BONTEMPI synth BS 2010
KE 3860
BS 3000
BS 2000		 8 bit waveform keyboard with simple lo-fi synth

Bontempi BS 2010

This 2 note polyphonic synth tablehooter of 1995 (PCB date code "25-95") was one of the successors of Bontempi B50. All non- percussion sounds are made from combinations of 2 static waveforms with volume and pitch envelopes. Most spectacular are the stereo crossfade and howl effects.

The sound of the internal speakers is quite tinny with coarse 8 bit waveform resolution, but fortunately at least the cruel beeping continuous tone of the B50 is gone. Most interesting is that the thing has a so-called synthesizer, which in fact only replaces the pitch- and/or volume crossfade envelope of a sound with 1 of 5 different preset ones. Also certain combinations of 2 preset sounds can be layered in a way that all together 2000 sounds (with hack even 3000 sounds) can be generated. While these simple combinations of presets not really deserve the term "synthesizer" (you can not tweak intermediate values), the instrument can make some funny, howling sounds and strange stereo crossfade and panning effects those are quite unusual and not often heard. Fans of Rob Hubbard's C64 SID compositions may easily fall in love with it.

Synth Section

The keys are very (almost too) responsive, so you can nicely glide on them. Unfortunately a design flaw often makes both polyphony channels cancel out each other during fast play, which limits usability. So it should be rather considered a monophonic instrument - not least because also some synthesizer settings or the use of rhythm forces monophonic play. But for lead synth voices or pads it is otherwise nice because it has certain unusual grainy special sounds those don't exist on average modern standard instruments. The fairly soft and thin digital waveforms have a clean drawbar organ appeal resembling Casiotone 1000P, although the coarse resolution and stereo crossfade/ howl effects make it unique.

The rhythms unfortunately all include a fixed key accompaniment pattern, which makes them quite useless for melody play.

main features:

(old picture from eBay)

eastereggs:

modifications:

notes:

The accompaniment repeats always the same pattern loop in a fixed key and thus is barely usable for playing given musics (even the built-in demo songs partly plays in a wrong key!) - only for improvising it may be of some use. Due to it is impossible to play rhythms without accompaniment, and accompaniment always limits the keyboard to monophonic play, there is only little serious use for the rhythm/ accompaniment feature of this tablehooter. It is hard to understand why Bontempi didn't add single finger chord or at least a mute button for that pesky accompaniment track to play rhythm-only.  Also the concept to prevent certain layered sound combinations in cheaper variants (BS 3000 was the real thing) sounds like a nasty ripoff to me.

All preset sounds, rhythms and demos are selected by a 4 digit number entered with cipher buttons {"1", "2" ,"3", "4"}. Entering a number sounds a snare drum, which disturbs live performance. Despite there are 56 empty fields in the "sounds bank" matrix table, entering other sound numbers than the official ones mostly do nothing. The only working easteregg number is 2211 'trumpet + el. piano' (found on KE 3860 list). The battery alarm plays a clang (resembles  'marimba+carillon' on note C4, but slightly thinner) in a preset sound that is not on the control panel of BS 2010 (nor KE 3860) and keeps that sound selected. (It has an electronic power switch and not much else to keep in memory - why didn't add Bontempi a proper auto-power-off?!)

The preset sounds partly include envelopes those are not selectable by the synth buttons. E.g. the 'telephone' sound uses a fast pitch switching envelope (squarewave vibrato) which can be replaced by pressing the 'frequency' buttons (i.e. the digital phone ring sound character disappears), but you can not switch back to the original envelope this way, because that special envelope is not available by these buttons. (Re-selecting the preset sound always gets back its default envelope.) The behaviour of the synth section is thus sometimes a bit confusing, particularly because after 5 presses they cycle back to the 1st setting, which is not necessarily the default for a sound, and there is no display to see which setting is currently selected.

