Search results

Thanks for your reply. Approximately how long should it take after releasing the button? Edit: I've got it now. It wasn't working with the 10000mah USB battery is was using. Solid red light remains for about 45 seconds which reiitializing the flash, the the orange LED starts Blinking.

I've had Teensy 4.0 since it was introduced, recently got a 4.1. I never got around to trying the Automatic Recovery feature. I've tried it a few times, and haven't gotten it to start running Blink. I hold the button for 17 seconds, then release. How long should it take to reinitialize and...

By itself, 10 ohms would limit the current to 500ma, assuming 5v supply. But it is part of a pi filter to condition the power supply. The $3 PJRC kit takes care of this. Great value.

Assuming the sparse PT8211 datasheet applies - are you using the recommended resistor and caps for the power supply?

I've only used PT8211 on Teensy 3.x and 4.0, always 3.3V - but these questions come to mind 'they keep burning up' - smoke? or just stop working? What's connected to the output of the PT8211? Does it make a difference? Do they 'burn up' immediately, or after a period of time?

Fragmentation is a different issue from imbalanced new/delete and malloc/free. And dynamic memory is handled by the library, not GCC itself. Yes, fragmentation can be an issue, but sloppy programming causing memory leaks is more common. But my point was more about the problems using...

I wonder if it’s really necessary to create all the AudioConnection objects that way? I think you could create an array of pointers: AudioConnection *cables[16]; // 16, for example And create the connections as needed using ‘new’. My interest in Teensy is mostly about building synth devices...

This might seem pedantic; I prefer the term 'precise'. :) @neurofun suggested the following, to create an array of 4 objects of type AudioSynthwaveform AudioSynthWaveform* waveform = new AudioSynthWaveform[4]; Yes, it creates an array, and the constructors are called. But it creates an...

I’m suspect your array initializers made copies of the AudioSynthWaveform objects. I’d try using pointers, or maybe references. I just checked - arrays of references aren't allowed in C++. So, pointers it is.

Raspberry Pi 3 was also a 64 bit CPU, like raspberry pi 4. Raspbian OS is still 32 bit, I think.

Update: I examined my EX-P a bit more closely, and found a schematic online (which only shows the circuit in the OTHER position). So - with the limit pot at minimum, it's effectively out of circuit. It's the 1K resistor on the wiper which causes the values to never reach 0.

re: the M-Audio EX-P The article linked below mentions: Anyway, for this application, just leave the switch in the OTHER position, and set the limit knob to the full anti-clockwise position to yield the full range. The resistance ring-to-sleeve varies from about 1K to about 11k as the pedal...

Of course, the purple boards sound better.

Funny! I almost wrote exactly that a little while ago!

It doesn't seem like a reasonable limit, to me. What DAC is being used?

Yesterday, I received 6 of the untested PT8211/T4 boards Paul linked in post #2, and a few days ago, got 6 PT8211 kits from PJRC. Today, I assembled two PT8211 for T4. Both work fine for me.

Understood - but I still wonder if the original program behaves differently on these sequences.

It’s been years since I had to dig deep into I2C, and then it was work, so I used a Beagle protocol analyzer - very handy. But I wonder, have you monitored the same transaction using kito’s original program, to see how it decodes the sequences in question?

Yes, but to do USB host correctly and supply DC, you need something like tpd3s014, as used on Teensy 3.6. Teensy 3.6 works fine as host for my Arturia Keystep.

How to invert the pedal polarity: analogRead returns 0-1023 , right? uninverted = analogRead(A10) / 8; inverted = (1023- analogRead(A10)) / 8;

Are you referring to the D+ and D- pads on the underside of the board? Those are for the USB HOST port, not connected to the micro-USB connector.

Unrelated to LED problem, but I’d use two local variables to eliminate unnecessary calls to analogRead().

I wonder why you chose Teensy 4.0, instead of a MCU that has Bluetooth support? I think I'd try an ESP32 first. What am I missing?

