Home > Ableton Live, controller, Modification, Tear Down > Novation Launchpad Tear Down

Novation Launchpad Tear Down

November 21st, 2009 Leave a comment Go to comments

For those of you that are curious like myself, here are some images of me disassembling the Novation Launchpad.   I hope this inspires some of you to do some cool modifications to the controller, and make sure to send us some images of your modified launchpad.  Click on the pictures to zoom in.

1. The launchpad

Launchpad unopen

2.  The launchpad flipped over.  There are only 9 screws, taking this thing apart is a piece of cake.

Launchpad from back

3.  The launchpad without the face plate.

Launchpad without face plate

4.  The Launchpad viewed from the top without button pads.

Launchpad open top view

5. The launchpad opened and viewed from the back.

Launchpad open from back

6.  Here is a close up of the mcu’s and the programing header on the board.  The bigger chip is a STMicroelectronics 72f63bk2m1, here is a link to it’s data sheet.  The smaller of the two chips is the NXP 74hc164d, and here is a link to it’s data sheet.

close up of mcu


  1. December 6th, 2009 at 01:29 | #1

    good article as usual!

  2. Bob Borries
    December 21st, 2009 at 17:14 | #2

    I hate red, green and yellow led combination, what if solder rgb leds instead, only connect 2 of the leds, like red and blue, too get red, blue and purple. With green and blue, you would get Green, Blue and Cyan.

  3. miketron
    December 21st, 2009 at 22:02 | #3

    @Bob Borries
    Agreed, I have been planning on replacing all the buttons with clear arcade buttons, and then replacing the regular leds with rgb.

  4. Ben
    January 1st, 2010 at 19:34 | #4

    @Bob, sounds like a good idea. Novation couldn’t do this as it’s far too costly to put down 80 RGB LEDs on a $199 controller. RG LEDs are far more available.

    @miketron, if you are thinking of wiring your own buttons and LEDs into it, good luck — it wasn’t designed for such hackery! Look at the Arduinome as an alternative, unless you’re really keen on the Ableton integration.

  5. RockMan Rock
    January 13th, 2010 at 20:52 | #5

    what is that next to the usb port?

  6. miketron
    January 13th, 2010 at 21:03 | #6

    @RockMan Rock
    That would be the programing header for when they are programing the IC in the Launchpad. I am currently messing around with this part of the Launchpad. I’m looking into rewriting some of the internal code to optimize the LP for usb 2.0

  7. May 3rd, 2010 at 00:49 | #7

    great post as usual!

  8. June 1st, 2010 at 10:15 | #8

    like red and blue, too get red, blue and purple. With green and blue, you would get Green, Blue and Cyan.

  9. Tom
    June 25th, 2010 at 02:24 | #9

    Hi. I am currently messing around with the launchpad on linux and looked at the specifications. My first reaction was “holy s**t, the protocol they designed is utter crap !”

    I mean, updating all the LEDs with 4 intensity level would optimally take 40 bytes, but somehow they can’t update the whole thing in less than 200 milliseconds ! I know USB 1.1 data rate is not so great, but come on !

    Moreover, they are using some kind of half-assed MIDI format for communication, which is not actually MIDI so you need a special driver so it can communicate with the rest of the world. What’s the point of using MIDI messages in the first place, then ?

    I see in the comments that you were planning to mess around with the firmware. Did you do anything on that part ? I think the launchpad would benefit greatly from a complete rewrite of the firmware …

  10. ben
    August 17th, 2010 at 19:23 | #10

    Hi there. I designed the Launchpad electronics and firmware, so that PCB is/was my handiwork. There were good reasons for doing it this way back in 2008, when the design cycle started.

    For economic reasons, we used an eight-bit processor that runs a variant of USB 1.1 called ‘low-speed USB’. It’s designed for economical eight-bit microcontrollers that cannot handle the transfer rates of full-speed USB, where interrupts come every millisecond and can spill hundreds of bytes of data at you. Hence it’s hamstrung to 800 bytes per endpoint per second. That’s 400 MIDI messages if you use your own spec; 200 if you insist on being class compliant. This is why we picked the former. I still think that the protocol we devised with Ableton is a good manoeuvre around the limitations of the hardware. Perhaps one day we’ll be able to revisit the design and make it better, but there’s always so much to do.

    Speaking of which, if you think you can do better and live in the UK, Novation are looking for more talented hardware/firmware designers, so send us your CV!

