IO10 melody B audio out (/MSB)
    57 VDAC2 (wired to 54)
    58 VDAC8 (wired to 54)
    59 GC
    60 GND5 ground 0V
    61 O9 (wired to pin 62)
    62 AIN model select in? (from pin 62)
    63 AO0 chord vcf envelope cv out
    64 AO1 chord vcf cutoff cv out
    65 AO2 chord vcf resonance cv out
    66 AO3 melody vcf resonance cv out
    67 AO4 melody vcf cutoff cv out
    68 AO5 melody vcf envelope cv out
    69 AO6 bass envelope out | HZ-600, HT-6000: (not used)
    70 AO7 stereo chorus lfo out
    71 P0 melody vcf /mute out | HZ-600, HT-6000: (not used)
    72 P1 chord vcf /mute out | HT-6000: (not used)
    73 P2 bass /mute out | HZ-600, HT-6000: (not used)
    74 P3 chord+lower+bass /mute request out |  HZ-600, HT-6000: (not used)
    75 /MUT melody /mute request out | HT-6000: melody B /mute out
    76 GND7 ground 0V (for pins 71..75)
    77 BAS bass wave out | HT-6000: melody A /mute out | HZ-600: also lower voice request?
    78 WTN (not used)
    79 MTN (not used)
    80 TST test (wired to ground 0V)

    Pin 77 (BAS) in HT-700 and HT-3000 service manual is described as "bass pitch signal", which looks confusing; the attached line is named "MUT2" and in schematics connected to CPU pin 23 (chord voice mute request). In keyboards without this bass it is indeed described as a melody mute control pin. The HT-700 block diagram text tells that BAS is the "pitch" (apparently squarewave), the bass envelope output from A06 gets mixed in a transistor, and while the sound IC generates no sound, it pulls P3 lo to eliminate noise. This makes no sense, unless it does something complicated like automatically muting the chord voice during bass or such things, or the CPU simply ignores pin 23 when it wants bass with chord.

    Pin 30 is clock out at half frequency if I3 is wired lo (done here), but becomes clock in if I3 is wired hi. The undocumented pin 62 is wired to 61, which looks like a secret model select feature. But I saw this in service manuals of HT-700, HT-3000, CT-630, MT-600, HZ-600 and eventually even on photos from inside my HT-6000, so it can't be a mode that was actually used.

    The Casio HT-6000 is much more complex and uses 4 of these sound ICs (each driving 2 VCA). It assigns them differently and so names the combined chord & melody outputs "melody A" and the "melody" outputs "melody B". It does not use the dedicated bass channel, because it layers the outputs from each 2 sound ICs for upper and lower (chord) keyboard section and so uses one of the melody channels as bass. In HT-6000 the LCD has 8 digits and is operated by sound IC 1(all pins) and 2 (omitting pins 26..28).

    A HT-700 variant without editable parameters was released as Casio MT-600 (similar case without LCD and entry dial). Casio also released an 61 fullsize keys variant of their HT-700 as Casio HT-3000, which had an additional modulation wheel (that may be addable to the HT-700 and KS 49 as a hardware easteregg); a light grey version came out as Hohner KS 61. A HT-3000 variant without editable parameters was the Casio CT-630 (same case, different button layout, no pitch/ modulation wheels). A speakerless HT-3000 without accompaniment was the Casio HZ-600. Another (much simpler) midsize keys synth with halfway similar digital ice-age sound is the Elta KE-491.

    Casio HZ-600

    This rare instrument from 1987 (embossed case date) was the earliest of all SD synthesizers.

    Technically it is a Casio HT-3000 without accompaniment (nor rhythm, only keysplit), no sequencer and no speakers. The main voice engine is the same like Hohner KS 49 midi with additional modulation wheel.

    different main features:

    modifications:

  • polarity protection diode added, power supply jack polarity corrected.

    • notes:

    It is claimed that the HZ-600 came out a little earlier in 1987 than the actual HT-series. It may be that Casio thought that releasing the SD sound engine first in form of a "serious" synthesizer (visually resembling Roland Alpha Juno) would improve its reputation. Unlike my expectation the PCB reveals that it was indeed an independent development and not just a finished HT-700 PCB with omitted percussion ICs. But various strange hardware details tell a different story. Despite the lack of speakers it already contains the big power amp IC LA4127 (without heatsink) and employs 6 heavy and quite oversized D-size batteries (perhaps for memory protection in the age of weak zinc-carbon cells) to power them. And the big power switch at the corner looks like when mains operation with internal power supply was planned but not finished (or should it just add some professional haptic to the somewhat cheapish plastic thing?!). The situation resembles a bit the Casiotone 401 with its mystery prequel 301. Examining the 32KB ROM contents brings some light into it. While my KS 49 (HT-700 variant) uses most of it, in HZ-600 the 2nd half is empty (but did not use a smaller ROM). This makes me conclude that likely the complex accompaniment software was not finished yet when Casio was urged to quickly release the "new" SD sound engine (that was threatened to be considered even more outdated if they had waited any longer) and so made it into a "pure" synth. Even stranger in this context appears that in my KS 49 the oldest embossed case date dots even reveal november 1986, but this likely means only that injection mould parts for the plastic bottom were reused from an earlier model.
     

