Fujitone 3-A
Angeltone DM-380
Karcher F2
Angeltone DM-100
analogue squarewave keyboard with warm C64 sound, accompaniment & programmable analogue rhythm

Fujitone 3-A

This mysterious and extremely rare analogue thing constitutes the 3rd hardware variant beside Letron MC-3 and HBATEC in a strange series of squarewave instruments with the same plastic case. It seems to be a simplified direct successor of the HBATEC and is in technology, behaviour and sound extremely similar with only few differences. Most severe is the missing organ chord mode, thus only the accompaniment can play automatic chord patterns.

different main features:

eastereggs:

notes:

This instrument sounds extremely similar like the HBATEC, but slightly harsher and the strange analogue intermodulation distortions and low tooting hum are almost inaudible here. Also the banjo (here called "mandolin") rings a bit faster and the "accordion"/ "organ" sound uses a dull 1:1 pulse ratio (HBATEC uses the same like with "piano"). Other sound glitches (including the strange tempo fluctuations) are still present.

The automatic accompaniment of this instrument plays in fingered chord mode when 2 chord section keys are pressed the same simple pattern like when only 1 key is pressed; as well the HBATEC as the Letron MC-3 play here a duet pattern instead. Unlike the HBATEC, with rhythm off the "fingered" and "single finger" buttons don't enable an organ chord mode despite their LED is lit. Instead of the "manual bass" button there is an "arpeggio" button; the arpeggio pattern is mixed into the main voice and thus can be also played without rhythm and accompaniment by simply setting these volume faders to zero. The arpeggio uses a slightly different (less cheesy) style than the arpeggio easteregg I found on my HBATEC, and its sound also employs here a slow decay envelope. The rhythm tempo can be set even yet another step higher than the already tremendously fast one of the HBATEC, which turns it fully into a rapidly purring drumroll; especially in combination with arpeggio this provides the famous fast arpeggiator patterns known from C64 homecomputer SID music.
 

circuit bending details

The Fujitone 3-A hardware is based on the Angletone DM-380 CPU "ANGELTONE KZ381" (40 pin DIL, crystal clocked @ 7.7 MHz).

I am not sure if this was a predecessor or successor of the HBATEC, because while the existence of an arpeggio and inferior features normally hints to an older version, this instrument is otherwise based on a strange no-name 40 pin CPU, while the HBATEC had the "KZ 283" by the brand quality manufacturer Zilog, which only 28 pins demanded a complex keyboard matrix scan technique. The yellow box of my Fujitone 3-A shows a photo of an instrument with the same case but no brand label and button names corresponding to my Letron MC-3 (but without its many LEDs), which is a digital and technically more modern instrument that apparently either got finished later or was imitated by mysterious reasons using costlier analogue hardware. Strange is also that Fujitone 3-A is the only mono instrument of this series despite it has a "stereo" label, and the writings are slightly different; e.g. there is nowhere the term "custom drummer" on it. Also the sustain function (i.e. transistorized flipflop) is missing. Most plausible to me appears a cheapened HBATEC successor that was made before the actual MC-3 when a brandless cheap 40 pin CPU (Zilog clone) became available.
Even cheaper keyboards of this hardware class resemble Angeltone DM-200, i.e. they completely lack the analogue percussion and so play only accompaniment. I don't know if you ever will get your hands on the actual Fujitone 3-A, but other variants aren't that rare, those often have omitted buttons too.
 
sound pulse width octave
piano 3:1  
oboe 15:1  
guitar 7:1  
xylophone 7:1  
piccolo 1:1 +2 
mandolin 1:1 +1
organ 1:1  
harpsichord 1:1 +1
     
bass 1:1  
chord 1:1 +1
arpeggio 1:1 +1
Some preset sounds play 1 or 2 octaves higher, but the waveforms only differ in the pulse width ratio of the squarewave. (The ratios were estimated on an analogue oscilloscope at the headphone jack of Angeltone DM-100, which isn't very accurate.)