All basic preset sounds are made from 2 layered static waveforms (resembling drawbar timbres) and thus not too realistic. High notes have noticeable DAC aliasing noise. When not mentioned otherwiese, they all have a mild 5hz vibrato (aparrently the default 'frequency' envelope). 'el. piano' sounds like a simple dull FM piano; 'piano' is the same in brighter and 'guitar' in even brighter. 'jazz organ' resembles a metal pipe organ rank (no clicking attack) with somewhat sonorous bass range. 'brass' is similar in even more sonorous with some phasing (LFO speed 8s); 'trumpet' is the same in slightly duller. The 'clarinet' sounds more like a dull harmonium. The 'accordion' with its sonorous bass range (and 8s phasing) sounds quite realistic; 'oboe' is the same in thinner. 'carrilon' has 4Hz tremolo and resembles tubular bells. 'flute' resembles a metal flute (with 8s phasing), but the reedy bass range rather a harmonium. The 'marimba' (ignores key press duration) has a picked string attack and thus rather resembles a harp or koto. The 'mandolin' rings at about 5hz, using the 'guitar' waveform (8s phasing); editing synth 'frequency' removes the ringing. The 'haw guitar' is 'guitar' with a grainy pitchbend up followed by strong 5hz vibrato. 'ghost' resembles a thin e-piano (1 octave higher) with strong siren-like 2hz vibrato (sets 'frequency' to supervibrato). 'telephone' is a digital telephone ring made from a flute-like waveform (1 octave higher) with rapidly purring strong square vibrato (depth -1 semitone?); editing synth 'frequency' removes the special LFO mode and so turns it into a piccolo flute.

The panel sound list has some bugs causing to doublets and missing combinations. While the rows normally count the numbers diagonally to the right down, the number 1333 is followed by 1134 instead of 1334 for 'haw guitar+marimba' which is the basic 'marimba' (same on KE 3860). The right end of the diagonal strip partly continues in the 2nd strip below the gap, but many combinations were (intentionally?) omitted. The "2000 sounds" instruments support no high sound numbers with 22xx, 23xx and 24xx (entering them does nothing). The only exception 2211 ('trumpet + el. piano') is not listed on BS 2010 panel.

The synth section buttons cycle each through 5 different modulations (in both groups the left one backward, the other forward) and so transform each preset sound into 25 variants. Selecting a new preset sound (basic or combination) always returns the synth mode to default. I found their official mode names (without explanation) in the BS 2000/ BS 3000 user manual. This is what they do.

fade

Each sound consists of 2 subvoices output through the stereo channels. The right channel seems to play the higher component, the left channel the 1 octave lower one. The fade parameter changes the volume envelope they crossfade with.
 
  name: function:
0 normal Both subvoices sound simultaneously.
1 echo Begins like 0, but after key release timbre toggles with 2hz between both subvoices (i.e. 2hz square DCO envelope at 100% depth on right channels and inverted on left channel) during the added 3s sustain. 
2 sweep Crossfades within 1s from the right (high) to the left (low) subvoice and stays there.
3 pan pot Continuously crossfades with 2hz between the right (high) and left (low) subvoice (i.e. 2hz triangular DCO envelope at 100% depth on right channels and inverted on left channel).
4 rebound like 3, but much slower 0.16Hz (6s) crossfade.

frequency

This controls the pitch LFO. In preset sounds those already contain special modulations (telephone vibrato or mandolin ring) the original LFO setting is removed when edited and will not return to default at 'normal'.
 
  name: function:
0 normal mild 5hz vibrato.
1 portamento This sets the keyboard monophonic and makes the pitch quickly (within 0.1s?) glide to the next pressed key.
2 glide This makes the attack phase pitch of each note rapidly (0.1s?) ascend to the played note (kind of bagpipe imitation).
3 hawaii Like 2 with slower pitch ascend (0.3s?) followed by additional square 5Hz vibrato.
4 supervibrato Strong (about +1 semitone) 2Hz vibrato.
- mandolin ring Not selectable. Retriggers the whole envelope with 5Hz (only used by 'mandoline').
- telephone ring Not selectable. Rapidly purring strong square vibrato (depth -1 semitone?, only used by 'telephone').

