808 Bass Drum Clone Eagle PCB & CNC Gcode

jbeuckm

Here is a bass drum layout for Eurorack power connector:

https://github.com/jbeuckm/808-Bass-Drum

I built this with 2n3904/6. I thought I was buying film capacitors but these turn out to be MLCC. I would be interested to hear opinions on how this sounds (sound sample attached) compared to the film caps / BC546 or original transistors. Sounds like an 808 to me!

808_2n3904_multilayer_ceramics.mp3

Description:

increase tone, then increase decay, then decrease tone. velocity cv pattern is sent to accent.

Filesize:

734.75 KB

Viewed:

2515 Time(s)

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Sounds good for me.
Freakin' good repo by the way. Will suck all of your files Very Happy

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Love the curved traces.
Did you etched this board?
Looks great Very Happy

Any "test-results" about the snare?

Future plans for hats?

jbeuckm · Posted: Thu Jan 15, 2015 8:45 am Post subject:

Thanks for the feedback! Yep - the sample is from the board I etched from the eagle files. I am debugging the snare right now - had some mistakes in the schematic. Then rimshot/clave and then hats.

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

jbeuckm · Posted: Thu Jan 15, 2015 10:57 am Post subject:

Ah that is a good add - then normal it to a switched accent input jack. Submit a pull request?

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Another great idea!

jbeuckm · Posted: Thu Jan 15, 2015 9:16 pm Post subject:

Don't etch that snare yet - the 72's are all flipped Sad

Here is a sound sample with a very ugly kludged board. Have to redo the layout...

808_sn_2n3904_mlcc.mp3

Description:

Filesize:

734.73 KB

Viewed:

2137 Time(s)

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

jbeuckm · Posted: Tue Jan 27, 2015 8:19 am Post subject:

Snaper,

I rerouted the snare but haven't built the new version yet - soon. I also built the rimshot/clave but only the clave is working. Would you want to build it and help me debug? I checked the schematic a lot so I might just have a bad part or connection.

https://github.com/jbeuckm/808-Rimshot-Clave

Open hihat sounds good. Will test the closed tonight:

https://github.com/jbeuckm/808-Hihats

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Will check them, but can't etch them right now, too much stuffs on the schefule!
Will look for the schemo;)

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Checked the updated schemos, they're getting better and better.
Still love the curved traces.

Made some minor changes, basicall put 16 mm pots and align them to make the PCB panel mountable.

Checked also the Clave Rim schemo. For the first, looks OK.

Fortunately, the problem should be somewhere before the switch.

Which one is the rimshot, top or bottom part?

jbeuckm · Posted: Wed Jan 28, 2015 8:16 am Post subject:

Hihats work well. Here is a sample of the bass/snare/clave/hihats together. I am moving everything toward 16mm pots for panel mount too. I have to reroute a few signals because there is only room for one trace in front of the pot holes.

808_bd_sn_cl_hh 1.mp3

Description:

Filesize:

734.73 KB

Viewed:

1983 Time(s)

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Oh, almost missed this thread?
Any news?
Im in the rack build business, would be great to build a separate box ...

jbeuckm · Posted: Mon Aug 17, 2015 9:46 pm Post subject:

Snaper,

The rimshort/clave is in the lower right part of page 9 here: http://privat.bahnhof.se/wb447909/dinsync/service_manuals/TR-808.pdf

There is a DPDT switch to change voices. It looks like the rimshot is the "top" part and a change to the bottom part too.

Also, in my layout, I sub a 5950 for the FET. My built prototype has a 5953 kludged in - not sure if that changes anything. What does the FET do here?

jbeuckm · Posted: Tue Aug 18, 2015 8:19 am Post subject:

Aaaand I found a bridge in the etched traces. Cleaned it up, and I hear rimshots! The rimshot is pretty quiet relative to the clave. Interesting, the value for R324 (connecting the upper rimshot circuit to the switch) is not labeled in the schematic. I might try to reduce that resistance and bring up the rimshot level to about where the clave is.

clave.wav

Description:

808 clave voice clone - normalized

Filesize:

4.09 KB

Viewed:

1776 Time(s)

rimshot.wav

Description:

808 rimshot voice clone - normalized

Filesize:

1.84 KB

Viewed:

1785 Time(s)

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Man, you're hardcore!!!
That means, you've made home etchable layouts for almost every voice cards???
You have to blog about this Very Happy

Are those layouts on your github tested?

I'm pretty sure I will build a patchable, modular 808 drum machine into a 3U high box, with an FCUK mini power!

jbeuckm · Posted: Mon May 02, 2016 12:27 pm Post subject:

@Snaper, All my 808 stuff on github is tested and working now. I am redoing the rimshot/clave but the previous version worked - just redoing for better parts and thicker traces. That' all voices now except for maraca!

I also have a MIDI interface that will generate the right kind of triggers and "accent" signals. I only found a practical 8-way DAC, so you would need two of these to run all 11 voices:

https://github.com/jbeuckm/atmega_midi_cv_drums

Maybe I could fit a 12-DAC version onto one board - will probably try at some point.

Now I need to figure out the rack situation. I'm thinking I will build a 15V section of eurorack and build some kind of back frame to mount the MIDI interface, noise, diginoise and output PCBs.

Snaper · Joined: Feb 28, 2014 Posts: 217 Location: Hungary

Awesome!!!

jbeuckm · Posted: Tue May 03, 2016 9:44 am Post subject:

I'm deciding whether to try a 12-channel MIDI-CV for this or just build another 8-channel one like the tested/working one above. This 12xDAC looks like it would work, but I still need some kind of 12-bit register to generate the triggers. Actually I guess I could just use two of these and only use the max value for the trigger outs. Then I would need THREE quad opamps to scale the accent signals and two MC14504B hex level shifters for the triggers. Kind of a monster...

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=196160928&uq=635978719106326224

jbeuckm · Posted: Sat May 07, 2016 8:05 pm Post subject:

Better idea: The outputs of the DAC for accent signals should range about 2-5V. So I could use a 1V comparator on each accent signal to generate the triggers. Six quad opamps on the board Shocked

elmegil · Posted: Sat May 07, 2016 9:14 pm Post subject:

jbeuckm · Posted: Sun May 08, 2016 11:33 am Post subject:

@elmegil Yes, I should explain. The DAC operates at 5V max. So the accent pulse has to be scaled x2.8 up to 4-14V. For each DAC output, I'll use an opamp to scale the accent and another opamp as a comparator to generate a trigger pulse. For the min (4V) accent, the pulse will hit about 1.4V. So a comparator with threshold around 1V (maybe .7V) should produce a hi pulse at its output - the trigger. This should be slightly faster than generating the trigger separately in software.

That's two opamps for each of 12 channels. I would be interested in any ideas for using fewer parts. I try to keep PCBs single sided and within the free Eagle footprint. One space-saving idea I had is to use the same voltage divider (two resistors) to feed all the comparators.

elmegil · Posted: Sun May 08, 2016 12:26 pm Post subject:

Multiplex your output.

Set the voltage, and tell a sample & hold (I like the LF358) to grab the voltage, then move to the next channel. One DAC, one scaling buffer, 12 small 8 pin chips and some assorted support components.

That's how I'd do it anyway Smile