jmarsh

If that is really a concern, put waveshape and lerpshift into a struct and dynamically allocate them together, using a single pointer to access both values. Heap fragmentation is really not a concern on the T4.x boards.

Why not set up the new array before deleting the old one, so the critical moment boils down to simply assigning a new pointer to this->waveshape.

This is exactly the sort of thing I was talking about when I mentioned "acting like a victim" earlier. When this thread began I had no idea who you were, nor Limor. Adafruit to me is just a business like any other, I don't care to know who runs...

Given OP's previous behaviour, closing the thread (for any reason other than their own request) would give them another excuse to act like a victim.

Are you sure? That looks like it's just declaring a pointer, not a class object.

The irony of someone with an account less than a week old, which was specifically created to announce a competitor to PJRC's products on their own forum, complaining about new accounts being used to "dunk" on them...

hi phil! i’m one of the new people who’s here to dunk on you! just wanted to say that i had a similar experience with your closed source products as CollinK a few years back. talked to yall a few times online and in person about it but neither...

I thought the OP would have come back after they'd cooled off and asked for the thread to be removed, because it definitely doesn't reflect well on them. But it seems they've doubled down and are directing people here from their own website/blog...

BITER/CITER are the number of loops, if you increase the number of bytes per loop (NBYTES) you can increase the total amount transferred.

That's why I'm saying I don't want this change made to the T4.x, unless there's a way to make them float when they're not wanted.

https://github.com/A-Dunstan/teensy4_usbhost There's not many device drivers but if you're familiar with libusb the interface is similar.

usb_configuration is from the core, the TeensyUSBHost1 class is from my own USB host library (TeensyUSBHost1 is for the micro port, TeensyUSBHost2 is for the host pins on top of the T4.1).

Yes.

I'd hope not, or at least if they do a way is found to have the port function as both device and host like the existing ones can. I have code that does this: if (usb_configuration == 0) { static TeensyUSBHost1 usb; usb.begin(); } else...

Huh? Feels like we're talking in circles here. They were talking about the molex connector that you ended up buying. The flex ribbon connector that the pads are actually designed for is this one, which has 1.0mm pitch and is used with the...

That's exactly what is meant to go there, as used for the T4 breakout board.

The problem with including pulldown resistors is that it locks the port into acting as a device. The existing micro usb port is possible to use as either a device or host port.

How fast would be "super-fast"? I have a USB solution already that gives me 3x 24MHz 8-bit DACs (based on https://github.com/osmocom/osmo-fl2k)

That's exactly what I'm saying. The MKL could be replaced with something else and while it wouldn't be possible to upload code with Teensyloader, another util (designed to work specifically with the MKL substitute) could perform the same job and...

It doesn't do this. It's not involved in the boot sequence (besides powering up DCDC at the correct time), that is purely handled by the IMXRT ROM based on the fuse settings. If you have an alternative way to startup DCDC and a method to write to...

In case it helps, the SPI peripherals on the Teensy 4.x have a special register setting that can switch the MOSI/MISO pins (the OUTCFG field in the CFGR1 register).

I had no idea who OP was at the beginning of the thread, because they didn't make any mention of it - I suspect that was intentional, trying to keep it "under the radar" and trying to make it seem like this was a case of a small open-source...

So for the json storage does your code use StaticJsonBuffer or DynamicJsonBuffer ?

You didn't divide the timer load value by the prescaler, you're trying to load it with 2.5e6 instead of 39061.

I'm seeing something similar with "Serial + MTP" config, the serial monitor stops working after an upload and needs to be closed/re-opened several times before it starts working again.

These seem to be popping up everywhere lately: https://www.luckfox.com/EN-Luckfox-Pico-Pro Cortex A7 is practically the bottom performer of the A series but still beats anything in the M series, plus with NEON it would be several orders of...

Slip of the tongue there? Perhaps you've been banned from purchasing Teensys because you're reselling them?

It is more or less, except you're charging more.

You can't just sell a pi pico 2 under the name "freensy" and declare it to be Teensy compatible...

It could be running out of dynamic memory.

Yes, it's only used to switch the USB Host current on/off.

It depends. Are you actually going to use the USB host functionality?

That's the same thing with a set of brackets removed. -(strlen(onebuf)+1) is the same as - strlen(onebuf) - 1

Which is why the size should always take space for the nul terminator into account by subtracting 1.

Use strncat to ensure the result doesn't overflow.

Except for the fact that the pico2 (and the original pico) is a dual core processor...

If you want to guarantee the arrays are stored sequentially (one after another) in memory, they need to be inside another object like a struct or class. The compiler is free to place global objects in any order it likes. With regards to the...

I would argue that systick really doesn't need to be running at 1000Hz, 100Hz would have a lot less impact.

The way it should be done, is that systick_isr (seen here) should be comparing ARM_DWT_CYCCNT with CYCLES_PER_MILLISEC*systick_millis_count to work out what amount to add to systick_millis_count instead of just adding 1. That way it wouldn't drop...

I think you mean millisecond, which is the period of the systick interrupt - and the flash procedures can easily disable interrupts for longer than that.

I have written an example that demonstrates the slowness: #include <LittleFS.h> #include <Entropy.h> // NOTE: Entropy class defines "randomByte" and "randomWord" methods that aren't implemented! LittleFS_Program progfs; static elapsedMillis...

The AudioInputI2Sslave object expects 16-bit audio, from the capture your device seems to be 32-bit?

Supposedly there are some nightly arduino IDE builds with a "copy output" button in the serial monitor as a workaround for their broken behaviour. I haven't tested them (because I still think the Serial Monitor is pretty rubbish; why isn't it a...

The serial monitor simply isn't very good, even when it does copy it only grabs what is currently visible. Any selected text that is scrolled off-screen doesn't get copied. You're better off using something like TyCommander if you want to...

That's HFW4R (4 pin), you want HFW8R (8 pin), e.g. https://www.digikey.com/en/products/detail/amphenol-cs-fci/HFW8R-1STE1H1LF/9695136

Yeah, I was about to ask if this is the bottom side of the board because that Teensy pinout is definitely flipped.

Sparkfun is responsible for producing it: https://www.sparkfun.com/teensy-4-audio-shield-rev-d.html

Back to the topic at hand: making flash writes not block interrupts... I've replaced the latter half of eeprom.c (everything after the eeprom_write_block function) like so: #define LUT0(opcode, pads, operand) (FLEXSPI_LUT_INSTRUCTION((opcode)...

Something I did wonder about before I ditched it: LittleFS allows using a statically assigned programming buffer rather than dynamically allocating one. FlexSPI apparently allows mapping the TX FIFO to AHB memory, although I'm a bit unclear on...

I'm using the Teensy to emulate a playstation1 memory card. They hold 128KB of memory arranged as 1024x 128 byte sectors. The Teensy communicates with the console using SPI(-ish) over FlexIO and loads/stores the data to a 128KB memory array, with...