I really don't understand why Bontempi didn't include this simple synth into their later System5 instruments. With real accompaniment and flawless polyphony it could have been much more fun to play, and their versatile "666 sounds" synthesis engine likely supports such internal parameters anyway.
 

circuit bending details

The Bontempi BS2010 is based on the single chip CPU "Texas Instruments A45016PH, ... , COMUS 2743341" (crystal clocked at 32MHz); the 2nd row "WX 54AETRW" seems to be only a serial number; the behaviour is identical with the CPU of Bontempi KE 3860, BS 2000 and BS 3000. So this hardware analysis is valid for them as well. 

Both stereo channels are output by the CPU through a single 8 bit resistor array DAC ("8BIT 103G") and get demultiplexed by a strange transistor circuit (sample & hold?, controlled by CPU pins 6 and 7). This adds some harsh overtones and HF residues to the audio signal. To smoothen it, you may upgrade it with something like the glitch-free Casio DAC circuit of Robin Whittle.

wiring differences

In BS 2010 there is capacitor C5 connecting pin 51 to GND. Pin 52 seems unused. Pin 55 is on GND. (Pins 56+57 are in all versions on GND.) In KE 3860 there are small PCB layout changes. Under capacitor C5 (pin 51 to GND) is a solder blob jumper connecting pins 52+55. An unused pad can instead connect 52 with GND. In BS 2000 there is a cut trace between pin 51 and empty solder holes for a resistor R4 to +Vs (pin 52), but there is still a (zener) diode D1 from pin 51 to 52 which is on +Vs. Instead of C2 there is a 10k resistor from pin 51 to GND. Pin 55 is connected through a solder blob jumper with GND. In BS 3000 there is even a cable running from the power switch to a disconnected cut trace(!) where R4 was omitted. It has (on solder side) the diode D1 from pin 51 to +Vs (pin 52 and through a solder blob 55), and instead of C2 a 10k resistor from pin 51 to GND. Connecting a resistor R4 turns the instrument on, so the cable to the power switch was possibly a leftover from a factory test or repair.

Some changes were likely result of different power switch wiring and desperate attempt of fixing a design flaw. That is to say, the BS 3000 case was apparently quickly re-designed in the last moment from a power-on button (like in B50) to a locking slide switch (holding down the same rubber contact driving a transistor) when Bontempi noticed too late that the new CPU has no auto-power-off and the case layout lacks a power-off button. The later BS 2010 and KE 3860 thus got equipped with electronic power-on and additional power-off button. Small Bontempi keyboards are generally notorious of making nonsense like crashing (shitshot by incomplete reset), buzzing (resetting in a loop) or not stay off at all (happens in my Arturo) when power buttons are pressed with batteries empty or conductivity properties have changed by ageing or battery leak corrosion. 

Pin 51 is reset. While BS 2010 and KE 3860 have here a capacitor (C5) to GND, both BS 2000 and BS 3000 have a 3.6V zener diode (BZX 3V6) to +Vs and 10k resistor to GND. Pin 52 in all models seems to be +Vs. 
 
model sounds on panel pin 55
BS 2010 65 GND
BS 2000 80 jumper to GND
KE 3860 105 jumper to +Vs
BS 3000 121 jumper to +Vs

3000 sounds mode

Manufacturing 4 different CPUs with only fewer sound combinations enabled makes commercially no sense. So it could be expected that wiring changes somehow setup the count of available preset sounds during power-on. First there seem to be indeed only 2 versions of behaviour ("2000 sounds" vs. "3000 sounds") because all sounds of the BS 2000 panel list do indeed also work on BS 2010, and those of BS 3000 also on KE 3860. When I found nothing in the keyboard matrix and the suspicious cable to a cut trace in my BS 3000 I did expect something evil. But fortunately there is no nasty stuff like flash rom reprogramming involved.

The entire magic is done by pin 55 that selects the 3000 sounds mode when pulled hi, else 2000 sounds. (When disconnected, it somehow seems to memorize its last setting until reset. It is unknown if this is by stray capacitance or an intentional polling algorithm.) All models except BS 2010 have a solder pad jumper for this. Pin 55 has to be disconnected from GND and rewired to +Vs (e.g. pin 41) to enable sound combination numbers beginning with 22xx..24xx. The highest supported sound number is 2431. (2432..2444 do nothing.)
 
