The Matrix-1000 was built in a time when there were no USB sticks and flash cards – the technology for storing small amounts permanently in electronics was a battery that kept the memory powered when the unit was switched off.
In the Matrix-1000, this is done with a very common type of battery – a CR2032 3-volt lithium battery, that has been working fine for more than two decades now but is destined to fail at last.
Unfortunately, one of the cost-cutting measures in the budget Matrix synth was to solder it in directly to save on the battery holder – so the replacement of the battery is a minor technical operation.
What you need to do it
A CR2032 battery – very common
A battery holder for the CR2032 – type for PCB through-hole soldering with 20mm pin distance (sample)
Soldering equipment for electronics: iron, solder, pliers, a good workplace light
Moderate soldering skills – soldering is no rocket science, but if you have never soldered before, maybe start by practising on an LED and a resistor
tweezers, a mechanical desoldering suction pump, desoldering mesh
A computer, or an iPad, with DIN MIDI to backup the sounds from memory
Replacing the battery is not hard, but there are some dangers involved, for you as well as for the old hardware.
Unplug the synth before opening it – do not electrocute yourself.
Avoid static charges (i.e. do not stroke your cat while repairing the synth) – static discharge can kill chips.
Avoid mechanical strain on the old PCB, and on the cables. Double-check whether you really removed all screws.
When replacing the firmware chip, be careful to insert the screwdriver blade between chip and socket, not underneath the socket. Do not scratch the PCB.
How to do it
Connect the Matrix to your computer and start a MIDI Sysex recorder (i.e. MIDI-OX on a PC or Sysex Librarian on the Mac). Select “Data Dump” on the Matrix, use the plus key to navigate to “2dA” (Dump All), press Enter. Stop recording when the Matrix is done.
Disconnect the synth from mains.
Remove the 9 screws from the lid, three on each side, three on the back. Remove the lid.
Cut the old battery from the board.
Desolder the remainder of the battery pins from the PCB; clean the PCB holes with the suction pump and/or desoldering mesh – this is the tricky part.
Although it might be possible to solder in the battery holder without removing the PCB if you are very skilled, it is probably easier to take the electronics board out:
Remove the 8 screws from the PCB, and the 5 screws holding the regulators to the right of the PCB to the cooling plate. Be careful not to lose the small distance discs from the PCB, and the plastic insulation on the regulator screws. There are also insulation layers between four of the regulators and the metal cooling plate – be careful not to remove those.
Remove the output connector to the back, and the PCB power connector to the right.
Put the battery holder into the PCB. The plus pole (the terminal on top) has to be facing the back of the synth, marked “Batt +” on the PCB.
Flip the board from the housing.
Solder in the battery holder from the back.
Check whether the memory chip – the one in the front left corner of the PCB – gets 3V in between pin 14 and pin 28.
Change the firmware EPROM by inserting a screwdriver blade betweet chip and socket, and wedge out the chip with great care. Avoid inserting the blade between socket and PCB, this might damage the PCB. When inserting the new chip, take care not to bend or fold any of its pins.
Reattach the PCB, and the regulators.
Close the lid.
Reconnect the synth.
Switch it on holding the ENTER key – this does a factory reset of the memory. You might have to redo this a couple of times.
Do a calibration run – navigate to EXT FUNC, select 7, ENTER, 0, ENTER. The display wil show CAL, the calibration takes about a minute.
The video of the process is quite detailed (i.e.: BOORING!), and it misses out the crucial detail of how to actually solder – the trick is that you use the soldering iron to heat up the wires you want to solder, and then melt the solder by touching those wires with it. The beginners’ technique of just holding the solder to the hot iron leads to ugly blobs of solder on your PCB. But if you really intend to watch this, you probably know how to solder anyway. (There is a short version of the video as well. With music!)
Want that ancient battery in your M-1000 replaced but have no clue how to do it? I am looking for one person to give me their Matrix for the operation – to do a tutorial. You won’t have to pay for anything but the shipping cost – but I’d need you to agree to the terms and conditions (see below)
These days, whenever somebody enquires about the improved firmware I am selling, I recommend changing not only the firmware chip but the battery as well. This battery, located in the lower left corner of the PCB next to the volume knob, is responsible for keeping the the Matrix’ memory chip running when the unit is switched off, so that your synth remembers all those meaty vintage sounds in banks 0 and 1. It has been doing that reliably for decades now, and it is a marvel that the battery is still working after all that time.
