Do you have any pictures? I personally found installing the LED to be a quirky part of the build, but once powered up and it worked I never thought about it again until this post 
One of the LEDs should activate whenever an effect is ON (there are some edge cases where if you switch effects while active it might not be accurate, but toggling off and back on always “fixes” it for me). The 2nd LED is used per-effect, the tremolo it pulses with the trem, the noise gate it represents whether the gate is active or not, the metronome it pulses to the beat, etc…
1 Like
Thanks for the reply. I’m not sure there’s much to see but here’s a pic. The positive terminals of the LEDs are soldered in the hole to the right as you’re looking down upon the pedal. I’ve checked the schematic and I can see the positive terminals of the LEDs are connected to the Daisy itself (pins 29 and 30) so I presume they’re under software control. I’ve built 2 pedals now and neither of them have working LEDs as far as I can see. The sound effects all work, though.
Oops, here’s the other side. I do the SMD soldering myself, in fact I prefer it to through hole. I build a lot of eurorack modules for modular synths, so I’m pretty sure my soldering is good, but perhaps I’ve put a component in the wrong spot? It’s not really possible to see in these pics, but I can try to get one through my microscope.
The positive LED leads are not connected directly to the daisy, there is a resistor in line with each, R24 and R25 under the daisy, can you check with a DMM the resistance between the positive leg (non square pad) and the pin on the daisy and make sure they show 1k? And then continuity test the other leg (not the pad for each of these, the leg of the LED since its easy to access) to GND.
Besides that, yea, not much to try…if it were me I would have flipped one around by now because it is such a common potential issue, but considering you solder SMD components, I will not insult you with such a suggestion 
All I can say is I have not had any issues with the LEDs and they should certainly be functional
Hi and thanks for the reply.
Yes, my measurements show there is 1k between the LED and the daisy. I actually reduced the value of the 1k resistor to 330R just to see if it made a difference and it doesn’t. No LED light at all.
I have built 3 pedals now. (JLCPCB sends you 5 boards when you make your 1st order.) I actually only need one, but I built the others just to see if my build quality was bad - but I get exactly the same results on all of them. Everything appears to work but the LEDs never light for any reason. So either I have done something wrong on all 3, or the BOM is somehow wrong, or the software has a bug. I’m using the absolute latest from GitHub. Compiled and loaded on MacOS. I may try building and loading from Windows just to see if there’s a difference somehow.
Thx for any help. To be perfectly honest, this pedal is for a close friend who plays guitar in an active band. It was to be a birthday gift, but unfortunately with the LEDs not working it kind of seems defective, so I am afraid I’ve just done something obviously wrong (three times now…)
1 Like
I would use a dmm on the daisy pin used for the main LED while powered on when the effect is bypassed and not bypassed. This will help rule out SW. I use a mac, I wouldn’t bother with recompiling on windows. I would also use a dmm (with an alligator clip maybe) on the leg of the LED as well.
Also I just checked briefly and all of these modules (others do too but I wanted to give an explicit list of a few I checked) have both LEDs illuminated at all times when on: the para/graphic EQ, the NAM module, the cloudseed module, tuner
I also just built and tested the latest to check out the new drum module and its really neat @keyth72
1 Like
I found the problem! It was indeed software. I now have 2 working pedals (the third got cannibalized in the process of making the others - but I only need 1 anyway )
I put everything on Linux. I’m using Ubuntu Mint under Parallels on a Mac. Yeah, lots of layers of stuff there. It was a real pain to get the usb and dfu-util set up correctly. And one other thing - if you follow the DaisySeed support site, and navigate to how to install on LInux, they tell you the libs will not compile if you use a version of arm-none-eabi-gcc equal to or greater than 10.3 I found this to be the opposite of truth. It did not compile with earlier versions than 10.3, and works fine with 10.3.1
Once compiled building the app was trivial - the makefile works just fine and there are tons of warnings about things about to be deprecated but no errors that stop compilation. I suspect in the future these deprecated C++ functions will become a problem in the code, but they are not now.
In any case - for reasons I can’t currently fathom I do not know why the LEDs would not light on versions built on MacOS and Windows - but do work fine with Linux.
Well, this took a month or so to debug and I was about to abandon the entire project and call it a lost cause failure on my part - but persistence paid off. I did wind up creating an stm32dfu rule in the /etc/udev/rules.d directory, and I found this to be necessary. Here are the contents of that file. I have no clue if this is generally necessary but it worked for me. Again, I’m using Ubuntu/Mint 21.2 on Parallels on a Mac. Lots of weird layers but the pedal works now!
I created a file named 50-stm32dfu.rules with these contents:
SUBSYSTEMS==“usb”
ATTRS{idVendor}==“0483”
ATTRS{idProduct}==“df11”
GROUP=“plugdev”
MODE=“0660”
TAG+=“uaccess”
It doesn’t seem that anyone else has got stuck on this, but maybe it will be useful for someone.
Cheers
Joe
2 Likes
What in the world! glad you stuck with it. I did put precompiled .bin files under the github releases, so I suppose that could have been a valid debugging step too.
Anyways, I am glad you got it resolved, definitely reach out if you have any other questions!
I actually tried the precompiled bin file, but that didn’t work for me either. However, I was trying to upload it from my Mac so it could be there’s some obscure usb transfer problem with MacOS. Well, I’m a keyboard player but my buddy is a guitarist. I made this for him for his birthday but I missed that so this will be something I give him over beers someday. I’ve got 2 of them now (and 2 more PCBs) so I’ll order another enclosure from Tayda and keep one for myself I can honestly say the SMD soldering wasn’t bad at all as the parts are 805 size. You wind up spending a lot of time with a microscope and tweezers, but I find it quite zen to sit there and put these little things in place. I wouldn’t shy away from the SMD if you’re so inclined, though when all is said and done, ordering the parts from Mouser and the boards separately probably costs as much as just having JLCPCB do the SMD stuff themselves. Though it would save some time.
Thanks for the replies.
Joe
1 Like
While I do have an EE background, I have been on the dark side (software) my entire career so I was eager to get straight to the software part of the project I also only wanted one pedal so I did an order of 5x pcb with only 1x PCBA. That means I do have a few boards laying around unpopulated but I dont think I have the patience to go for it haha
Hey same here. I’m an EE by education, but have been in software (EDA) my entire adult life. I kept wanting to switch to a design job but the commercial software side of things always paid better and as time went by, I was kind of pigeon holed and I became a “software guy” instead of a “design guy.” And that’s how it went for 40 years or so until I retired. I don’t regret my choice but I keep doing these DIY projects just to remind myself that my goal as a young person was to be in electronics (and I sort of was - I did software for the chip design/manufacturing business. Funny how life turns out.
1 Like
I’m just wondering how it became necessary to build under Linux just to get LEDs working, while other stuff seems to work. Or why running pre-built binaries doesn’t seem to work.
Well, there might be some old pre-built binaries which might predate the newer codec used for the past couple years. But that probably isn’t the case here, as these boards are newer than the codec change.
I have two Macs: an old MacBook Pro 2015, and a much newer MacBook Air M1. Both work fine for building Daisy programs, with no bizarre exceptions, and no special tricks installing the toolchain. I build for Seed, patch.Init(), Hothouse, terrarium, Bluemchen…
1 Like
Yeah, that’s super weird it didn’t work when building for a mac for you. I used a Mac a the whole time while working on this project never ran into an issue like that. So strange 
Glad you got it working though!