In my BS 2010 pins 55+56+57 are on the same broad GND trace (impossible to figure out without comparison). Thus pin 55 first has to be carefully cut out of that trace (e.g. with small screwdriver or nail scissor, gap here marked black) before you can solder that part (here marked yellow)  with a thin enamelled wire to +Vs. Clean up copper debris thoroughly and verify not to short +Vs and GND before you insert batteries. (An overlooked short circuit would likely destroy power switch transistors Q1 or Q2.)

The jumper inside KE 3860 suggests that a later BS 2010 version was made with the same (easier to modify) mainboard. But it also may be that this was done only to exchange faulty mainboards during warranty repair and switch them to the same count of sounds.

keyboard matrix

When I originally wrote down the keyboard matrix, I badly messed up the CPU pin count and numbering (they count anticlockwise). I later reconstructed their order by PCB photos. Likely other instruments on this page have the same keyboard matrix. So I verified and re-sorted the matrix by comparison with my BS 3000. The instruments itself have no matrix diodes (thus restricts polyphony to 2 notes) and so make matrix mess when panel buttons are held during keyboard play.
 
40
39
38
37
36
35
34
33
 
CPU pin
in 1
in 2
in 3
in 4
in 5
in 6
in 7
in 8
in / out
 
o
C1
o
C#1
o
D1
o
D#1
o
E1
o
F1
o
F#1
o
G1
out 1
23
o
G#1
o
A1
o
A#1
o
B1
o
C2
o
C#2
o
D2
o
D#2
out 2
21
o
E2
o
F2
o
F#2
o
G2
o
G#2
o
A2
o
A#2
o
B2
out 3
19
o
C3
o
C#3
o
D3
o
D#3
o
E3
o
F3
o
F#3
o
G3
out 4
22
o
G#3
o
A3
o
A#3
o
B3
o
C4
o
C#4
o
D4
o
D#4
out 5
11
-
volume
+/-
'1'
'2'
'3'
'4'
S.
fade -
S.
fade +
out 6
15
T.
juke box
T.
plus play
T.
start/stop
S.
frequency -
S.
frequency +
tempo
-
tempo
+
rhythm
start/stop
out 7
8
reset
click 1
click 2
reset
click 3
accomp
ring
reset
rhythm
start/stop 2
out 8
17

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
T.
 = track control
S.
 = synth section
orange
background 		 = easteregg (unconnected feature)
grey
background		 = unconnected doublet

  • alternate button functions
    The CPU keyboard matrix inputs are pin 33 to 40. The outputs for keys are pin 11, 19, 21, 22, 23. (Pin 11 supports 3 additional higher note keys.) The control panel button outputs are pin 8 and 15. At the unused pin 17 a set of 8 additional buttons can be added, those do strange sound changes. I instead added a switch that switches the line to the control panel from pin 8 to pin 17. This changes the function of 8 buttons in the following way:
     
    normal
    juke box
    plus play
    track
    start/ stop
    frequency -
    frequency +
    tempo -
    tempo +
    rhythm
    start/stop
    alternate
    reset
    click 1
    click 2
    reset
    click 3
    accomp. ring
    reset
    rhythm
    start/stop 2
The 'accomp. ring' button exchanges (until reset) the bass of the accompaniment with a bell (carillon?), which sounds particularly nice with the "Jingle Bells" demo tune. After a stopped rhythm it makes its last accompaniment note hang. Also held keys can make a note hang. In 'juke box' demo mode it can turn the next song into sound chaos (a dangling pointer bug?). The 'click' buttons make a click noise (volume depends on the button). When these buttons are pressed together with keys, sometimes strange sound changes appear. The "reset" buttons resets everything to power-on condition (like after pressing "on"). The "rhythm start/stop 2" button behaves like the normal one, except that it can pause and resume the juke box mode; after resume the main voice is missing (like "plus play", which normally does not work here).

