latigid on

Glad that you got it sorted! Either a hot-air reflow station or TK.'s special trick with a thin wire underneath the SOIC is recommended (this was done for harvesting old chips used in OPL3 modules).

Pictures? You can also use the dedicated thread here: Did you solder in an isolated resistor network rather than a bussed one? Is the polarity correct? The random data might indicate floating 165 inputs. Other ideas: Try also with set debug on to see the actual SR positions. Is the NG file the LH one, assuming no other boards are connected? Is there any conductive flux residue left around the pins? Are you using the wCore or something else for testing? What is the length of the ribbon cable? BLM positions are here. SR side is as you would expect, but also attached here for convenience. Best from snowy Boston, Andy

I agree with Zam: probably those DAC channels are dead (but do check directly at the pins). The TLV chip is quite sensitive to heat.

Great work and nice pictures! Are the encoder pins in question shorting out on the pinheaders below? If so, trim the pins a little. I'd recommend Reichelt for MEC switches. Conrad also sells them, and the caps I think are cheaper there.

I've done exactly this (with the board outline at the edge) at OSHPark. About half of my boards had missing copper though, as I suppose the router can pull the plating off.

Maybe you can get it to a single layer by routing through the SIP header. For pads to the edge, you could consider a castellated pad/via. Normally the fab will charge extra for it. I'm not sure if it's a good idea mechanically; the simple pin header would be stronger. Or maybe I don't understand your idea.

Drum labels sound like an extremely useful improvement!

When you look at a male IDC pinheader from the top, the pins are numbered ________________________ |02 04 06 08 10 12 14 16 | |01 03 05 07 09 11 13 15 | __________----___________ Because we normally connect displays on the rear side, the Core J15 is mirrored, so it's actually ________________________ |16 14 12 10 08 06 04 02 | |15 13 11 09 07 05 03 01 | __________----___________ Thus the pin numbering starts at 01 in the bottom right corner. Could you check that these pin numbers correspond to those on the LCD when the cable is connected (with the power off)? Or are the pairs inverted as Bruno says? Otherwise the pinning on the LCD (unless it's mislabeled) looks to correspond to the Hitachi standard.

Ah right, these can be very confusing! Also, the way CMOS labels power connections as Vcc Vdd Vss etc. can also lead to mistakes. I didn't check the pin compatibility (please check!), but this PCB might help:

Did you connect the IDC16 cable on the top or the bottom of the display?

Awesome! I keep noticing the blue/green flashes underneath the USB cable :)

A good idea and one I had considered, but I think the cutout would weaken the plate too much there. Plus, the clearance to the I2C board will be quite close when the case is in rack mode. I had a look at headers mounted to a tiny PCB, but I think there's just not enough room, even with micromatch or "paddleboard" type. Consider 10mm spacing, 2.54mm plastic standoff on a straight-pin header, then a 1.6mm carrier PCB. I didn't find a header lower than that height difference. I would suggest to put a sliver of kapton tape on the components/IDC.

If you wish, it is possible to bridge the pairs of pins 4/4 and 5/5 as shown (viewed from the top of the board). It isn't done by default, as simultaneous use with a BLM would cause a conflict and perhaps an expectation of an "extra" MIDI port. If you and any future owner of the SEQ are aware of this it should be okay.

Great that you found the solution! The mounting is very tight indeed, and insulation seems like a very good idea. I wonder if it's better to rather mount the IDC directly to the OLED, for instance using: HIF2E-14D-2.54RB(20) (only 14 pins, but it would work, as the backlight is not required. HIF2E-16D-2.54RB(20) is not in stock at Mouser, but could be ordered. There's also the RA type (grey colour I think) https://www.mouser.de/ProductDetail/Hirose-Connector/HIF2E-14D-254RB20?qs=sGAEpiMZZMs%252bGHln7q6pm7F1NlCFAzwf%2fz35lHmVYIJ7sqU4G2Cmfw%3d%3d

Have fun! Keen to see a video of the jam/performance (if available) :). Are you using the Host port (USBA) of the v4+? Lots of people are asking for CV/gates for SEQ v4+, so I will work on a DIY solution using the intended Line Driver. I think I have a way of generating the +5V inside the Eurorack case that should avoid sending power down the DB-25. Will need to be tested of course! Not intended to discount the FH-1 solution in any way! It's great if it works for you.

Great job! This is with the 62083 chip to drive the activity matrix, correct? I have almost everything ready, but I've been doing PCB design/mods the last days... I'm happy to be beaten to the finish!

I will request that the fab cuts the next lot a bit smaller. We're at the lower limit of the corner radius for the router bit though, so there's not much to be gained. If you centre the superflux legs on the SMT pads, you can control the diagonal displacement to get it a bit more in the middle. Personally, I never noticed an issue with alignment, even with the LED at the side of the cutout. It is actually possible to increase the light transmission by wrapping the sides of the Matias switch with alu foil. Although the risk is low, I wouldn't recommend, just in case the foil falls off and shorts a component. Maybe there's some sort of foil tape?

Nice machine Bruno! Great mods!

Ace! Very nice job!

Just what I would've suggested. Great that it's going now! Best, Andy

Merry Xmas! 62083 is available at TME: https://www.tme.eu/at/details/tbd62083afwg/treiber-integrierte-schaltungen/toshiba/ Best, Andy P.S. fantastic work so far!

62083 is available at TME: https://www.tme.eu/at/details/tbd62083afwg/treiber-integrierte-schaltungen/toshiba/

Is there something like an "enter" button? Seems it would be useful e.g. to confirm a "flashing" menu item when using the jog encoder push function.

For best brightness balance (and LED lifetime), use the latest specified values. The others will work though. You could also get creative with adding two 22R together if you happen to have spares. What is your colour scheme? Normal red-green or something else? Note that the firmware will allow to swap LED colours around. Obviously this only works with the connected LEDs. Peter did a "cyan-red" version for his first build. You can do this by doubling up green and blue for example and use the extra pad located on the RJ.