| Electronics > Repair |
| [Beginner] Troubleshooting my XMOS audio woes |
| (1/2) > >> |
| rthorntn:
Update: I've taken the minidsp apart the chip is the XMOS 16L7C10, the ifi is a XMOS 8U6C5. Update: I should add I have a DMM, oscilloscopes (Rigol digital and Tek analogue) and some other bits and bobs, my scope skills are amateur. Hi, This is the sequel to the hugely unsuccessful https://www.eevblog.com/forum/beginners/borked-usb-connector-advice/. The ifi went into the too hard basket. And then a few days ago I start a multichannel audio project and grab my MiniDSP U-DAC8, pretty lightly used and been in storage for a couple of years, exactly the same issue as the ifi, I swear I'm cursed. In addition to evaluating why I have so much stuff in my spares bin, I'm also starting to think this culprit is XMOS. The XMOS chips in the ifi and minidsp are the same vintage circa 2014-2016. I just can't see how this is a driver issue, Mac, Windows and Linux all show the same behaviour, the USB device is just not detected. I swear I saw the minidsp show up once on the mac yesterday and it worked on the mac when I first tried it a few days ago. I can remember the ifi was (very) occasionally detected as well. The times when I get detections seem to be on the first connect after the DAC has been unplugged and dormant for days and after that connect, if I unplug it and plug it in again I get nothing. I've tried lots of different cables. The ifi works fine through SPDIF. I'm starting to think that it could be something like flash degradation on the XMOS, a bad "batch" of capacitors, electromigration, planned obsolescence or solar radiation for crying out loud. I'm thinking of going deep on this one, the second DAC broke the camels back. Where to start, these things are USB 2.0 and I'm happy to spend couple of hundred bucks to get bits that will help me troubleshoot this because it will be a learning experience? Or the XMOS DACs could be fine and I'm doing something wrong but I've never come across anything like that, something I'm using could be killing them perhaps? Please help, I've just learned bits and bobs about electronics using the internet, my background is IT. Thanks! |
| shabaz:
Those microcontrollers are still available from Digikey (that generation of XMOS parts are more than a decade old) but there's a good chance that you wouldn't be able to swap it out, because, although there's no Flash in the XMOS package (it will be on a separate chip), there is one-time programmable memory on the XMOS part and you likely wouldn't be able to read it to copy out. In short, if you suspect microcontroller failure, then the device wouldn't usually be repairable. I don't know what ESD protection is on the board for the USB interface. It's expected for that to be present when using USB with most microcontrollers. Issues with that (could even be a design issue by the hardware design engineers for that audio product) might have caused the symptom that you observed. You could also probe signals to see if the processor appears to be operating (i.e. to confirm that it has booted something at least). The XMOS chip doesn't contain a DAC, so there should be a (probably) I2S interface to probe too. And probably an I2C bus for chip config for other ICs on the board. All those would be candidate pins to probe, to see if code appears to be executing. Also, all the voltage rails on the board should be checked of course. The typical tools you'd need are 'scope, multimeter, and perhaps a hot air tool if you need to replace parts. If you're only recently finding several of these devices failing, there's also a good chance that it could be a system type of issue, e.g. something zapping the board when plugging in etc., that seems a higher probability if you've experienced more than once in a short period. I'm speculating though. |
| rthorntn:
update: attached an image, any ideas on the undocumented 8-pin and 2-pin headers, they aren't mentioned in the manual, maybe a debug or programming interface? Thanks, I've disassembled the minidsp completely, this is the USB to DAC interface, it looks identical: https://www.minidsp.com/products/usb-audio-interface/usbstreamer I downloaded the manual to take a look at it. The usbstreamer doesn't get detected when connected to my mac on it's own (DAC disconnected). I suppose worst case I could go down the path of trying to shoehorn a new mchstreamer in there, I've no idea whether loading the usbstreamer firmware on to the XMOS U11692C20 on the mch* would work. Does the USB signal go through an intermediate IC before it goes into the XMOS chip, there are tiny (5-6mm?) 24-pin & 16-pin ICs on there? Real newbie question inbound...how to I even measure points on such a small board, I worry about shorting things out, jigs? I just posted on the minidsp forums asking a bunch of questions. |
| shabaz:
It looks different from what I expected. I thought that the USB connection would have gone straight into the XMOS chip, with some ESD protection placed physically in between. However, in the photo, there's another IC (blue arrow), which looks like an external USB PHY chip, and a red item that might be a choke. If anything has been zapped by ESD, there's a chance it could be that blue IC, although it has some internal ESD protection. It will be hard to probe the pins there to confirm the operation of that IC because it's a tiny QFN part (ideally needs special tiny probes hooked up to 'scope; an example is called Sensepeek PCBite), or sometimes people use ultra-fine wires tacked on with solder. I think the best you can hope for is to check the power supplies on the board (e.g. probe the pins on the connectors since that's easier, to see if any are at (say) 3.3V. Hopefully, the 0V connection is easy to tell, maybe from the underside by looking for a ground plane, or there's a chance the metal shell is directly connected to 0V. That will confirm that the ICs are likely getting power. Other than that, if the USB connection is not getting detected (and if you've confirmed it's not a PC or USB cable issue, of course), then there's a chance it might be a fault with the blue-arrow IC or the firmware for the microcontroller (the IC with the green arrow is possibly a memory chip containing that), or the microcontroller itself could be faulty. I don't see a solution if the memory chip or the microcontroller is faulty. I don't know anything about these boards to know if different firmware would work (and it would need some tools to reprogram the memory chip anyway; for sure, something like an Arduino might be repurposable to act as a programmer, but that doesn't address all the other issues with attempting to put on firmware in this way (the firmware might be protected for instance). |
| rthorntn:
Thanks! I absolutely love my PCBite for "workholding", I've never used the two probes that came with it! I just checked and found that I backed the Kickstarter in 2017 using Swedish Krona. |
| Navigation |
| Message Index |
| Next page |