[Need Help with SPI] Euxoloti for axoloti (Eurorack Module)

modular
#77

I wasn't aware such micro-sd extenders exist. But I expect one issue, the sdcard holder has a little switch that is used to detect card insertion/removal in firmware. The card will be mounted on startup without switch, but no hotswapping cards...
What's wrong with sideways mounting Axoloti Core (with DIN and power section cut off)? That 'd give you access to sdcard, usb host+device ports, bootloader switch, status leds. The jacks are probably of less use, but can easily be modded into nice extra's like footswitch or expression pedal input... Or removed to save a bit space, the MIDI part could be sandwich-mounted upside-down, the mounting holes align...

Clarifying my motivation for mentioning this: if the eurorack power supply uses a linear regulator from the 12V line to produce +5V, it will get very hot when a USB device is connected to the host port that pulls 500mA, as it will dissipate 3.5 Watt. The usb host port will still have a stmps2141 to protect against over-current.

Do not try to widen the mounting holes in the power section
#78

@gloom_and_glare
your right, the op amp clips the voltage. a extra 3,3V vref for mcp3208 vref would be useful. actual jitter of cv ins are +/-0.05 of 64. in axoloti. pots are less jittering till nothing.

@toneburst
the sd card extension will be cool, but tricky to mount at the panel. an other solution can be axoloti in card reader mode.
the max11300, which @fma mentioned, would be also a nice one for a monster expander.

the pots are 9mm alpha 10K lin. split shaft. I bought them at banzaimusic.com. But this shop is horrible. You can easily cut away one part of the shaft to get a D shaft.

I'm excited about how everyone will use and extend the project in the future.

@johannes thank you.

#79

Ah, I hadn't thought it was possible to use a split-shaft pot as a d-shaft like that.

Those pots at Banzai look a lot like the ones Synthrotek sell. I bought some for my Turing Machine, so I could use Rogan T18-fitting knobs. Worked nicely, and they have a great action. I'll consider getting then from Banzai though in future.

Assuming I'm not in a hurry, of course...

a|x

#80

I updated some files for the pcb people…
* B.O.M. + Mouser and thonk links
* Panel + Size and link to whitewall product

Please. also check the Readme at B.O.M.

edit: I also forget the knobs for the pots…

edit 2: mouser carts are online at github too…

edit 3: euxoloti AXO objects for pushbutton, led and gate functions are at github. And there is a jpg file of the panel to upload it directly to white wall service.

#81

Will White Wall make the complete panel?

Their site is a bit confusing, but they seem to just do printing onto aluminium dibond, so I guess the panels have to be drilled by hand afterwards.

I do wonder how durable they would be, too, since their products aren't really designed for this kind of use..

a|x

#82

yes, their site is a bit confusing.
WW only do the UV printing and cutting the outlines.
the holes have to be drilled by yourself. but it is really simple. the position of each hole is marked by a cross.
You can also order a metal printed plate. it's only 1mm thick but with pcb and connection behind it, it should be less bendable.
Some of the buchla people use it. search at muffwiggler diy forum for white wall.

I haven't much experience with panel services. I know there are some services for that e.g. Schaefer AG or Re:Synthesis

acrylic panels are also no problems or pcb panels.

maybe we start a poll before the final run with the corrected pcbs are ready and start a euxoloti thread at muffwiggler. so more people are interested in.

The current pcb testers are: @toneburst (you), @fma, mxmxmx (no forum member) and me.

By the way. A detailed modification overview is online at the wiki and some of the euxoloti objects are also online. you can find them in the future at the community contributions (Axoloti 1.0.7>=)

#83

Cool. I will look into panel options.

Alu dibond look a bit thick, but 1mm aluminium sheet sounds flimsy to me.

Laser-cut Acrylic may be the cheapest option.

alx

#84

Incidentally, Paul, I have some questions about CV I/o.

Are you planning to produce a V/Oct input object, to allow for pitch-accurate note CV tracking? It might need some kind of calibration option to correct for component value differences.

Are the outputs AC or DC coupled, and would they be suitable for producing accurate V/Oct output (again, possibly using a custom patcher object with some kind of calibration system)?