keyboard matrix

 17
18
19
20
36
37
38
5
12
30
10
 
CPU pin
in 1
in 2
in 3
in 4
in 5
in 6
in 7
in 8
in 9
in 10
out P
in / out
 
o
o
o
o
o
o
o
o
o
o
L.
/tempo
out 1
13
o
o
o
o
o
o
o
o
o
o
L.
/demo
out 2
14
o
o
o
o
o
o
o
o
o
o
L.
/program
out 3
15
o
o
o
o
o
o
o
o
o
o
L.
/chord
out 4
16
o
o
o
o
o
o
o
o
o
C.
Y-C
 
out 5
29
C.
arpeggio 1
C.
arpeggio 2
C.
single
C.
finger
C.
off
tempo
-
tempo
+
R.
rhumba
R.
march
R.
pops
T.
hihat open
out 6
31
R.
bossanova
R.
tango
R.
disco
R.
16 beat
R.
samba
R.
swing
R.
rock'n'roll
R.
salsa
R.
slow rock
R.
heavy metal
T.
base
out 7
32 
R.
country
R.
big band
R.
waltz
R.
bank
R.
key drumkit
R.
sync
R.
start
R.
stop
R.
start/stop
O.
piano
T.
snare
out 8
33
O.
oboe
O.
guitar
O.
xylophone
O.
piccolo
O.
mandolin
O.
organ
O.
harpsichord
R.
program
R.
play/space
demo
1
T.
hihat closed
out 9
34
demo
2
demo
3
P.
base
P.
snare
P.
hihat closed
P.
conga
P.
hihat open
R.
program
R.
play/space
 
T.
conga
out 10
35

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 an "in" to an "out" pin. The matrix was sorted by ascending key notes.
 

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 ('orchestra')
C.
= chord
L.
= led out
T.
= percussion trigger out
orange
background 
= easteregg (unconnected feature)

eastereggs:

  • 5 lower note keys addable
  • 8 additional rhythms & bank switch button
    There are 8 unused preset rhythms buried in the matrix. Instead of individual OBS buttons you can add the rhythm bank switch button to select these through the normal rhythm buttons.
  • 2 different arpeggios
    Only one arpeggio is wired to the button.
  • rhythm on/off
    uses a single button instead of 2.
  • keyboard drumkit mode
    this turns the rightmost 5 white keys into drumpads (likely intended for case variants without drumpads).
  • Y-C switch
    a locking switch here changes the single finger chord playing method from Yamaha to Casio (see here).
Like with HBATEC, there are LEDs and percussion trigger outs hidden in the matrix. The pulses can be seen on a diode from CPU pin KP to the keyboard matrix outs. But additionally they are also multiplexed from KO6 to KO10. 

LED from any other pin among this group flashes with the percussion:

to KO6 = hihat open
to KO7 = base
to KO8 = snare
to KO9 = hihat closed
to KO10 = conga

Keyboards without analogue percussion circuit play only accompaniment; they do this even (in fixed key) with accompaniment turned "off". I haven't figured out how the CPU detects the lack of percussion hardware to start accompaniment always. It is no fixed matrix diode. Possibly it finds the envelope capacitors in the matrix by timing differences (similar like a C64 paddle port).

Due to similar analogue circuits, there are certainly similar mods possible like with my HBATEC keyboard.
 

pinout KZ381 & KZ281

The "ANGELTONE KZ381" (40 pin DIL) is a squarewave home keyboard CPU with 2 note polyphonic main voice and accompaniment. CPU "ANGELTONE KZ281" (found in Angletone DM-100) seems identical. It was apparently a direct successor of the KZ283 in the HBATEC keyboard, but has 40 pins to support a conventional keyboard matrix. The 2 main voice channels have analogue capacitor envelopes controlled through each a short envelope trigger pulse /ME and an inverted analogue envelope shape signal /MS. Unlike KZ283,  there is no manual chord mode and arpeggio is output through the 2nd main voice channel (which stays mute with accompaniment). Pin 9 outputs something that looks like the clock divided by 2. Pin 10 combined with the keyboard matrix outs can control LEDs and 5 percussion triggers. Pin 28 stays hi when accompaniment is on. I found no difference between KZ381 and KZ281. Possibly the 281 is a 381 with a software bug (custom drummer mode locked up sometimes during my tests?) and so was used for percussionless versions.

