MarkT

Either use I2S MEMS microphones or you'll need a couple of microphone preamps...

The neopixel chips define the output current, so the only way to get less bright ones is to find less efficient LEDs, not easy these days, and I'm not sure you can get the neopixel chips separately from an RGB LED package.

I presume the fader uses PWM output (analogWrite), which is not compatible with digitalWrite on the Teensy if I remember right - on the ARM digitalWrite has the pin mapped to GPIO, analogWrite has it mapped to some timer internally.

The zener will blow as there is no current limiting before it. Perhaps split R1 into two parts, one before the zener, one after? Measuring after the rectifier will mean there is some lost voltage compared to the actual output of the dynamo...

Hyphen in python is an operator, namely subtraction. So D_n, O_n are what you need, not D-24 or O-24

Use the line outputs if connecting to anything other than headphones/earbuds. BTW I have measured some artifacts on the SGTL5000 microphone inputs, IIRC something like a low level tone around 13kHz or some such - plotting the spectrum is a good...

Well lookee here ... it's yet another Teensy-based synth ... or might be, if I ever finish it. But it says "Blog" in the category heading, so this may serve to stimulate discussion if nothing else, and if I'm moved to do so I can report progress...

You need to test for 2 bytes being available: if (mySerial.available() >= 2)

BTW that conditioning circuit looks wrong to me, asymmetrical. Change the 2k2 to 10k as well, then the signal will be mid-rail for the ADC.

You can't drive a speaker direct from a line-level audio signal, speakers are 8 ohm, line level is 50k or 100k... (Or do you mean a powered speaker?) And you certainly shouldn't try to drive a speaker direct from a logic pin, that can blow up...

You have at least one while loop inside a critical section (i.e. with interrupts turned off). You should be turning off interrupts briefly for critical section stuff only (a few microseconds). Also that while loop is waiting for a condition...

Yes a (poly)fuse would be a good idea rather than just a zener diode.

Good to hear its all sorted for you now...

A "linear" optocoupler has pretty poor linearity in fact (from an audio perspective), that particular one has a 2mA to 10mA figure of 0.25% which is -52dB, whereas 16 bit audio can have 98dB SINAD and you'd expect the signal path to not throw...

Every board with high pin-count BGA parts is at least 4 layer, though usually 6 or 8 or more - its the only way to route all those pins from the middle out!

All logic families require fast inputs unless they are Schmitt-trigger. Most often used is the 74HC14 which is a hex-inverter with Schmitt-trigger inputs. If you don't need inversion you simply chain 2 stages together. I'd guess the 74xx14 are...

Piezo sensors have a high source impedance, which might cause issues, but it looks like this one has built-in buffer amplifier as it requires power. So yes I'd try it as you describe.

Very slow edges can generate 100's or 1000's of edges in some cases, many logic inputs are not designed to handle slow inputs (a CMOS gate acts as a high gain analog amplifier if inputs are near the mid-rail voltage).

100k are to stop thumps with plugging in, i.e. draining the capacitor charge. Its always good design practice in audio to stop such thumps as they can break loudspeakers in the worse scenario! I2C pullups are usually 4k7 or 10k unless the...

5 layer I've not heard of, usually come in pairs, for dense SMT with BGA parts 6 or 8 layers is more likely.

Yes the raw signal from the sensor's needed Schmitt-trigger inputs to handle the very slow rise times (for CMOS logic signals anything over 30ns rise time is "slow", and over 100ns rise time may completely misbehave).

Simply poll the input at 1kHz or so, that'll hide any bouncing on a 15µs timescale completely, but be fast enough to handle 30Hz with ease. I think the scope showed bounce, not debounce(!)

Not sure why you need magnetics for a simple Teensy-to-teensy link - or are they a long way apart and need galvanic isolation? BTW m is milli, M is mega. 95mbits/s isn't so impressive :)

The Teensy 4 controller is only 1.8V internally I think, 3.3V is its I/O voltage - put another way its a 1.8V processor that's 3.3V tolerant! This is why 5V can easily destroy the whole chip.

