Add more CV input/outputs to Patch?

jbeda · October 12, 2020, 4:12pm

Hi all,

I just got my Daisy Patch and I’m starting to brainstorm what I want to do with it. Looking at perhaps porting some other open source eurorack modules to the Daisy.

One thing that I’m hitting is that the Patch is a bit short on CV inputs/outputs (and gates/triggers).

I’m wondering if it would be possible to expand the patch with some more ports. I know that the headers on the patch have more female headers for expanding. But I’m having a hard time seeing any free pins that could be used. If there were, I’d be looking at finding an IO expander chip and address it via SPI or I2C. I’m kind of a noob here so this would be a big project for me.

Are there any pointers/examples/advice here? Has anyone done anything to expand the various daisy systems using those extra female headers?

Thank you!

antisvin · October 12, 2020, 4:30pm

Those “extra” headers are just parallel connections to Daisy pins. You could use it to replace some of the module’s peripherals, but that means that you’ll be loosing some of its functionality. IIRC, every pin on Daisy was used for something on Patch.

If you want to control more parameters without sacrificing anything, you should probably consider using MIDI.

Jaffasplaffa · October 12, 2020, 4:54pm

If there was a button, you could make a page system, where you can push one button, to switch between page 1 and page 2, ehich would give you twice as many pots.

Does the encoder have a push function?

jbeda · October 12, 2020, 5:08pm

If I try and use those pins will the existing connections interfere?

It looks like the MIDI pins can also be used for I2C so that is a possibility. Those are all passive components so likely there won’t be any interference from them as long as there is nothing plugged into those jacks.

jbeda · October 12, 2020, 5:09pm

Hi @Jaffasplaffa – not looking for more pots, really but instead looking for more 3.5mm inputs and outputs that are DC coupled for CV (and gate/trigger).

Jaffasplaffa · October 12, 2020, 5:10pm

Ah yes, sorry…

erwincoumans · October 14, 2020, 2:33am

Yes, you can use the encoder and push button to switch functionality, I plan on doing this for my Daisy Path granular synth/sample player/cloud reverb.

ryan_pwm · October 3, 2021, 10:28pm

Could you just bypass the dc filtering caps on the daisy patch? I’m not sure if making something DC coupled is as simple as that. Or maybe they are on the daisy seed itself?

BHAudio · October 30, 2021, 3:12pm

Hi @jbeda ,

I am currently creating a custom piece of hardware for a EuroRack setup. What I am looking at is creating more Gate/CV outputs by using this chip in my design:

DAC7578SPWR
Datasheet

Please take this suggestion with caution as I am only in the early phases of board layout.

Thanks,
Brett

tele_player · October 31, 2021, 2:06am

If I’m not mistaken, that’s a very expensive part. Is the expense justified?

recursinging · October 31, 2021, 7:38am

If your looking for Multi-Channel DAC, consider the MAX11300. I have a driver in the works for libDaisy.:

