wackazong

I did some tests today on this topic. I had very good success by simply debouncing the encoders with two caps of 0.1uF between the common pin and the output pins. This can even be done on the encoders pins themselves, no need to change the PCB. Now I have not had a single bounce so far during my tests. Seems to work and is simple enough for me to do in hardware.

Hello again, I would like to do some kind of software debouncing for my encoders. I have them at DETENTED3 and FAST 2 now, this gives the right frequency and step size of value changes in relation to the turning velocity. However, the introduction of FAST mode brings a problem: when the encoder bounces, the value suddenly moves very fast, as the algorithm notices very rapid value changes in the SRIO chain, then applies acceleration and generates a big value jump. Is it possible to debounce the encoders? I can see in the code that only buttons are debounced now. I was wondering whether there is some way to activate it for encoders. Thanks! Ander.

Yes, Station is already updated :)

Wow, much better! I like the new feeling very much. Thanks for that. For me speed_par 1 is best. Almost full 360 degree turn when I turn fast. As most of my encoders are set to the middle value and I deviate from that, this is enough for me. A tad slower even could be nice, but this is already a big improvement.

Well.... I just finished my own Midibox, now a friend wants one, too :) I would like to teach him, although he has never soldered or programmed before. Will be interesting. He is looking for a relatively simple Midibox 64E kind of thing, but he wants LED rings. Is there an existing PCB layout for LED rings which he could use? I searched forum and Wiki but could not find anything.

Thanks :) Was not really easy... 1) Record video and audio 2) sync both in Ableton live 3) render snippets which are multiples of 8 or four bars, add some overhead on both sides 4) manually do the camera changes in iMovie using the "precision" editor

The encoders are standard Bourns PEC11, no special quality there. I did not find any suitable SMD encoders, so for theses parts I had to put some holes into the PCB. This makes it more rugged, so maybe it was a good decision after all. The push buttons are standard SMD pushbuttons, I needed some with very low actuation force, therefore I took the DTSM-65K-V-B from Diptronics. No surprises here, theses parts would work in a normal midibox without problems.

Yeah, the routing was difficult. I did it with kicad, it took me i think about 5 attempts. Regarding your questions: 1) yes, correct. White soldermask is just a tiny bit more expensive than green at http://www.multipcb.de/. I can really recommend them by the way. 2) I don't get that, sorry. I have all module connected in one row, therefore there is one input and one output on each PCB. It connects the DIN and the I2C bus (for the LEDs) 3) No, 0.125W would not be enough. Each array drives 16 LEDs. Unfortunaltely, the PC!9635 requires resistors, I would have liked a part with constant current regulator, but I did not find one that suited me. 4) Yes, they determine the I2C address of the driver chip, which has to be different for each chip. There are three on each PCB, one for each color. 5) Yes, more or less. one signal layer, one ground layer, a mixed power/additional PWM layer and one layer only for PWM (of the LEDs)

Yes, I use Ableton Live for music production.

Well spotted ;) Your are right, the PCB is the same for all modules. It can hold either one encoder or two pushbuttons in the same area, or any kind of sensor with analog output. One PCB can hold 16 pushbuttons or 8 encoders/sensors, or any combination. Actually, one module consists of all the necessary PCB stuff for - one STM32 core - 2 DINs (in a SMD version) - one step down converter (24V to 5V) - three driver chips to drive 16 RGB LEDs (PCA9635). This is a design of my own, I do not use DOUT modules The core parts are of course not soldered in on every module, but of course this one PCB design made developing and producing the PCB much cheaper. Its actually 4 Layers, 16 RGB LEDs need quite a lot of routing...

Thanks Wilba for posting the pictures. BTW, this was all hand-soldered by me.... Very meditative. I have made a quick drawing of the button mechanism. It works quite well, you can push on it everywhere on the surface and it activates the button. I selected a pushbutton with only 150g of force.

While I am at it: Many thanks to you for supporting me during this project. I could not have done it without the help of you and the great MBHP. I will try to post some photos, source and docs here in due course.

The website seems to be down, anyone have the photos so we can host them somewhere else?

