latigid on

Good to know! I have a PCB with some tight placement, so every mm counts!

Okay, the datasheet gives 23.37mm long, is that what you have? So did you stagger the UDNs in the vertical dimension by spacing with extra sockets? Another way might be to put one chip on the top of the PCB, one on the bottom. You have to ensure the inputs and outputs are the correct way around and the power connections will also be mirrored. Also just to mention; Nico did a few tests with fault testing the UDNs like shorting to ground, bipolar (i.e. negative voltage) CV signals and connecting two UDN outputs together. Everything seems to be okay with the 1k resistors in place. It feels like this could be a more robust solution than using CMOS.

I can think of one way to build this into a regular DOUT module; solder the UDN chip in standing up on one row, then wire through hole 1k resistors standing up on the other row. Connect in the middle, avoiding shorts of course. Comme toujours! But this looks to be a very good solution for level shifting multiple gates/clocks. CMOS chips like the 4504 and 7407 are an order of magnitude faster, but I don't think it's too relevant in our case. Plus these are hex chips with annoying pinouts and unnecessary /OE pins etc. Good job Nico!

To me this points out the ADC performance differences between the PIC and STM. Isn't J5 on the latter 3v3 only? You might be overdriving the converter/clipping the diodes etc.

I'm working on a sort of hybrid DINX6/DOUTX4 board and I have a question on how to treat SC and RC. The BLM is proof that chaining the clock works even when the DIN/DOUT parts are not equal in length. But on DIN/OUTX4 boards sold by SmashTV normally one splits the J8/9 ribbon with an intermediate header. On this board I have the choice to go serial or parallel, so I wonder what the best way is, or if it really matters. Many thanks,

Before going further, you'll want to check how the reset behaviour impacts the device as Thorsten mentioned above. As all of the reset circuits are linked, applying a signal may interact with other parts with unpredictable results. And looking a bit closer, reset on U20 is an active high signal, so a +5V signal probably has to be applied to your pin 10. Formally this also drives the bases of Q6 and Q7 and activates the two other reset nodes. One way might be to disconnect that pin 10 out of circuit by cutting the trace (if possible) and using a Core pin to drive reset high or low, but I couldn't tell you how to do that exactly (maybe via a suitable push pull pin + pull down resistor?). For MIDI in and out, it's not sufficient to wire directly to J11E because MIDI in should run through an optocoupler. It's quite an easy circuit and could be built on a little bit of perfboard. A ready-made solution is at the midibox-shop but I don't know how much room you've got in your box. Having a MIDI out is useful even just as a pass through port for other devices. You could also go old school Core8 which ends up being smaller and has MIDI in and out, but requires a different approach to programming/updating. It springs to mind that a more compact MCU breakout board could be quite useful for basic "one time control" applications such as this.

Hehe :)

Surprised you didn't throw a WiFi router in there and have, like, totally, no wires, man! I see some room about the TBD for a future expansion. :)

Schade! IC20: pin 1 is not connected, pin 15 is not connected, you are lucky this time! Should still work.

Please write in English or at least German! miniCore is just the Core8 minus all unnecessary parts, mostly headers. All pullups are in place and you can easily trace the circuit because it's on one layer only. Those LED pins are common to RN20 pin 12 then the resistor pin opposite (pin 5) goes to IC20 pin 7.

You could use a miniDIN connector, as found on old skool PS/2 mice and keyboards. I think Altitude did exactly that with his BLM (4*16) extension.

Are you at the config page and do you assign the BLM port correctly? If a valid MIDI signal is detected it should say "found" within, say, five seconds. Is the optocoupler installed on the quad I2C board? What connections are you using? A crimped DIL plug? Can you verify these?

Nice, is the idea to use the switches bare (no caps)? How are they to play? Well done!

For any past or future travellers, I measured the DIN and DOUT r5 boards (just one example of each): DIN 100.1*45.2 mm DOUT 100.25*45.1 mm Thickness is about 1.6-1.8 mm. Both boards have exactly the some mounting hole positions relative to each other, i.e. does not rely on the outer milling. 4.3mm (prepared for M4 screws) centres using the x,y centre of the PCB as the origin with the long axis horizontal: (x,y) = (-45.525,-18.01) (-45.525,+18.01) (+45.525,-18.01) (+45.525,+18.01)

Well, on the hardware side you could check resistance from the common cathodes (each group of four are all connected right?) and the mentioned pins on IC18. Left of centre (viewed from the PCB rear) to pin 2, right of centre to pin 4. I'd still look at RN18 and plug in a miniCore to be sure.

@Phatline Gnarly! No warranty for non-standard usage! ;) @TK. The unmounted pads are for terminating the SC/RC lines. Not required as far as I've seen.

Many thanks for FW updates. I've added them to the OP. @Phatline Check that all diodes and LEDs have the correct orientation Those middle 8 LED pairs have their cathodes sunk on the transistors nearby (Q37 and Q35) so check the soldering there, also the adjacent resistors. Do all sets of four cathodes show continuity? Are any shorted to ground? Bases of the transistors go to RN18 (not labelled on your PCB) pins 12 and 14 respectively RN outputs to IC18 pins 4 and 2 respectively

Could be wrong, but I think this is already possible by using the MIDI router inside the SEQ. You would still need an app on the other end to work with the BLM MIDI notes. It's a cool idea to re-purpose the matrix though!

I told you the 4148s would be the most painful part! But they do actually provide a "bridging" function that you don't get with SMT parts. You've got it right: RN17 is on the top side, bottom right corner. RN18 is to the left of RN17, top side, just before the button pads start. Schematics available here: http://www.ucapps.de/mbhp/mbhp_blm_scalar.pdf http://www.ucapps.de/mbhp/mbhp_blm_map.pdf I don't intend to publish the layout files; seeing as there are very few users I'm happy to help on a case-by-case basis.

Strange! I won't be able to help much in the next few days/weeks as I'm moving, but I advise to check around all power connections: capacitors (including tants), pin 16 of ICs and RN1..5, check headers and the R1 bridge. The best way is with a hot air rework station and leaving some sort of continuity check across 5V/0V. Thus when you find the fault you only go as far as needed to rectify it. You possibly have a bridge under a capacitor or the soldermask is scratched and the short joins to the groundplane there. Seeing as you're the second to have this issue, I need to consider more plane isolation in the next run... Hope that you can get it back to a good state!

Seems like the icons in reply are messed up? Link gets cog, emoticon gets chain and code gets smiley face. Or is it just me?

It might be that you just need to play more :) For me it's clear if the track is not in the current 16 step window because the cursor is not scrolling (assumes the sequence is running). I then glance down to the bottom row and choose the appropriate bar. There are also instances where you might want to fill in notes and then expand the track length. I find the BLM works like a "macro view" of the SEQ. I.e. there is bi-directional communication when a parameter/screen is updated. (Try to fill up a grid and then adjust the scale for instance!) So display behaviour on the BLM can't be fundamentally different from the SEQ without significant changes to the latter (I think). Oh, and don't forget 303 mode for easy access to four octaves! (May I suggest feature requests are discussed in the thread here? Just to keep this one related to build experiences.)

Something that isn't documented as far as I can see: Shift+3 and Shift+4 toggle Mute and Solo modes. Could we think about a more optimal way? It seems like the best usage is described on the ucapps page above but not implemented yet?

Aha, I was going to suggest checking the optocoupler! For usage, check out the overview here. An expanded walkthrough could be good at some point. Have fun!

Ha, well maybe the next time I'm in Munich! Sounds like I will have to sit down and try out all of the possibilities as a New Year project.