github.com/electro-smith/libDaisy

Add a driver implementation for the MAX11300

electro-smith:master ← recursinging:MAX11300

opened 08:58AM - 13 Oct 21 UTC

recursinging

+2091 -0

This is a very libDaisy and Eurorack oriented driver for the [MAX11300](https://…datasheets.maximintegrated.com/en/ds/MAX11300.pdf) DAC/ADC/GPIO device. ### MAX11300 > The MAX11300 integrates a PIXI™, 12-bit, multichannel, analog-to-digital converter (ADC) and a 12-bit, multichannel, buffered digital-to-analog converter (DAC) in a single integrated circuit. This device offers 20 mixed-signal high voltage, bipolar ports, which are configurable as an ADC analog input, a DAC analog output, a general-purpose input (GPI), a general-purpose output (GPO), or an analog switch terminal. The MAX11300 is an interesting device for Eurorack applications not only because of its configurability, but also because of the voltage ranges it can support. It is possible to power the MAX11300 from the +12V and -12V Eurorack standard power rails, which allows the DAC and ADC to operate at Eurorack voltage levels, without any external scaling circuitry. This essentially means one could hook up the 20 port pins of the MAX11300 directly (or better perhaps through a small value resistor) to the 3.5mm jacks of a Eurorack module. This flexibility comes with some limitations. The configurable voltage ranges are as follows: * 0V to +10V * -5V to +5V * -10V to 0V ...all of these ranges span 10 volts. At 12 bits of DAC and ADC resolution, the accuracy is probably not good enough for V/Oct use (Although the MAX11300 based [Polyend Poly2 module](https://polyend.com/manuals/poly2-manual/#inputs-outputs) claims better than 5mV accuracy through calibration). Another limitation of the device DAC is the settling time. The datasheet recommends allowing for 40us between DAC updates. In practice I discovered that this time is cumulative per configured DAC pin. That means the more DAC pins that are configured, the longer the wait between updates, which reduces the time resolution considerably. The update rate is fine for LFOs and envelopes, but not for audio rate (>100Hz) modulation. ### API The device can be attached to libDaisy hardware via one of the SPI peripherals. By default the MAX11300 will use `SPI_1` and its standard pin configuration, but this can be changed by providing your own `SpiHandle::Config`. ```c++ MAX11300 max11300; MAX11300::Config max11300_config; max11300_config.Defaults(); max11300.Init(max11300_config); ``` The API of this driver was designed to provide intuitive access to the features most relevant to Eurorack. This includes configuring and re-configuring pins at run-time. Pins can be configured in four modes: * DigitalRead * DigitalWrite * AnalogRead * AnalogWrite When configuring pins to `DigitalRead` or `DigitalWrite` modes, then an additional configuration `threshold` allows you to set the voltage threshold/output of boolean true. For example, `8.0f;` would configure a `DigitalWrite` pin to switch between 0V (false) and 8V (true), whereas `2.5f;` would configure a `DigitalRead` pin to return boolean true when the voltage detected at that pin is higher than 2.5V. ```c++ max11300.ConfigurePinAsDigitalWrite(MAX11300::PIN_12, 8.0f); max11300.ConfigurePinAsDigitalRead(MAX11300::PIN_8, 2.5f); ``` When configured as `DigitalRead` or `DigitalWrite`, pins can be written to and read from as boolean values. ```c++ max11300.WriteDigitalPin(MAX11300::PIN_12, true); bool value = max11300.ReadDigitalPin(MAX11300::PIN_8); ``` When configuring pins as `AnalogRead`, or `AnalogWrite`, a `VoltageRange` is included to configure what voltage range the pin will read from, or write to. ```cpp max11300.ConfigurePinAsAnalogRead(MAX11300::PIN_14, MAX11300::VoltageRange::ZERO_TO_10); max11300.ConfigurePinAsAnalogWrite(MAX11300::PIN_5, MAX11300::VoltageRange::NEGATIVE_5_TO_5); ``` When configured as `AnalogRead`, or `AnalogWrite`, pins can be written to and read from using the raw 12 bit codes, or a by voltage, which is then clamped and scaled according to the voltage range of the configured pin. ```cpp max11300.WriteAnalogPinVolts(MAX11300::PIN_5, -3.25f); ``` All pin `Read*` and `Write*` method operate locally on memory locations that need to be synchronized with the device. This is done by invoking the `Update()` method at regular intervals. ```cpp max11300.Update(); ``` ### Implementation details After receiving some helpful advice, I decided to make the driver implementation templated with an abstract "transport". This allowed for more thorough unit testing, as well as support for a DMA based transfers in the future. The (quasi) full-duplex nature of communication with this device required extending the API of the `SpiHandle` class in libDaisy. I added a `BlockingTransmitAndReceive` method to support sending and receiving data in the context of a single transaction (CS/NSS low). The impact of this addition should be low, as it is simply a wrapper for the underlying HAL pimpl method of the same name. I discovered that if you try and update DAC configured pins without respecting the datasheet recommended 40us settling time, the MAX11300 will simply lock up until it is power cycled again. In order to prevent this, I added a `wait_us` parameter to the transport `Transmit` interface. This allows the driver to communicate the max update frequency to the transport implementation - which will be necessary when implementing a DMA based transport. ### Out of scope The driver transport is currently using blocking methods of the `SpiHandle` class. This is not optimal, and the nature of the driver API should make a DMA based transport possible. There are a couple of things to be aware of when tackling this. * The `SpiHandle` class does not yet have the DMA equivalent to `BlockingTransmitAndReceive` * The TX and RX buffers are not in the `DMA_BUFFER_MEM_SECTION`, nor does the API allow them to be passed in. This means the DMA TX and RX method impls will need to do cache flushing and invalidation. * The `wait_us` parameter passed from the driver will need to be respected when scheduling DMA transactions. ### Conclusion I hope the addition of this driver will open up a number of new options for people designing Eurorack modules on the Daisy platform. I welcome any and all critique and contributions!

The device has 20 channels which can be freely assigned as DAC, ADC, GPI or GPO - All at Eurorack voltage levels.

Manysounds · November 1, 2021, 6:25am

O my yes this is a wonderful idea

On a related note, what do we think the best solution would be to get 16 channels of 48k 16bit audio into a Daisy, and could it even handle such a stream? I don’t really know what the real time I/O limits are but my brain says “build a multichannel mixer”.

BHAudio · November 2, 2021, 2:39am

SWEET! I’ll check it out right now.

Thank you for the info, this is great.