Hi, regarding the buttons: Its an acrylic button I had made for me, 6mm of opaque acrylic plastic. It fits directly into the case and is secured by a small notch so it cannot fall out. One side is supported by a small piece of elastic foam, the other side has the pushbutton below it, so that the led can sit in the middle. Quite simplistic, actually, but it works very well so far. Next time I would put notches on both sides and tink MUCH more about the mechanics before building them, but as long as it works now I am happy :).

I have the Safari option "Accept cookie only from sites that I visit" on. Maybe that is blocking the cookie. I will try.

One thing about the new forum: Wouldn't it be possible to add a login possibility directly from the start page? I don't like having to click on login first, actually.

Ahh, that was it... I had it figured out once. Thanks!

BTW, how do you get MIOSStudio to save the window layout for the next launch? It will always open with an empty workspace here....

Funny thing, on Mac 10.6.2 it will not only unpack the zip if you double click, it will unpack the .jar file as well, leaving with a directory with the jave classes... Not really useful. Do not use the integrated Archive Utility for unpacking, use something like 7zX, that one will work correctly.

More: The STM Primer 2 can only record at 16kHz, which is not enough for me. Anybody know a good audio codec? I ordered some samples of PCM5310.

OK, did some more research: The STM32 Primer 2 looks like it could do all of that. Any chance to get that functionality on the MBHP STM32 core as well? It seems like the CircleOS is available in source code. As far as I can see I could port the audio stuff to MIOS. It is written for a specific codec chip though, the STw5094a. But that's only in BGA.... I will have a look for a suitable codec chip with input and output. I would like CD quality (16bit, 44kHz). Is that a problem with the available memory of the STM chip maybe? Would continuous playback/record be possible?

Hello, for another project of mine I want to construct an interactive Sound box. I want to base it on the MBHP, with the STM32 (as I have some experience now and also a working toolchain :) The features are: The Box consists of an old but working loudspeaker chassis, with speaker and everything. Inside is all the electronic stuff, out comes just the AC power and a USB cable. The box has IR distance sensors and a microphone input. It reacts to people coming closer and to incoming sound signals by playing short samples (from an SD Card) and/or replaying the sound that comes in. It is going to be based on the STM32 core PCB with as few additional components as possible. What I need to know: 1) How can I make a line out from the STM32? Do I use the I2S for that? 2) How can I play sound files from a connected SD-Card? Could I use MP3 or do I have to use WAV/AIFF? 3) How do I make a line in? 4) How do I load sound files on the card? I would like to plug in the USB cable from the STM and just use the whole thing as a USB mass storage. I would use an encapsuled amplifier for sound output from the line out to the speaker, these things are really cheap nowadays and extremely small. For the Input I would also use an encapsuled preamp to bring the signal from mic to line level. The IR distance sensors would be connected to AINs. Thanks for ideas, answers, pointers, posts, whatever, ALEXander.

Very nice. But what does it do exactly? Depending on that, kit, yes please! Best, ALEXander.

Hey, I ran into some serious issues with my board design today, maybe somebody here can help me out. I have various modules in my MIDIBox-based controller, which are supplied by 24V from a AC/DC converter. On each PCB, a step down-converter (Texas Instruments TPS5420) converts the 24V DC to 5V DC. This works beautifully. I designed the PCB exactly like it is proposed in the datasheet, and followed the parts advisor in the Texas Instruments SWIFT tool. The 5V come out ok. To supply my MIOS32 core, I then convert the 5V down to 3.3V using a linear converter (ST Microelectronics LK112SM33TR). All wired like proposed in the datasheet (this one is easy). Now, if I supply my board directly with 5V (for testing purposes I just get 5V directly from the USB port), everything works, and the core gets 3.3V. This has been working for a couple of weeks now. Then I plugged in the 24V power supply. Things worked for a couple of hours. Now suddenly I have only about 0.2V coming out of the linear regulator, instead of 3.3V, and it gets extremely hot as soon as I plug in the 24V. Does anybody have an idea why 5V coming from the step down converter is different to 5V coming from a USB port? I changed to linear regulator and put in a new one, but that did not change anything. Should I maybe take another linear regulator model? Thanks for any pointers, ALEXander.