High performance steppers run at 80V or 120V, go much faster as its inductance and back-EMF that limit speed for a stepper, so increasing voltage counters both of these effects. 5000rpm is still very high for a stepper, but 3000 isn't by all...

And any 74LS series chip will risk being fried if the supply voltage peaks above 7V... Definitely use 5V for any TTL. 3.3V for Teensy 4 means you can't directly interface Teensy and TTL.

Line level isn't designed to drive 600 ohm loads (no one uses 600 ohm in audio these days) but may be able to (check SGTL5000 datasheet for this). Balanced signal cables are used professionally to avoid any ground loop issues, but with...

I don't understand the use of D1 which seems to protect C6 rather than U1. Also why is there protection on the tip but not the ring of the jack?

In the loop in main that calls loop() there is a call to yield()... In the Arduino this does serial event handling, I presume Teensy framework does similar in yield?

The MISO line is sampled by the SPI master only on the relevant clock edges - the master doesn't do anything else with it as its never driven by the master. Its the slave SPI devices that have to be clear about MISO handling.

It has 2 I2S busses, not 2 I2S pins! For 8 separate channels I think you need to use TDM output to an 8-way DAC and 8 amplifiers for the speakers. Check the tdm output class in the Audio Tool: https://www.pjrc.com/teensy/gui/?info=AudioOutputTDM

But you didn't keep loop() empty, you put a delay into it which will stall the event handling. Why not simply use: void loop() { }

You can also use a low-dropout regulator down from 5V to 3.3V and save a more expensive DC-DC converter. The TracoPower devices are good, some have MTBF measured in 10's of millions of hours!

The FFT requires all the data up-front before it can start. But you don't have to do all the FFT computation in one go once you have that data, you could do a few stages at a time on each update, but that would increase latency. A 1024 point...

The design tool is implemented in the gui/index.html file in the Audio library - when you add a class to the library you add entries in the gui to match...

Volatile is required whenever a variable may change in a way the compiler cannot see (interrupts and hardware registers being the classic examples). If you don't turn on any optimizations it might not matter, but your code behaviour is strictly...

Doesn't look cut to me. The solder-mask covers the trace at that point.

The standard audio adapter boards are too low impedance (30k) for guitar pickups (which want more like 500k or 1M), meaning the higher frequencies will be attenuated. I know some people have made guitar-capable audio boards for the Teensy, try...

Handy tip: connect to opposite corners of the push-button and you don't have to remember which pins are common...

About 0.7µs, not 0.5µs, with standard cables and dielectrics - your calculation works for air or vacuum :) Just seeing the waveform on a 'scope for a live system will say a lot as unwanted reflections would show up.

Sounds like you want a Time-domain Reflectometer?

If only the USB standard had thought about this issue and figured out a different icon or something for the plugs on such a cable... Probably the correct answer is to cut any power-only cables in two whenever you receive one, since as I...

I see, thanks for explaining :). I confess I simply switched to using a different microcontroller board for this functionality, rather than delve deeply into T4 machine registers...

You could check the cable end-to-end with out of use, or you could check than an RDM capable device is responding from the far end if in use. DMX512 uses RS-485 signalling levels into 120 ohm termination. https://en.wikipedia.org/wiki/RS-485 I...

There are 4 SPI modes, 2 sample on clock rising edge, 2 sample on clock falling edge.

For a digital signal you have to poll at greater than 1 / minimum_pulse_width. So if the minimum pulse width (positive or negative) is 100µs, you must poll faster than 10kHz to see every transition, or simply use interrupt driven input. A 100Hz...

Can you post an _accurate_ schematic - the one you've provided shows no ground connection the Teensy, the photoresistor is an antenna hanging off one wire, its either incorrect and thus we can't figure out what your circuit is, or its correct and...

Actually you _can_ use those low profile IC sockets (as pictured in post #2) for 0.65mm(*) Dupont header pins, but it will permanently stretch them and isn't good practice. (*) definitely not 0.064" as stated above!!

If you look at the Audio library docs here: https://www.pjrc.com/teensy/gui/?info=AudioInputAnalog you'll see the AudioInputAnalog class uses pin A2 on T4 hardware

Ah, I see - ambiguous in that context!