Although there is no Zilog brand anywhere, conclusions from predecessors and pinout comparison with the "Zilog - Z8® Family Design Handbook" of June 1988 strongly hint that it is one of the following:

Z8601 MCU (2KB ROM, 128 byte RAM, 12.5MHz, NMOS)
Z8611 MCU (4KB ROM, dto.)
Z86C11 MCU (4KB ROM, 256 byte RAM, 12 or 16MHz, CMOS)

The analogue envelope shape is likely result of external capacitors, since this MCU only supports digital signals. But despite +5V supply, I measured only about +3V hi level at various pins. This pinout is based on my own hardware observation combined with above MCU. Line names were chosen by me to describe their functions (partly inspired by Casio naming conventions).
 
pin name line purpose
1 +5V   supply voltage +5V
2 XTAL2   crystal out
3 XTAL1   crystal in / clock in (7.7MHz)
4 P3.7 /MS1 melody channel 1 envelope shape
5 P3.0 /KI8 key matrix in
6 /RESET   reset
7 R//W   read/write out (NC)
8 /DS   data strobe out (NC)
9 /AS   address strobe out (NC, half clock frequency?)
10 P3.5 /KP matrix percussion & led out
11 GND   ground 0V
12 P3.2 /KI9 key matrix in
13 P0.0 /KO1 key matrix out
14 P0.1 /KO2 key matrix out
15 P0.2 /KO3 key matrix out
16 P0.3 /KO4 key matrix out
17 P0.4 /KI1 key matrix in
18 P0.5 /KI2 key matrix in
19 P0.6 /KI3 key matrix in
20 P0.7 /KI4 key matrix in
21 P1.0 /ME2 melody channel 2 & arpeggio envelope pulse
22 P1.1 M2 melody channel 2 & arpeggio tone out
23 P1.2 /ME1 melody channel 1 envelope pulse
24 P1.3 /MS1 melody channel 2 envelope shape
25 P1.4 C1 chord tone 1 out
26 P1.5 C2 chord tone 2 out
27 P1.6 B bass tone out
28 P1.7 AON accompaniment on out
29 P3.4 /KO5 key matrix out
30 P3.3 /KI10 key matrix in
31 P2.0 /KO6 key matrix out
32 P2.1 /KO7 key matrix out
33 P2.2 /KO8 key matrix out
34 P2.3 /KO9 key matrix out
35 P2.4 /KO10 key matrix out
36 P2.5 /KI5 key matrix in
37 P2.6 /KI6 key matrix in
38 P2.7 /KI7 key matrix in
39 P3.1   (not used)
40 P3.6 M1 melody channel 1 tone out

The pinout is very different from the quite similar KZ282 (Z86C40 MCU) used in toy keyboards. Perhaps Angeltone changed the underlying CPU type because the latter had fixed input-only and output-only ports those restricted PCB layout design.

 
The Fujitone 3-A has brown instead of grey OBS sound and rhythm preset buttons, and it has to the right 3 separate green demo buttons {"I", "II", "III"} (with an additional LED) where the others have the green custom drummer {"program", "play/ space"} buttons and a single grey demo button. An old eBay photo of a "Fujitone M/3A" shows the same colours like my 3-A, but with also that single grey button.

attention: On eBay I saw later a version with type label "Fujitone 3A" (note the missing hyphen) that case (and button writing) looked very much like the photo on my Fujitone 3-A box (but with printed model name on the case), thus this seems to be yet another different instrument that likely contains just plain ordinary MC-3 hardware like the Fujitone 3. The same MC-3 thing was also released as Angeltone DM-280 (in a black nylon bag with red "ANGELTONE" logo, seen on eBay). Few DM-280 even contain actual Fujitone 3-A hardware (by button names - thanks Traktor for info); like the real Fujitone 3-A they seems to be extremely rare.