I don't know what this stuff shall be - it was likely a test feature. They also don't change the count of preset sounds or other obvious things when held as a fixed diode during power-on.

distortion

The stereo amplifier consists od 2 identical channels which each 3 transistors. The transistor Q10 is the preamp, Q9 + Q10 form the power amplifier of the right channel; Q8 and Q6 + Q7 do the same for the left channel. To add distortion, solder a diode from the base of transistor Q10 to the wiper of a 50 kOhm pot, and the left end of the pot to GND; do the same with the other channel. 

volume control

For real volume control, cut the trace between C14 (100 nF) and the 1µF electrolytic cap. Connect the wiper of a logarithmic 100 kOhm stereo potentiometer with the now open end of C14, and the right end of the pot with the open end of the 1µF cap; connect the left pot end with GND. Connect the same way the 2nd amplifier channel with the 2nd channel of the stereo pot.
 

pinout A45016PH

The single chip CPU "Texas Instruments A45016PH, COMUS2743341" (80 pin SMD pins count anticlockwise, COMUS... = software number of internal ROM) is the CPU of the "Bontempi Synth" cheap mini keyboard series. It is crystal clocked at 32 MHz and apparently has 6 internal sound channels (2 melody voice, 3 chord, 1 bass?) and very low resolution sampled(?) digital percussion. It supports a keyboard matrix of at least 8 in and out pins. Sound is output through an 8 bit resistor ladder DAC, multiplexing both stereo channels.

This (likely incomplete) pinout was made by myself, because I have no service manuals. So all pin names were chosen by me (partly inspired by Casio naming conventions), so do not depend on them - they may change if I find schematics. The table is based on own BS 3000 examination and my old BS 2010 description.
 
pin name purpose
1 DATA1 hf data out (not used)
2   (hi out)
3   (hi out)
4 DATA2 hf data out (not used)
5 +Vs supply voltage (+4.3V)
6 MUXL dac demux left sound channel
7 MUXR dac demux right sound channel
8 KO7 key matrix out
9   (hi)
10   (hi)
11 KO5 key matrix out
12 GND ground 0V 
13 +Vs supply voltage (+4.3V)
14   (hi)
15 KO6 key matrix out
16   (hi)
17 KO8 key matrix out
18   (hi)
19 KO3 key matrix out
20 DATA3 data out (phase modulated 200Hz, not used)
21 KO2 key matrix out
22 KO4 key matrix out
23 KO1 key matrix out
24   (hi out)
25   (hi out)
26   (hi out)
27   (hi out)
28   (hi out)
29   (hi out)
30   (hi out)
31   (hi out)
32 +Vs supply voltage (+4.3V)
33 KI8 key matrix in
34 KI7 key matrix in
35 KI6 key matrix in
36 KI5 key matrix in
37 KI4 key matrix in
38 KI3 key matrix in
39 KI2 key matrix in
40 KI1 key matrix in
pin name purpose
41 +Vs  supply voltage (+4.3V)
42 +Vs supply voltage (+4.3V)
43   (lo)
44   (lo)
45   (lo)
46   (lo)
47   (lo)
48   (lo)
49   (lo)
50   (lo)
51 /RESET reset
52 +Vs supply voltage (+4.3V)
53 GND ground 0V
54 CLK3 hf clock + hi out (not used)
55 SEL select 3000 sounds mode
56 GND ground 0V
57 GND ground 0V
58 O8 dac out (MSB)
59 O7 dac out
60 O6 dac out
61 O5 dac out
62 O4 dac out
63 O3 dac out
64 O2 dac out
65 O1 dac out (LSB)
66   (hi out)
67   (lo out)
68 CLK1 clock out (0.26us = 3.8 MHz?, not used) 
69 CLK2 clock out (0.15us = 6.66 MHz?, not used) 
70 OSC1 crystal in (32 MHz)
71 OSC2 crystal out
72 GNDC1 (crystal center pin, wired to GND)
73 GNDC2 (wired to 72)
74 GND ground 0V
75 GND ground 0V
76 GND ground 0V
77 GND ground 0V
78 GND ground 0V
79 GND ground 0V
80 GND  ground 0V

Some of the many +Vs and GND pins may be in fact hidden configuration inputs (like pin 55) to setup software behaviour. Unused pins marked with "(hi out)" or "(lo out)" have a low resistance output, while normal pins are high resistance (i.e. strongly change level when burdened with 1k pullup resistor against GND or +Vs). The keyboard matrix outputs however are high resistance anyway to avoid damage when many keys or buttons are held down in the diodeless matrix. (Don't ask me why the key matrix out pin order is such a mess. - Casio often did the same.)