    PS. We’re always willing to help out third-party supporters if you’ve got your own questions.

  11. October 9th, 2010 at 14:58 | #11

    Hi Ben, I read that you was involved in the design of the Novation launchpad and I have a question. It would be possible to disable the mixer and session button and then assign to those button anothers user modes via a new firmware update or something similar, do you know an engineer from novation that can make those changes? For my project “The Audiovisual sampler” would be great, because session view & mixer are useless for this:

  12. Tom
    October 10th, 2010 at 03:34 | #12

    @ben: thanks for the insight on design choices ! So this low-speed USB thing put a hard limit at 800bytes/sec.

    Then, with such a limit in mind, why is communication done with not fully compliant 2 or 3 bytes MIDI messages ? I would understand this choice if it made the device a regular MIDI-USB controller that can be used without any drivers, but it doesn’t seems to be the case.

    My guess is: you guys started with a nice regular MIDI controller in mind, but then you had to lower the cost and go for the low-speed USB thing. Suddenly you were stuck with a 200 MIDI messages/sec, which is definitely not enough. But it was already pretty late on the development cycle, so you improved the protocol quite a bit with the 2 bytes MIDI messages trick. By doing so, you had to drop the MIDI compliance, thus introducing a driver on the computer side.

    I would really like to hear more about all this – please contact me at girodt at gmail.

  13. October 16th, 2010 at 04:50 | #13

    I was just browsing for relevant blog posts for my project research when I came across yours. Thanks for the useful information!

  14. ben
    November 13th, 2010 at 20:20 | #14

    Hello again, and apologies for leaving you hanging on for a month: Novation are keeping me fruitfully busy!

    Anyway, to tackle the questions, we started with the microcontroller we used in Nocturn, which was more than adequate for the task, as we’d already proved this concept with Nocturn. So it was designed from the ground up with this compromise in mind. It meant we could keep the cost and development time of Launchpad down, making it economical enough to make and sell, and getting it to market fairly rapidly.

    This may seem like a strange decision until you remember our rationale at the time: remember, this was conceived as an Ableton controller, and although we published the full MIDI spec, we didn’t anticipate the huge general appeal it would have as a hackable controller — especially in the Linux community. If we had seen this coming, we would certainly have given more thought to upping the chip’s spec to make it more versatile than it is, although this would also have compromised the retail price, and one has to sell products to survive! We don’t make life more difficult for our customers deliberately; it just so happens that the best decisions are frequently made in hindsight.

    We have already moved to a new generation of microcontroller for the next range of products (so watch this space!): I can’t really say what the future will hold but we’ll definitely be leaving low-speed USB behind forever. The economics of supply and demand have pretty much levelled the cost discrepancies between many of these microcontrollers, and one can now buy 32-bit technology that can handle full-speed USB fairly cheaply. This wasn’t the case two years ago.

    In response to Fernando, the messages sent by the mode buttons are just standard MIDI messages: the interpretation of this is all controlled by Ableton, which is really outside my area of expertise — you’d have to ask them directly (via their forum) about disabling the ‘mixer mode’ button, which you could then consign it to whichever function you wanted. The only button feature that’s hard-wired into Launchpad is the lightshow you get by pressing the four cursors while tapping ‘view’. A silly little touch, but I’m quite pleased with it 🙂

  15. November 27th, 2010 at 10:51 | #15

    Hey Ben, thanks for making such a cool controller! I’m looking into using it for live video production switching. I bet you did not have that in mind when designing the Launchpad, but it has some interesting possibilities. If you want to follow progress, or even join, check out http://forum.vidblaster.com/showthread.php?t=1676.

  16. Sam Parker
    February 11th, 2011 at 14:04 | #16

    Hey Ben, nice to talk to someone on the novation team.
    I basically use nothing but novation products in my sets and started to experiment with the idea of just using a whole bunch of launchpads and nothing else. It seemed unique, economic, creatively freeing, and also would look damn cool 😀
    However I had to dismiss this idea for the very reason mentioned by several people in this thread: that the launchpad was just a little too slow. I had a 6 launchpad map doing everything from controlling multiple drumkits, controlling keyboards in Native Instruments Massive and Reaktor, as well as controlling the Maschine and Traktor S4 software (this was before the new Traktor Pro 2), controlling a highly complex effects automation system…and of course ableton live all at once and working within each other to create what I thought was the ultimate live performance setup. However… when doing things such as the drum kits, beat juggling in Traktor, or faster keyboard stuff in any synth programs, the lag was just too much. I am sure it wasn’t my computer because A) its a beast, and B) I tried it with just ONE launchpad and it still lagged.
    I understand that the launchpad is an absolutley phenomenal piece for the price, but its age is starting to show 😛
    Could you tell me if (and when) novation is going to make Launchpad Mk II? My hopes are that the buttons will be more like those found on drum machines so they are more performable, as well as of course a faster system.
    novation fan