Unfortunately, replacing this battery is no simple switch operation, as the old battery is soldered in without a battery holder. (Remember, the Matrix-1000 was a budget machine back then.) So a minimum amount of soldering is required – no big deal, and certainly doable for almost everybody, but I would like to do a (video) tutorial, and I would need another Matrix-1000 for that. Maybe yours?
Terms and Conditions
Here is what you would need to agree to:
I am picking ONE from the first five people to contact me about this. My decision is arbitrary.
You would need to bear the shipping cost. I would need to get a battery holder and a battery for you. I might also replace the firmware if needed.
You, the owner, would also bear the risk – loss of memory, shipping damage, or a CEM3396 voice chip that decides to die right now, I won’t do anything about those. I will, of course, make up for any damage caused by recklessness and negligience – so if I drop my soldering iron into a live Matrix-1000, or set it on fire by connecting to the wrong voltage, I will make good for it.
I will only exchange the battery, and the firmware. I cannot do any other modifications or repairs.
I will possibly add a small back plate glued to the cardboard to give the Dino’s output sockets more stability.
If you use the USB bus master to connect to a powered hub, you could use a Lightning cable to charge the iPad while playing. (Well, technically you could also solder aforementioned cable to the +5V supply on the board, but is it really worth doing that?)
I DO like the TB Midi Stuff app, but as of this time of writing, it is almost abandonware. The author promised to do an iOS12/13 update, but I would not blame him if he didn’t.
Modstep is a great sequencer and beatbox with a more than passing resemblance to Live, but like Live, it is no simple tool. And there have been no updates for a while – once again, I do not blame the programmer. Do people realise that it is impossible to survive on writing special-interest apps?
A Minimoog, a Prophet and an Odyssey enter a bar… and it’s only 200 Euros. — I know, if you don’t understand German, there is not much the last post can offer you but that is where I collected the first batch of info on the DinoPark VA PCB synth. In the meantime, controller panels have been programmed with the TB Midi Stuff app for iPad, and it works.
A Novation SL MKII controller keyboard connected via a USB printer cable to the DinoPark board; the Dino is the USB master, and the power supply for the keyboard.
The DinoPark’s USB Slave connector – the Micro USB socket – is used to connect an iPad via the Camera Connection Kit. It runs TB MIDI Stuff; the panels enable me to change the settings, and show me the parameter settings for the actual sound.
It would be possible to use the Novation controller only; this just about works for the Minimoog emulation as a Minimoog has not got that many dials to turn; for the Prophet clone, the 24 dials and faders of the SL are simple not enough; you would need two templates and would have to jump between them. Moreover, the Novation controller does not evaluate the board’s CC messages and is thus unable to show the actual values for a parameter. Nothing new for VA synths.
Dino Park is switched to OMNI mode, receiving on all channels.
Switching between the synth models is by doing a full MIDI Bank Select: CC#0, model #, CC#32, Preset Bank (0 = Presets, 1 = User). BTW: This is wrong in MPA’s MIDI documentation for the Dino Park.
These models are included:
Bank 0: Minimoog model, „Minimax“
Bank 1: Prophet-5 model, „Pro-12“
Bank 2: ARP Odyssey model, „Prodyssey“
More models available for the Creamware technology: Bank 3 – “B4000” organ emulation closely related to Ferrofish’s tech, a VS-inspired synth called Lightwave, the “Vocodizer”, a FM/PD synthesizer called FMAGIA, and a „Drums ’n‘ Bass“ model.
Saving a preset is done by sending CC#111 followed by the preset to save to.
MIDI implementation is documented for the four first models on the DINOPARK-Support page. Some small errors and omissions: The CC#0, CC#32 sequence for model switching is mixed up, Prodyssey model VCA Mod is CC#11 (not CC#51), VCA Gain is missing from doc but can be reached by CC#51.
Rediscovering an old love of mine, for a simple reason: The R3 is an ideal piece of gear to take on a holiday – compact yet with a full-sized 3-octave keyboard, extremely versatile, powerful sound – and really good fun.
The setup: The R3, vocoder mic inserted, my Volca Beats drum machine synched to the R3 via a sparkling blue MIDI cable, and inserted with another audio cable into the R3’s two audio inputs, and optionally, the ModStep sequencer app on my iPad as a phrase recorder, connected via the new Camera Connection Kit and an USB cable. Apart from my headphones, there is even a small autonomous Bluetooth speaker that can be wired to the R3’s output, via another audio cable. As the R3 is a full-blown synth rather a sound toy (also known as “Tischhupe”), and everything mentioned fits effortlessly into a small gig bag, and thus in my overstuffed car, I conclude that the R3 is ideal travel gear, and suitable even for camping holidays.