Comments are good. Comments that state the obvious just add clutter. It's a fine line. The loop in your NoteOff() function should terminate on first match: // Function named NoteOff, receives 3 variables that can be use in the function. MIDI channel, MIDI note, and note velocity void...

Some controllers have encoders. Example: Arturia Minilab. There are tricks to writing code for encoders to change the increment based on the speed at which you turn. Finally, if the goal was as described in the original post, I could see making a dedicated MIDI controller for a specific...

Thanks for responding - I'll try some newer code.

I’m familiar with OSHpark , I bought a purple T4 from them, but never any bare boards. Mail is fast from Oregon to Northern California.

Latest update... My first attempt at MicroDexed with T4 (using PT8211) worked great, except for occasional glitches in the audio, which I suspect were caused by my sloppy wiring between T4 and PT8211; there are warnings about avoiding wires for this... So, I decided to rebuild, using a Teensy...

Interesting, that detail isn’t on the OSH Park page for the item, and makes a big difference. I've already built several of the 3.x PT8211 boards, I wish I'd left one unassembled. But that's OK, I need to order some other stuff from PJRC anyway.

I haven't tried it, but the PJRC documentation for SPI suggests using transactional configuration: https://www.pjrc.com/teensy/td_libs_SPI.html I'm sure it's also possible to use one SPI bus for each Mode, but seems unnecessary.

Thanks Paul! I just wish it was cheap, like the ones on your site.

Is there any plan to produce a board for this?

Interesting, I never investigated extending the GUI.

Wcalvert: in your stuff, have you come up with something better than cascaded 4channel mixers to handle many voices?

Strictly speaking, a double is a floating point number. But in C and C++, a float is a lower precision FP number than a double.

I’m thinking out loud here... Wouldn’t it be clearer if you wrote a class which encapsulates a voiceless ? oscillators,, amps, filters, envelopes? While the setups produced by the online tool are very handy, for something like this, it seems to me some basic C++ stuff could really clean things up.

Got it. I completely missed TX6 and RX6 when I looked at the board picture on my phone...

How are the pins used? Software serial asynchronous? Something else?

Can you be specific? What program ran and displayed on serial monitor? What program didn’t work?

I realized that after reading #10, then looking at the code. Well, AudioAmplifier transmits nothing if mult==0. But, AudioMixer4 will call transmit() if there is input on any channel, regardless of multiplier.

I don't expect that anything you can do with the level controls on a mixer object will affect CPU utilization of anything before or after. Maybe I'll test it to satisfy my curiosity.

I haven’t looked at audio library source code, but I would be surprised if turning down a mixer before a reverb has any effect on the reverbs cpu utilization .

"what would be the best way to bypass the reverb or the "noteFreq" module to reduce CPU usage? " Does bypass using a mixer reduce CPU usage?

8 channels 12 bit (two bytes, unless data is packed) 10khz sample rate 8*2*10k=160k Definitely agree, it’s worth some easy testing.

I just want to point out that in the original post, you mentioned storing 12-bit data sampled at 10Khz, and now in post #4 mention that it is 8 channels.

I googled ‘Arduino key matrix example’ , and the top hit was an article which explains it clearly: https://www.baldengineer.com/arduino-keyboard-matrix-tutorial.html

For that you’ll need to test if voice is active, but in the release phase.

There are certainly variations possible in a note stealing scheme: Oldest note, oldest note in release phase, same note # as new note... maybe others. But they’ll probably all need oldest note as a last resort. So, you’d need to keep track of when each note came in. Maybe somebody else will...

You will need to implement voice stealing in your onNoteOn function. If an idleVoice is not found for the incoming note, find the oldest note that is playing, shut it off, give the newly freed voice to the incoming note.

Pots always have limited turns, most are less than 360 degrees, and some are 10 turn and more, for special applications requiring fine adjustment. Digital pots is a misnomer sometimes used to refer to rotary encoders. If you really really want endless turns, you do need rotary encoders. I...