  17. May 20th, 2011 at 08:10 | #17

    Hi Ben,

    I’d be interested in working with Novation lol, how do i go about that then…


  18. May 20th, 2011 at 10:11 | #18

    It is possible, you have to write a new midi remote script for the launchpad in ableton.
    http://remotescripts.blogspot.com/ this is a good start for python script programming and the liveapi.
    I wrote a new script for the launchpad with all the normal functions but with the Combination Mode feature of the AKAI APC’s, so you can tie together the famous red boxes of various control surfaces so they move along.

  19. gurek
    July 11th, 2011 at 14:25 | #19

    Seems like no one is taking notice about it.. 🙂 Honestly, I would also be happy with the LP2 and it looks that they are busy now in Novation factory so maybe they are finishing new prototypes… I am here for testing it, please, gimme one!!.. 🙂

    I have a question too. I’d like to customize my own LPad, I am focusing on the overall case. Is there any chance to get the measures for the faceplate? Would be great to have the raster of those buttons to avoid wrong prototyping by myself.. .)

    Ben?.. Ha?.. .)

  20. ben
    August 10th, 2011 at 10:25 | #20

    @Darren, want to work for Novation? Send us a CV … the ‘genius’ role at http://www.focusrite.com/careers is pretty much an invitation for speculative applications. We’re growing, and always on the look-out for talent. Who isn’t?

    @gurek … no Launchpad 2 in the immediate-immediate future, but there’s some other fun stuff coming out. Wouldn’t rule out a Launchpad rework one day though: we speak about it fairly often, and other products are giving us opportunities to develop improved technologies for button grids, the communications layer, and so on. I’m a product engineer and have little say in what we design, but perhaps it’ll fall out when it’s ready.

    @gurek again … Drawings: don’t have my computer on me, but drop an email to ben@[the obvious place] and I’ll see if I have the simple PDF I used as a guide (and also if you’re allowed to have it: I don’t see why not). I remember the grid spacing being a round number of millimetres so it’d be hard to go wrong.

  21. Chris Hnah
    March 27th, 2013 at 13:59 | #21


    Ben, now that it’s 2013, I was curious if you knew if there has been any slight revision to the LP? I just saw one in a local retailer and it looks like the form factor was slightly different, and the buttons felt less clicky? Do you know if there have been any changes made to it, and what those would be?

  22. James
    September 4th, 2013 at 11:01 | #22

    we plan to do another teardown of the LP 2 it has been awhile for you guys. It has a number of new features. I will refer you to the wonderful Create digital music artical on the subject.


    we plan to have another post out in a couple or so on the subject

    The coolest feature is probably “the Launchpad is no longer tied to specific drivers and software. Class-compliant operation means you can plug it into any Mac, Windows, or Linux machine (hello, Raspberry Pi), as well as, via the USB Camera Connection Kit, an iPad.” -Create Digital Music.

  23. January 12th, 2014 at 11:35 | #23

    you inspired me to make a little mod with arduino and midi out, now it’s a standalone stepsequencer: http://www.charon.sk/penelope

  24. Geoff
    March 15th, 2014 at 15:57 | #24

    I just wondered if you can help. I managed to plug 12v into my launchpad and it stopped working…DAMN! Using a usb hub, and the wrong adapter…I took my pad apart and noticed that the resistor near the the usb socket was a bit burned looking.
    labelled on the board as R9.
    I don’t suppose you could have a look and let me know what the coloured bands are… It looks like Red black black grey Red…. but the grey bit could just be burned-ness…
    I obviously a bit narked at myself for killing the pad, but if a resistor is all it needs (fingers crossed) then all will be well again.

    I’d really appreciate your help.

  1. November 23rd, 2009 at 06:33 | #1
  2. November 23rd, 2009 at 08:50 | #2
  3. November 23rd, 2009 at 11:02 | #3