Apologies if some/all the above has already been answered earlier in this thread.

a|x

#85

The E-Switch illuminated switches are on backorder at Mouser, now, estimated restock 21st April.
Can't seem to find them anywhere else, unfortunately.
Do you know of any other, preferably EU suppliers?

a|x

#86

Yes, acryl panels can be done really easy and cheap at our fab lab. Maybe I will visit the Hacker Space nearby me. They have a screen printing and a CNC milling machine. It would be perfect for a little batch of panels.

Audio inputs are dc coupled like org. axoloti.
audio out isn't coupled. I haven't checked if the audio dac do it internal like the headphone amp IC.

1V/oct input should be also possible. The best cv in jack for this would be CV1 as it has a limit range by 0-5V. you can expand it by changing resistors R19(66,5K) to 47K and R20(200K) to 180K and get a -1.5V to 5.5V range like MI clouds (page2 "IC4A").

I will think about, how to realize a calibration routine (incl. safe settings on SD card)…if you have any experience or ideas, let me know.

#87

No, the audio inputs are AC coupled on Axoloti. The outputs, both headphone and line are DC coupled. The axoloti schematic shows capacitor symbols between codec and the TPA6132 but they really are resistors as indicated by the component value.

#88

@johannes does that mean V/Oct CV output isn't going to work well for Euxoloti, without an extra DAC?

a|x

#89

The audio DAC path is DC coupled, that's requirement #1 for CV out.
For V/Oct CV out, another requirement could be tuning stability. The audio DAC does not use an accurate or stable voltage reference. I'm not sure how much tuning drift to expect.

#90

I see. So an external DAC, and a stable voltage-reference would be the best way to go.

Maybe an expander with an MCP4822 or similar 12bit DAC, then.

a|x

#91

@paul Got my PCBs today. Very quick! Haven't ordered any of the components yet, so it's going to be a while before I'm up and running. Still haven't found a non-US supplier for the illuminated switches.

a|x

#92

@paul the MLCC caps https://www.mouser.co.uk/Search/ProductDetail.aspx?R=C322C102F5G5TAvirtualkey64600000virtualkey80-C322C102F5G are pretty expensive, I guess because of the 1% tolerance. Could a cheaper alternative be used instead, or is the low tolerance required?

a|x

#93

Ordered everything from DigiKey in the end. They had everything except the 14-way DIP socket (think I have 2, anyway) and MCP6004-I/P. I went with the MCP6004-E/P-ND instead, which seemed to have an identical spec. (apart from operating-temperature), as far as I could see. Hope this works.

a|x

#94

As long as it's a NP0/C0G type, it's not entirely evident why they would have to cost that much or be 1%. I'd guess it's because the Mutable Instruments BOMs specify ≤ 2%? But what's the rationale?

#95

Dunno. I ordered them, anyway. Still saved, as they bumped my order over the £50 free-shipping threshold, so I'm not complaining, really. I'll bear that in mind next time, though. I have a feeling I may be building a couple of these. I'll reserve judgement until I've built and tested this one, but the Axoloti platform seems to offer the best solution for a relatively powerful generic/programmable Euro module, at the moment, at least.

a|x

#96

@toneburst
sorry for late answering. I don't know any EU source for the pushbutton. I know, that digikey sells them, too. Can you upload your digikey shopping cart? It will be helpful for other people.

@gloom_and_glare and @toneburst
The 1nF 1% capacitors are very expensive and I haven't checked every alternatives at mouser or digikey. I can't give a competent explanation about why better as 2%.

The used circuit is a First-Order inverting low pass filter.

Bildschirmfoto 2016-02-18 um 12.30.50.png797x808 61.2 KB

(orginal source/page 16-3)

So what have read is, that real op amps produce more noise above 10kHz (german wikipedia) and supply voltage drops can influence the output voltage depending of the gained voltages. So I have to check the PSRR and CMRR of the used mcp6004:

My max cutoff frequency should be 10kHz or less…If, I'm using a capacitor with more tolerance it will be possible to exceed the 10kHz max freq..or I have to put less cutoff frequency and use a capacitor with higher tolerance.