The DAC demultiplexing pins 6, 7 output a strange waveform (multipulse squarewave 111010?) to switch the DAC between both stereo channels. It is unknown why no simple squarewave is used (overtones?, HF interferences?).

Pins 1 and 4 output very high frequency data. Pin 20 apparently outputs data through a phase modulated 200Hz squarewave signal (may be IRQ, frequency resembles keyboard matrix). Pin 54 looks like hi, but on my scope there is a pale shade hinting that it is mixed with extremely high (clock?) frequency that my old analogue oscilloscope can not enlarge further. The unused output pins 24..31 become lo when the 'accomp ring' matrix easteregg is on. Also during shitshot unused pins output strange stuff, and pinouts of Bontempi single chip CPUs differ strongly, which hints that it may be a generic Texas Instruments microcontroller with internal ROM. Shitshot (touching crystal pins, keeping reset pin hi) on this CPU usually results in total lockup (neither keys nor panel buttons work). This sometimes producing a grainy digital noise loop (may contain internal ROM code worth to capture and analyse) but nothing playable (typical for small Bontempi sound bank keyboards), which hints that the algorithm strongly depends on RAM contents without much prewired hardware functions. I.e. this is likely a softsynth on a chip (like Casio SA-Series), similar like the (later?) "666 sounds" engine.

The instrument has 12 demo songs with standard accompaniment:

  1. Jingle Bell
  2. Twinkle Twinkle Little Star
  3. La Cucaracha
  4. Can Can
  5. Mozart Op. 40
  6. Silent Night
  7. Cantata
  8. Ode To Joy
  9. The Carnival Of Venice
  10. Romanze In F
  11. Fascination
  12. Patricia
They can be started either in a big loop with the 'juke box' button or individually with the track control 'start stop' button. Only in the latter mode the main voice can be muted with the 'plus play' button to permit manual keyboard play training to their accompaniment track; the user plays the main voice while the instrument selects accompaniment and timbre by itself.

Bontempi - On Tour KE 3860

This instrument of the "Bontempi On Tour" series of 1996 (PCB date code "16-96") was the direct successor of the Bontempi BS 2010. It is almost identical (including the annoying polyphony flaw and beeping demultiplexer noise) except that it has a blue case, 9V DC power supply jack and includes instead of 80 now 121 layered preset sound combinations. So it provides 3000 sounds in total (despite the label "electronic synth" is gone) because its "sounds bank" matrix table contains far less empty entries (those identical with BS 2010 even have same numbers).

Strange is that there are still 15 empty fields (2 diagonal rows) in the matrix; I guess it was no technical limitation but marketing reasons because it wouldn't have resulted in a round thousands number of sounds. The hardware accepts the same 121 sounds like BS 3000.

different main features:

(old picture from eBay)
Differences to BS2010 hardware are small... The pin 55 jumper above the CPU selects 3000 sounds mode.
...although this one has an AC adapter jack. The empty solder holes were likely designed for a discrete voltage regulator. The same PCB in BS 3000 is fully equipped and only lacks capacitor C2.
Sound and behaviour of this instrument (including rhythms and demos) are fully identical with my BS 2010 and BS 3000. The only technical difference to BS 2010 is a jumper for the 3000 sounds mode.

Other Bontempi keyboards with the same case style like KE 3860 were released in blue as Bontempi On Tour KE 3760 (contains B40 hardware) and in orange (different demo) as Confetti Kla4.

 
BONTEMPI synth   BS 3000 & BS 2000

The top of the line model of this hardware family was the mysterious "6 note polyphonic" Bontempi BS3000 mini synth. Unfortunately it seems to be terribly rare, so I expected it to be vapourware of that only some prototypes were made (like with the Commodore C65 home computer). The only hint to its existence was this advertisement on the Bontempi homepage of the year 2000.
BS3000 Synth.

