latigid on

https://www.chk-electronics.com/P401.html or equivalent.

Yes. I'm using 19mm spacing.

As explained above, the notch of the connector only matters when both PCB headers are keyed (allowing only one orientation). Considering putting the header on one end of the cable, the same pins are connected regardless of whether the notch on the IDC is "correctly" aligned with pin one (lower left as viewed from the top) or rotated about 180 degrees. But in TK.'s example above, rotating the cable about 180 degrees at the display end will mean the backlight circuitry will be where the power rails are supposed to be and vise versa, and this is not good for the LCD in most cases.

Doesn't really matter. +5V draws off the USB power buss, so less pressure on the regulator. Most OLEDs use a charge pump for higher voltage anyway; there is no energy to be saved.

Verify each of the intended connection points on J15 are made correctly with the display. Check for shorts between the cable pins (can happen with misaligned header) You are fiddling with the contrast and brightness controls each time, right? What firmware is loaded on the Core?

Thanks for the positive feedback! Indeed, the limitation of this version is that it is 3U standard but sadly not Eurorack-mountable owing to the rails as @eptheca states. Here's an interesting compromise though: The PCBs fit in a standard 3U panel, 85-86 HP width. Use 1U tiles ~40HP (i.e. 4U/7U case) to mount the displays above. The CVs/gates/clocks/triggers etc. could easily fit below the PCBs with room for a MIDI socket or two. This wouldn't be too hard to interface and could even be hand-wired. Still necessary would be mounting points (brackets?) for the panel and Core/peripheral PCBs. Personally I still think the power and space requirements of a full SEQ are suited to a desktop or 3U rack unit. But I can understand the wants and desires of self-containment and intermingling with the modular goodness. :)

Great! The structure already looks much flatter and quite useful for pattern changing. Nice work from your side too.

All looks standard. On an IDC header, pin 1 is on the bottom left when the notch is on the bottom. As the header is soldered to the rear of the display but PCBs are mostly designed from the front, the J15 Core header is mirrored to account for this.

As long as the LCD has a standard pinout, what you're doing looks correct. You should test another LCD.

A common problem with Eurorack power headers is they only work if the notches are polarised the same way on both the bussboard and the module. If the module has "pin 1" on the opposite side as the power output header and both headers are keyed, there is no way to correctly plug in the cable. The Core schematic is correctly drawn as a "mirrored" header.

It's easy to fry LCDs, e.g. if you get the backlight circuit going full blast onto the logic pins, reverse the power connector or a data input etc. I'd try another LCD. Ensure the cables are connected; it can be confusing without polarised headers (always missing from LCDs). Don't trust the notch on the Core board, use the stripe of the cable to align the power or backlight.

NPN transistors don't source current very well. I'd suggest moving the diode in series with the resistor on the collector and tying the emitter to 0V. This way the current can be controlled with R1, otherwise it's not doing much. The capacitor you've shown in the Falstad sim won't do anything. R3 isn't necessary if you have the 220R in place. D1 will block reverse voltages but will also drop the gate voltage. Voltage < 0V will pass through. As long as the current is limited the 595 output will probably be okay with another output connected, even excluding D1. Diode clamps are better done in parallel with the node to protect, diodes connected e.g. to the power rails. I'd suggest testing the circuit on a breadboard or perfboard/stripboard before ordering the PCBs.

Can you post the actual circuit diagram you're going off?

Meh, all of the details are there if you look.

Pin 3 has no isolation so is part of the ground (0V) plane.

http://wiki.midibox.org/doku.php?id=wcore http://wiki.midibox.org/doku.php?id=wcoref4_pcb

Check continuity from pin PC6 on the Disco board to R3 on the second MIDIO. This will resolve any cable interconnection issues.

http://www.midibox.org/dokuwiki/doku.php?id=wcore&s[]=wcore You can't use the USB port without flashing first, unless you have a USB to UART programmer. If you have a Discovery board you can use that, or else pick up a cheap ST-Link clone.

Hi Yannick, Absolutely, Adrian has the PCBs and is beginning his work ! Best, Andy

Where's your Disco board from?

The LCDs expect an "enabled" signal, so I wonder if it's a problem having only one connected. Load up something easy such as MB_NG and see if it works. Reflow the joints on the 595 socket (chip out first). You could have some cold joints; ideally they should not "bulge" around the PTH pad. What is the scratch between J15A/B headers on the solder side?

You don't need an SD card to display LCD messages. Photos of the PCB please. What LCD are you using? Is the voltage jumper correct? Does one LCD work by itself? Is the shift register 74HC595 (not HCT)? Can you swap it for another to test?

Did you adjust the contrast pot?

Not for this iteration, the height of everything is about 124mm. But a Eurorack-mounted version would take up ~84HP and consume a lot of your +5V rail. The limiting factor here is the OLEDs. If there were smaller displays, the whole thing could be more compact.

It's turned out well! Just needs some nice OLEDs! And keycaps, but these will come a little later.