That’s awesome! I looked up info on that kit pedal and I remember seeing it on diypedals on Reddit, very cool to see you can buy one, here is that post I saw:
https://www.reddit.com/r/diypedals/comments/1ccz8xq/an_smd_version_of_the_daisy_seed_terrarium_pcb/
Smd version of Funbox may need to be the next step. Kind of hoping someone else takes that on… ha!
@newkular did a good job on the Hothouse. He should do SMD Funbox kit…sometimes, I’m just not in the mood for BOMS and orders, though I’ve done TONS of that.
@newkular peer pressure 
That’s exactly how I feel most of the time and why the kit was created (along with making the enclosure and hardware bits turnkey.)
The other example you contributed is in the repo now (cheers again!): HothouseExamples/src/TremVerb at main · clevelandmusicco/HothouseExamples · GitHub. Maybe you could share your port of the VENUS code and we’ll add it?
You know, I’ve already thought about that and the economics of it as a kit are something of a question mark at the moment. When an SMD layout is roughed in and the parts library sorted (using JLC), that question could be mostly answered. Cost-per-board depends on just so many factors.
I suspect the value added with stereo, midi and expression input will support a price increase, but I make no claims on marketing skills.
I’ll post the Venus/Hothouse code after I clean up the crude hacking. I also don’t have the patience to do a proper #define/#ifdef for this, partly because then I’d have to truly understand how the expression and MIDI stuff works.
With the extra room from using smd parts you could also add true bypass relays. That’s one feature I couldn’t add to the through hole board due to space in the 125B.
New video about the Funbox project, just me talking in this one! Going through the GitHub repo, so probably a little dull unless you’re interested in building the pedal.
Awesome documentation video! And that build guide website looks super detailed 
Thank you for sharing these resources to the community!!
Has anyone else had issues with the whine using these PCBs? Just bought one and i’m getting the 1khz whine i’ve heard people talking about. Sadly it doesn’t seem like anyone has a fix for this
@patrick398 That’s a bummer, could you provide some details about your build? Which manufacturer for the pcb, and what quad op amp did you use. If you have a quick way to test the line out from the daisy seed audio out pin, I’d be curious if you also get the noise there, or if it’s occurring within the output buffer. Is the pcb in an enclosure and all boxed up? The one other person I know of that built one had noise before boxing it up, but “dead quiet” after putting it all in the enclosure.
Thanks!
I used JLCPCB for the boards, i’ve been using them for years and always found their quality to be pretty top notch. I just probed the L and R out on the Daisy and the whine is there too, along with some more fizzy noise which i’m guessing is being filtered out by the op amps.
The board is in an enclosure which itself is grounded. I’ve tried with a few different power supples, including the one for my boss IR-200, and a Truetone Onespot which have always been totally silent for me.
Been reading a few threads online and seems to be a problem which occurs across a few different PCBs including the PedalPCB Terrarium and some of the official electro-smith ‘pedals’ which suggests the noise is within the seed itself. I wonder if hardwiring some groundwires from the seed straight to power ground might help.
Is yours completely silent? It’s quite frustrating if this is a problem that only occurs randomly as it prevents me ever being able to put anything into production using the seed.
The whine also seems to be related to how complicated the patch is. The one i’m working on has a couple of delay lines, a load of LFOs and some filters, which i wouldn’t class as particularly complex, but if i just wire the adc to the dac it’s silent.
I’ll attach my compile setting in case that offers any clues.
This is all pretty much default. The patch itself would maybe offer some more information, but probably not much if it works fine otherwise.
Yeah it all works fine and nice and quiet in plugdata
That’s all very helpful, thanks for the detailed response! Yes I’ve also noticed the noise seems to go up the harder the processor is working. Mine isn’t 100% silent, if I turn the volume way up I can hear some noise, but it’s very quiet under normal conditions, and the noise is fairly even across the audible spectrum, not at a certain point like 1khz.
Which op amps are you using? I had originally used a TL074 but I found using an MCP6024 to be quieter, and I update the board and BOM to reflect that change in version 3.2.
I meant on the device of course.
Remind me - does Funbox have the LPF on the codec output?
Yeah i built the 3.2 version using the upgraded op amps. At first didn’t notice the noise when I was running into my mixer and into my hi-fi, but as soon as I went into my guitar amp it was pretty obvious at normal playing levels
Ok tried a few different things i thought might help including rewiring the power and audio jacks and hardwiring the AGND and DGND straight to power ground, to no avail.
I just tried playing around with the blocksize though and bumped it up to 224 and that seems to have significantly improved things. There is still something there, but its a much less annoying frequency and seems quieter also.
What’s happening there then? I don’t really understand the blocksize thing. Is using a high value like this going to cause problems elsewhere later down the line?
That’s really interesting, the block size is how many audio samples it processes at a time, so a higher value increases latency. But it gives the processor more time to do the DSP before converting it back to an analog signal, so it’s not working as hard I suppose. Also it would change the rate that each block goes through, I’ve seen some speculation in the forums that a 48 block size with a 48000 samplerate can cause the 1kHz spike, because 48000/48 = 1000
Higher block sizes reduce overhead (wasted cycles) associated with context switching entering and leaving the ISR for the audio callback. The cost of this efficiency is increased latency.
Early in Daisy history, the understanding was that frequency of noise was directly related to the AudioCallbackRate, presumably because of spikes in the power supply caused by the STM32 power management responding to CPU going to 100% while in the callback. Lowering the block size to 4 raised the frequency of the noise to beyond audible. I don’t recall if it also affected amplitude of the noise signal.
