Look at the TB schematic for answers. Power from the USB-C socket (DCIN_VBUS) is used as a signal by U12 to switch the USB data connection to the XMOS between the USB-C socket and the GPIO header input. You need 5V on the USB-C for its data connection to be used. Diodes D1 and D2 prevent either ‘5V’ supply backfeeding into the other, but there’s no formal switching - whichever is the higher voltage ‘wins’. In your case the easiest option is to get a USB B connector to wire to the GPIO header USB data connection, while supplying power from your linear supply to the GPIO power pins.
Incidentally I just noticed there are no resistors on the USB-C CC1 and CC2 pins. This will probably prevent a USB C to C connection working because there’s no indication to the host that a device is connected.
I made my own connector between phone and TB. With power on pins 1 and 21 the USB +5V wire is still needed, the 0 wire is not needed. So only three of the USB wires need to be connected for USB input with external power.
Thank you!!!
I can not implement what you are saying by myself but I hope I’ll get some help to try it!
I guess that now it makes sense to me! From what I can understand!!
I was communicating with one of our Tone Board users on Facebook. He took the extreme method of powering his Tone Board using a power bank. I did not put this into the original blog post, because it seemed to me that (a) poorly-designed power bank circuits could be noisier, or (b) it would be inconvenient to use a power bank for long periods of time.
I am investigating how to power the generic tone board from a good power supply directly on the card and keep up as bridge from RPI (not through Y cable which I already manage).
I am confused on what to do to feed power. I understand it is through pin 1 and 21 as presented in this thread, here and on examples on picture here.
However, here it is also mention that a ship needs to be to de-soldered.
What is the exact method? Need to de-solder? Actually what is the purpose of this ship.
I made this circuit and it works extremely well. Its not exactly what i use (I use more filtering) but this is the basic principle. And it is powered by external linear transformer with LM 5v ultra low noise regulator.
Problem i have is whenever the inductive loads in home like fan, grinder etc turned off, khadas will disconnect from pc. I know it is very sensitive to voltage fluctuations but never thought it would be on this scale. With khadas connected directly to pc, this problem never occurred. Can Someone from khadas shed some light? Adding more caps is not solving the problem. Is there any other way around? Thanks
I finally got to the bottom of the problem (I think)
Problem was my external linear regulated PSU: When it’s “cold” and I switch it on, it takes it some time to stabilise at 5VDC. It’s probably less than 1 second, but it’s long enough to put the Tone board is some kind of wrong state in which the sound is heavily distorted.
Easy workaround for me to switch off and on the “warm” PSU quickly, just so the Tone board is reset and gets the stable 5VDC upon restart.
Is this an issue that could be addressed somehow on the Tone board?
Perhaps you could add a turn on delay to your power supply. Could be a simple R-C circuit with relay. Or search the Bay using term “turn on delay circuit”. Caution, some poorly designed circuits may add noise.