If the breakout board could be mounted inside then I’d settle for a DN25 connector. It would be nice if there was room for it on the back panel but if not either side would do.
Thank you for expressing your interest and excitement! It serves as the fuel I need to make progress on this
I have not disassembled a Digitone but I suspect there is not much space in it. If there is, a DB25 version will not be hard to make.
I took some measurements today. After some tweaks to my code I was able to get the latency down to ~6.6msec. I believe the final version will be somewhere between that and 10msec (if I find out that 6.6msec is unstable). In comparison, my (admittedly out-of-the-box unoptimized in any way) Ableton installation on my 2018 Macbook Pro reports 57.6ms latency - but I am not sure if that is just Overbridge, or does other stuff like the onboard audio interface play into that.
Here’s a capture of the scope when I made the measurement today:
The top (yellow) trace is the Digitone master output and the blue trace is my board. Some things to note are:
The output level of my device is lower than the Digitone (doesn’t seem to be an issue when I plug it to my mixer). I believe I will be able to improve this.
The signal is more distorted coming out from my device. I can’t hear a difference when I’m comparing the two, but this is certainly not optimal. I will try and improve this, but I am not an expect on analog circuit design so this might be the best I can offer for the upcoming revision.
I’m curious if these would be a deal breaker for those that have shown interest.
How do you feel about ~6.6-10msec of latency? Please note that since the board breaks out the master output as well, you won’t have any latency issues between sounds coming out of the Digitone if you only get them through this board. The latency will come into play when mixing with other equipment, but I am hoping that <10msec is acceptable.
This is what I’m seeing in Ableton when I hover on the Digitone plugin:
to check whether it is noticeable or not, i’d try to mix two tracks with percussive content in Ableton, and quicly swich the second track with a 10ms track offset and check if the groove feels noticeabily different.
The inaudible distortion is probably high frequency ringing of the output opamp.
The damping capacitor on the feedback path need to be adjusted.
For good audio quality the power line should be as noise free as possible.
this works ITB only, though you also do need overbreidge, or an audio to midi sync box to keep the sequencers in sync.
OTB you could have the digitone playing percussive parts in sync with an analog drum machine, so latency could be an issue, unless you have a dedicated Sync clock box that is able to offset the clocks like the ERM multiclock.
So cool. Think i already responded in the thread but i no longer have a DTone, thinking of getting a DTakt though and if so this would be very cool. But i still maintain i´d want to solder it myself, more fun that way. Unless some really complex IC´s are there is could be pretty simple.
Not so many components either. What bothers me is fitting it into the machine. I know i can handle it myself but it could be disastrous and will of course void warranty if one does not have the right tools and skills.
Just because of this, i would prefer an external box instead of messing with the internals, if you´re not knowing your stuff. The elektron people will not be very happy either i guess.
There are four recordings, all going through my Mackie 1642VLZ3 and recording using a Motu Ultralite mk3 using Audacity. Only post-processing applied is Audacity’s normalization with its default settings. The sounds I used are factory DT sounds. Kick is “Thug Kick SM” and snare is “Class Snare SM”.
Kick comes from the DT master channel, Snare comes from my board. DT configured to not output the snare to the master channel.
Both kick and snare come from my board (I grabbed the master through my board as opposed to from the DT as the previous recording)
Kick+Snare coming from the DT master (my board not involved)
Left channel is the kick through my board, right channel is the kick through the master output of the DT
Please let me know if you want me to try any other stuff.
Thank you for those who are trying to help with the analog design - that is very welcome. Here’s the output stage from the DACs:
It sounds like I need to experiment with differnt values for the feedback capacitors C9/C13. It’s also possible the +/-12V supply is problematic. It is generated using a DC-DC stepup module from the 5V usb device input that is powering the board. Not optimal but I didn’t want to require another power brick. I figured in the future I could offer a slightly more expensive version with a higher-quality PSU and balanced outputs but for now I am trying hard to keep costs low in small quantity batches.
I am without access to my workshop for a few days so I won’t get to experimenting with different capacitor values until that is resolved, but would of course love to hear any feedback or suggestions for things to try out.
Those caps seem reasonable actually. I typically use 22pF to 100pF for those purposes depending on op amp and function.
Are C11 and C15 for AC coupling? If so, you could try 22uF there. Might generate slightly less distortion.
Also, if you have an account on Muff’s or Electro-Music there are tons of helpful people in the DIY forums. Some are absolute experts in these fields. I’m fairly experienced, but some of those guys are wizard level.
Yup, C11/C15 are for AC coupling. I haven’t spent much time on electro-music but I have been reading random posts on Muff’s DIY forum throughout the year. Good idea, I can post there. A thing I want to try doing before I do that, once I find the time, is to try and write code that plays a sine or square wave through the DAC, removing the DT/USB from the equation.