Yet the R3 is so much more.
Bass and melody tracks in Modstep, you can even override the Volca Beat‘s patterns and control the sounds directly from a Modstep grid
Came as a substitute, stayed as a champion
Many, many years ago, I was looking for a small beginners’ synth to get back into music. I had been looking for a MicroKorg but without the MK’s obvious limitations – four-voice, monotimbral, small keys simply not suited for my clumsy fingers. I realized that there was a MK XL, and then I realized that there was a slightly larger version of the same engine packaged with full-size keys, more FX, the capability to produce two different sounds at once, and a slightly better user interface, called R3. Which was to be had cheap. So I bought a second-hand one on eBay.
It was a good choice for getting back into music. The same qualities that make it a good holiday companion made it a good choice for jamming: It’s portable, it’s playable, and it’s capable. And gradually I started falling in love with the R3. Let me tell you why.
Genesis fans from the Prog-Rock as well as the „Mama“ factions hate it. Well, „Duke“ it‘s not, but it contains some truly unique pieces of music.
For one, I do love underdogs. And the R3 is an underdog synth. You see, the R3 is a bit like Genesis’ “abacab” album. Prog-Rock Fans hate this record for, well, Phil Collins, you know. Pop fans hate it for its handful of admittedly truly horrible songs. Just like Microkorg enthusiasts hate the R3 for sounding cleaner than a first-gen MK or an MS-2000; VA enthusiasts hate it for not sounding like an analog. I love “abacab” for its unique, electronic, Prophet-5-heavy sound, and for showing off what a terrific musician Phil Collins actually is, whatever you think of his songwriting.
Well, the analogy stops here – as I have mentioned the magic word “Prophet-5”, let’s be honest: The R3 is not a good VA to replicate vintage P5 sounds. In case you should be desperate to replicate the sound from Genesis’ “Lonely Man At the Corner”, better known as the sound from PC’s “In The Air Tonight”: There are much, much better VAs for that – please do look at Creamware’s Pro-12. So let me break down what I like – and dislike – about the R3.
The depth of the Radias engine
The R3’s basic layout is a traditional substractive synthesis setup, with two oscillators, two multimode filters, 3 EGs, 2 LFOs. The first oscillator features classical as well as sampled wave forms. The filters are not too bad for digital filters. A small mod matrix complements a couple of hard-wired modulations such as velocity/EG and LFO2/pitch. Some extensions and variations on the classic theme, under Korg’s “MMT” label from the Oasis line.
Some of that stuff is weird – Format wave forms, anybody? – some is really good. I love switching OSC1 to VPM – this is Korg’s variant of Casio’s Phase Distortion synthesis from the 80s, which was Casio’s variant of DX7-like FM synthesis for normal humans without a second brain or PhDs in acoustic physics. A wave shaper, technically a lookup table but very versatile. Filter2 working as a comb filter, i.e. a resonator for Karplus-Strong type string plucking sounds. And up to three independent effects per sound, including another bit crusher as well as a ring modulator.
The best thing of the R3’s mod matrix may be the feature that is not part of it: a sixteen-step modulation sequencer. Apart from sequencing, well, pitch, it can animate parameters like OSC1’s harmonics in VPM mode, which you cannot reach from the mod matrix. It can even modulate the mod matrix’s modulation depths. I would wish that the mod matrix could reach all the targets the mod sequencer can reach, but it is unfortunately much more limited. They fixed that in the Radias – a Radias is two R3s packaged with an Electribe and many, many knobs – but the software update enabling the additional targets never was back-ported to the R3. Underdog indeed.
I do own a Radias, by the way. It understands R3 sounds, and they are among the best for this synth.
Let’s face it: in terms of digital technology, the R3 is old. 8 voices – 4 per split – are not enough. (Still, its even more limited Microkorg brethren are sold to this day.) There are fresher VA alternatives that are just as cheap, like Novation’s Mininova. And there are some things to hate about this small machine.
The plastic housing is cheap. The full-size keybed is playable but nasty. (Take my word for it: the keys don’t feature the usual metal springs but are attached by flexible plastic working as a spring load – ewww!) They tend do turn yellow over the years as well. Really, really nasty. They could also do with an aftertouch, but even if the R3 HAD aftertouch, the limitations of the mod matrix would mean that you could not program it to control vibrato independently of the mod wheel.
In summary: Rational causes for a a love affair
Full-sized, 37-key keyboard
Light and compact
Deep VA engine capable of some unique sounds
While not knobby, it is still easy to program without an editor