    hardware details

    The Casio HZ-600 is the most basic synth of the HT-3000 hardware family. The complex multi-chip hardware is built around the CPU NEC D78C10G that controls the SD sound IC NEC D935G but there is neither chord nor percussion, but only keysplit. The melody voice of upper and lower keys is routed from the sound IC through each a VCF JRC 2090D, followed by the stereo chorus. The sound ic also controls the LCD.
    By lack of accompaniment and sequencer, half of the 32KB ROM (I dumped it) is empty, which suggests that this was an early incomplete prototype software version and not just a crippled cheaper variant derived from the finished HT-3000.
     
    In my specimen a 2200uF 16V electrolytic capacitor was bulged and someone had cut and shortened the battery box cable, so I had to replace them.

    keyboard matrix

    The matrix has the same layout like HT-3000, but lacks all rhythm and accompaniment related features. They are unlikely to exist as eastereggs because the ROM contents is much smaller. Also some preset sounds differ.

    I haven't analyzed this matrix by myself. This is based on the Casio HZ-600 service manual.
     
    KC1
    KC2
    KC3
    KC4
    KC5
    KC6
    KC7
    KC8
    KC9
    KC10
    KC11
     
    pin
    out 1
    out 2
    out 3
    out 4
    out 5
    out 6
    out 7
    out 8
    out 9
    out 10
    out 11
    out / in
     
    o
    C1
    o
    F#1
    o
    C2
    o
    F#2
    o
    C3
    o
    F#3
    o
    C4
    o
    F#4
    o
    C5
    o
    F#5
    o
    C6
    in 1
    KI1
    o
    C#1
    o
    G1
    o
    C#2
    o
    G2
    o
    C#3
    o
    G3
    o
    C#4
    o
    G4
    o
    C#5
    o
    G5
     
    in 2
    KI2
    o
    D1
    o
    G#1
    o
    D2
    o
    G#2
    o
    D3
    o
    G#3
    o
    D4
    o
    G#4
    o
    D5
    o
    G#5
     
    in 3
    KI3
    o
    D#1
    o
    A1
    o
    D#2
    o
    A2
    o
    D#3
    o
    A3
    o
    D#4
    o
    A4
    o
    D#5
    o
    A5
     
    in 4
    KI4
    o
    E1
    o
    A#1
    o
    E2
    o
    A#2
    o
    E3
    o
    A#3
    o
    E4
    o
    A#4
    o
    E5
    o
    A#5
     
    in 5
    KI5
    o
    F1
    o
    B1
    o
    F2
    o
    B2
    o
    F3
    o
    B3
    o
    F4
    o
    B4
    o
    F5
    o
    B5
     
    in 6
    KI6
    O.
    piano
    O.
    synth. bells
    OL.
    brass ens.
     
     
    O.
    preset
    OL.
    preset
     
     
    		    
    in 7
    SI1
    O.
    harpsichord
    O.
    magical wind
    OL.
    jazz organ
     
     
    O.
    internal
    OL.
    internal
     
     
    		    
    in 8
    SI2
    O.
    jazz organ
    O.
    blues harmonica
    OL.
    elec. piano
     
     
    O.
    card
    OL.
    card
    C.
    normal
     
    		    
    in 9
    SI3
    O.
    brass ens.
    O.
    light harp
    OL.
    clarinet
     
     
    mode
     
    C.
    split
     
    		    
    in 10
    SI4
    O.
    symphony
    O.
    plunk extend
    OL.
    synth. ens.
     
     
    data/ entry 
     
     
     
    		    
    in 11
    SI5
     
    O.
    select
    OL.
    select
     
     
    sustain
     
     
     
    		    
    in 12
    SI6

    The input lines are active-low, i.e. react on GND. Any functions can be triggered by a non- locking switch in series to a diode from one "in" to one "out" pin.
     

    legend:
    "o"
    		 = keyboard key
    underlined
    		 = function needs locking switch (i.e. stays active only so long the switch is closed)
    O.
    		 = preset sound upper keys ('orchestra')
    OL.
    		 = preset sound lower keys
    C.
    		 = chord (only keysplit)
    S.
    		 = sequencer
    orange
    background     		 = easteregg (unconnected feature)
    grey
    background    		 = unconnected doublet

    The flagship of the HT-series was the Casio HT-6000 (61 fullsize keys, aka Hohner KS 610/TR as a light grey variant), which had a more versatile synthesis engine (64 waveforms, 4 detuneable suboscillators per sound, different velocity curves, ring modulator, separate VCA per channel etc.), which is a quite serious synth and does not sound squarewave-like.
     

     removal of these screws voids warranty...    
    WarrantyVoid
    back to tablehooters collection
    		  
     
    back