Likely a fullsize variant of Elta KE-3 in Fujitone 3-A case without drumpads and 'rec' & 'play' instead of custom drummer 'play/space' & 'program'  buttons was released by Hua Li (only 2 Chinese characters) as DM-250. (Box shows brand "discover MTC" and "MusiTech Collection" with an MC-3 case (dummy?) MusiTech MT-4000, seen on internet.) I expect this to have a KZ282 CPU due to the sequencer buttons. Due to omitted drumpads it likely lacks percussion hardware.

Angeltone DM-380

Angeltone DM-380 - electronic keyboard

This seems to be the original brand instrument of the Fujitone 3-A hardware class (see there), because case and CPU are both labelled Angeltone, and in comparison to the Fujitone 3-A it has many additional features those make almost full use of its keyboard matrix. The case of the DM-380 looks more elegant and moderner than the 3-A, and its smooth, bulged plastic drumpads are wonderful responsive. (It has some style elements common with the Hing Hon EK-001.) Unfortunately the volume controls work lousy and should be upgraded.

The Angeltone DM-380 was one of the last genuine analogue keyboards not specially designed for analogue synth enthusiasts. On German eBay around 2005 it was very common (although often fraudulently advertised with an absurdly high 'original retail price'), but nowadays they are all fairly rare, although not expensive. Thus if you don't own a variant of this hardware class yet, do not miss the occasion to buy one cheaply, because this tablehooter has programmable rhythm with fully analogue percussion (much like HBATEC) that doesn't sound bad. They sound somewhat brighter than with HBATEC/ Fujitone 3-A and IMO even a bit more natural. Annoying is only the strong static key matrix buzz that can not be turned off.

different main features:

notes:

Like my TrendLine MC 3700, this instrument has no master volume control (despite it is labelled so) but only 2 separate "faders" for main voice and rhythm volumes, those behave totally independent (i.e. reducing the "main volume" does not also reduce the rhythm volume). There is no separate chord volume control but chord and rhythm share one slider and the rhythm knocks unpleasantly loud in comparison to the rest and it is generally difficult to set the volume very low. (This can be certainly modified.) But you can muffle the analogue percussion by reducing the supply voltage at the AC adapter. The arpeggio is mixed into the main voice. In the demo tunes the 2nd voice (obligato?) volume is a bit too low.

On the main PCB there are 3 trimmer pots for 2 tom and the cymbal/ snare decay envelope of the analogue percussion, but only the latter can be accessed from the back. The PCB is a really lousy mess of bad solder joints, bridges, patched wires and possibly recycled junk components that looks like assembled somewhere in a Chinese concentration camp.

warning: The ribbon cables at the PCBs are brittle and crack off very easy when bent at their solder joints, thus do not attempt to remove the PCBs unless you have prior hotglued all ribbon cable joints firmly to their PCBs. I broke mine when I attempted to examine the PCB component side, and it took many hours to re-solder these awful things. I soldered the ribbon cables now to component wires on the solder side of the PCB, where damages can be seen and repaired much easier. Also the tiny Y-C switch tends to catch and break apart when pulling the PCB out (took hours to fix) and the rectangular LED is a bit delicate to put back into its hole.

A very bizarre feature is that this instrument has a tiny slide switch "Y-C" at the back (that very easily falls apart when accidentally bent during PCB removal) to switch the single finger accompaniment between 2 different sets of chord patterns (i.e. the same key combinations select different chords). I likely would never have figured this out as a keyboard matrix easteregg without manual. Likely the "Y" and "C" genuinely stand for "Yamaha" and "Casio" single finger chord system, but the manufacturer avoided to mention these to prevent legal brand name or copyright problems. (But I doubt that the key presses to select chords are anything still copyrightable or patentable, and a decent formulation in the manual like "inspired by [brand]" or "like used by [brand] keyboard players" would be certainly sufficient to avoid legal trouble. E.g. the shortcut schemes of word processors are also switchable to emulate others without legal problems, since everything else would result in undemandable monopolies when all secretaries or musicians in the world would be forced to re-learn typing each time before they can use a different brand product.)