37 Mini Key Digital Keyboard * (C-C) * 120 individual sounds * 24 Rhythms * Synthesiser function to create 3000 sounds * Polyphony: up to 6 notes * 12 Songs with Plus Play function * Controls ; volume & tempo * 15 keys for selection of functions * Pre-set Effects : Sustain, Reverb, Vibrato, Chorus * Stereo Amplification & High Performance Loudspeakers * On/Off switch  * Switch Off Reminder *


Requires 4 x R6/AA 1,5V batteries to operate (not included)  Can be used with AD170 mains adaptor (not included)
It had a similar case like my B50, but was claimed to have 3000 sound combinations, 120 individual sounds, 24 rhythms, 12 demos and up to 6 note polyphony - apparently it was the top of the line model of Bontempi's pre "System 5" mini keyboards. 

However later I indeed found and bought a BS3000 and BS2000 on eBay. Their behaviour is very much like the later KE 3860 and BS 2010, i.e. they are only 2 (not 6) note polyphonic and have the same sounds, rhythms and playability flaw. Even the demos are the same.

Likely the case style with mini keys, rubber buttons and paper printed control panel didn't sell well and so they got soon replaced with a more conventional midsize design.
 

Bontempi BS 3000

This odd shaped instrument was basically a Bontempi KE3860 with small keys. It is one of the rarest and the holy grail of Bontempi mini keyboards, because with 121 layered sound combinations on its control panel it has the most complete sound set of the series.

Like KE3860 it has the nice little synth with many siren-like howling envelopes, and also here the annoying keyboard bug makes both polyphony channels cancel out each other during fast play. Unlike the ad suggests, it is not 6 note polyphonic. Possibly Bontempi counted internal sound channels of the accompaniment or suboscillators as a similar customer misleading scam like their fantasy name "midi keys" for midsize keys.
The control panel lists all supported preset sound combinations with own names, counting without gaps and so making the available number range more obvious. The highest supported sound number is 2431. (2432..2444 indeed do nothing.)
synth section Only the BS3000 has a power supply jack.
An unused cable at power switch ends on cut trace near CPU. Unlike KE 3860 the PSU jack PCB apparently has a voltage regulator circuit.
The CPU of my BS 3000 is "Texas Instruments A45016PH, WX 55A1JNW, COMUS 2743341".

Bontempi BS 2000

This was the cheaper and crippled variant of the Bontempi BS 3000. Its only difference are fewer preset sounds (only 80 listed on control panel) and the missing AC adapter jack. Also this instrument is extremely rare. Basically it is a BS 2010 with small keys.

The continuous control panel numbering scheme is like BS 3000 and so makes it obvious that all preset sounds beginning with 22xx, 23xx or 24xx are missing. But even here the still supported number 2211 ('trumpet + el. piano') is not mentioned.
 
Both the BS 2000 and 3000 came with the same combined manual, which indicates that they were not technical successors of each other, but that BS 3000 was designed as the more expensive variant. The only hardware difference is a jumper setting. The manual claims that BS 3000 has 120 (I count 121) and BS 2000 only 80 sounds.

eastereggs:

hardware details

The CPU of my Bontempi BS 2000 is "Texas Instruments A45016PH, WX 54A6XTW, COMUS 2743341". See BS 2010 for general hardware description. How to open the case without damage see Bontempi B50.
Only the solder jumper in the middle above the CPU (pin 55) makes it have fewer sounds. 

3000 sounds mode

In BS 2000 a solder pad jumper connects pin 55 with GND. Resolder it to +Vs like in BS 3000 (here marked yellow) to enable 3000 sounds. (Do not connect all 3 contacts, else you short supply voltage with GND and likely destroy the power switch transistor Q1.)

If you like the sound style of these instruments, also watch out for Bontempi ET 202. (It has no synth, but 4 note polyphony and nice, arpeggio- like accompaniment.) Another Bontempi Keyboard with nice howling synth sounds is the Bontempi GT 770. Another great digital lo-fi synth with strange howling sounds is Casio SA-35.
 

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