Because it is so unobvious, here is a (slightly cleaned up) quote from the manual what the Y-C switch does:
 

Single automatic bass chord playing method:

Press single automatic bass chord button (S. FINGER) it enters into the single automatic bass chord accompany state, Y single finger automatic bass chord playing method or C single finger automatic bass chord playing method, it can be selected by moving the C-Y switch.

Y system of single finger automatic chord playing method: [this would be Yamaha]
 
Major chord: When you press any one key on the A. B. C. Keyboard [i.e. in the accompaniment section], you will get its Major chord.
Minor chord: When you press the key and any black key on its left side at the same time, you will get its Minor chord.
Seventh chord: When you press the key and any one white key on its left side at the same time, you will get its seventh chord.
Minor seventh chord: When you press the key and any one white key and any one black key on its left side at the same time, you will get its Minor seventh chord.

C system of single finger automatic chord playing method: [this would be Casio]
 
Major chord: When you press any one key on the A. B. C. Keyboard, you will get its Major chord.
Minor chord: When you press the key and any one white key on its right side at the same time, you will get its Minor chord.
Seventh chord: When you press the key and two white keys on its right side at the same time, you will get its Seventh chord.
Minor seventh chord: When you press the key and three white keys on its right side at the same time, you will get its Minor seventh chord.

There are also 2 different OBS arpeggios on this instrument and 16 OBS rhythms while the Fujitone 3-A has only 8 (rest as eastereggs). Most rhythms are like HBATEC; only the "slow rock" is very different, and the "16 beat" here has an annoying snare staccato.

Apparently a predecessor of the DM-380 was the Angeltone DM-450 (aka Fujitone 4-B, seen on eBay), which likely contained the same hardware (same OBS rhythm and sound names), but had a less stylish case with different button layout (2 grey rows for sound & rhythm presets) and blue drumpads (shaped like with Casio SK-8) at the upper right control panel rim.

Karcher F2

This bizarre cheap variant is almost identical with the Angeltone DM-380 but is missing the analogue drum section on its PCB (solder holes were left empty) and thus plays no percussion instruments at all despite it has the same KZ381 CPU (here an "ANGELTONE KZ381, Z 90 39" like the Fujitone 3-A). But the accompaniments of the 16 "rhythms" are fully functional (including "Y-C" switch), and when neither single finger nor fingered chord is pressed, they behave like Angeltone DM-200 (which has a different CPU KZ282).

The case of the Karcher F2 looks like the DM-380 except that it has no drumpads and thus no custom drummer section. Instead of the "program" and "play/ space" buttons it has 2 additional demo buttons (5 in total) although the buttons 4 and 5 are fake doublets and do the same like button 1 and 2. Next to the buttons are even 2 embossed fake LED shapes where the DM-380 has genuine LEDs.
Theoretically it would be likely possible to upgrade the F2 PCB with real analogue percussion (solder holes are still there), but it would be a lot of work and I haven't analyzed this further.

Angeltone DM-100

This crippled squarewave tablehooter is very similar like Karcher F2. Despite Angeltone DM-380 CPU it lacks analogue percussion, so what appears to be a row of 7 drumpads, only selects the 3 demos (multiple pads wired together). Another flaw is that despite the key matrix supports it, instead of "rock'n'roll" it selects a doublet of "swing" (haven't checked if this is a wiring bug or short circuit). The timbres sound a little rougher and less filtered than in Fujitone 3-A, which stresses its rough C64 appeal.

The DM-100 seems to be extremely rare. (I never saw another.) The wedge shaped case resembles Bestar MC3300.
Unlike Karcher F2, the PCB is much smaller and was never designed to have analogue percussion. The keys PCB has the manufacturer name "ANDA ELECTRONIC LTD.", which may be the same company like Xin Anda - creator of the absurd 8-Melody Letter Study Piano. Very strange is that the CPU type is "ANGELTONE KZ 281, 2163, 2 9127", not KZ 381 despite identical pinout and behaviour. Possibly it had a software bug (custom drummer mode locks up sometimes?) and so was used for a percussionless cheapo version. "2163" may allude to Yamaha YM2163 although none was installed.

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