TUNE troubleshooting

[epijdemic]

Hi. I need some troubleshooting references for TUNE (which is hard enough without Eagle files).

When i try to connect it via USBasp the port shuts down, like in a polarity issue. I do not have a short between VCC and GND, DMM measures 180ohms in continuity mode but that does not mean anything. Except mine is way off of successful builds circuit measurements.

Anyhow.. i cannot find the problem. Its the second Atmega already , because i suspected a fried chip. Second one was even flashed with the TUNE firmware before i swapped it off an old Nano. Still nothing. LEDs keep dark, not even a blink. Connecting 12V also does nothing.

how do i go from here?

[epijdemic]

UPDATE: I resoldered the first Atmega back onto a Nano to check if it was actually dead, but heck no, it's fine. I can connect to it and flash it. So the fault is not with the Atmega chip but something else.

Image

[pjbulls]

Were you powering the module via another way than just the programmer? These can provide a little bit of power, normally enough for the MCU, but maybe not for an extra 12 LEDs. Depending on how the shift registers turn on that might've put quite a bit of additional load on the programmer causing it to shut down. Don't think I've had that with an AVR ISP but the STM ones definitely can't provide much current.

I usually program the MCU before I join my boards together or even fully populate them to avoid these kinds of issues, another nice one is decoupling capacitors on the serial lines that make it impossible to get a good signal across.

[epijdemic]

i tried to flash the atmega before joining and had the same problem. programmer went dark and shut down the usb port. i also tried powering the board with 12v and then connect to the programmer (although it seemed silly dangerous with an obvious fault) but still the same result.

the current atmega is already flashed with TUNE firmware (flashed it on a devboard and swapped the mcu over) but the module still does nothing on regular eurorack power.

i had a faulty trace on another antumbra module in the past (SMOG) that caused me troubleshooting for weeks until i found the issue, so i am bit nervous about this one too. the SMOG i could at least troubleshoot with the mutable schematics and eagle files, but here.. its a flight in the dark, thats why i would need some common references (i.e voltages on certain pins)

[pjbulls]

Are you sure you don't have a (weak) short somewhere? Your soldering and orientation looks good but 180 ohms between +5V and ground is not a lot (~30 mA idle?) and the fact that the module does nothing with the firmware definitely on there isn't good news.

The ICSP header on an AVR has a standard pinout, so I think that you should see the following connections (with no shorts between them):

Some of these (power=2 and ground=6) already look OK visually.

[murtsi]

Hi All,
here is my story with Tune build. I had same symptoms, added with U4 getting really hot. Measured voltage in output of U4, only app. 800mV.
Then lifted power pins of IC:s (starting with cheapest, U5 and voilà, power line back to 5V. Then googling datasheets and there it were: MCP6001 is manufactured three versions: MCP6001, MCP6001R and MCP6001U. All same package but different pinout. I accidentally bought the R- version (with reversed Vdd and Vss) because of this f****ing chip shortage, thought that R means Reel as it is in many typemarkings (at least Mouser uses it). Mouser had that time none of 6001 chips, only R versions. Fortunately had ordered the 6001 versions from Tayda for next projects, dropped it in and programming succeeded, leds are blinking and I'm happy again!
The resolution of photos above (and my display) is not very good, but seems that epijdemic might have same problem. According to datasheet (if I understand it right) the 2 first markings on chip tells the type of pinout (AA and CD are "normal" 6001, AD and CE are 6001R).
In my build the first U5 was stamped CECV, the new U5 is AA89. On my display epijdemic's build might have marking ADRT, looked with my 64 years old eyes with old monitor...
The teaching of this story to me: ALWAYS check the datasheets carefully, when the exact BOM part is not found!

[pjbulls]

MCP6001 is manufactured three versions: MCP6001, MCP6001R and MCP6001U. All same package but different pinout.
[...]
According to datasheet (if I understand it right) the 2 first markings on chip tells the type of pinout (AA and CD are "normal" 6001, AD and CE are 6001R).

Okay, wow, I didn't even know that was a possibility. In the picture it looks like U5 is marked 'ADRT', so that may indeed be an MCP6001R. Nice find, & always keep checking your datasheets!

[ashley27]

Hey guys.

I'm building a Tune as we speak and am about to solder in the MCP6001. The one I have here is a MCP6001UT. Will the pinout be correct for this build?

Thanks!

[pjbulls]

No, that won't work unfortunately. U is a different pinout from the regular MCP6001, you'll need one of those. Don't get an R version either, they have the power supply pins swapped. (the T means it's delivered on tape & reel, that doesn't matter)

[funkytransport]

My Tune is out of tune ?! at middle C it's way off.. ive tried both 440 and 432 modes and both are not as they should be. it's exactly 40 cents sharp... I'm guessing its a clock speed issue? whats the best way to fix this?

thanks for any help?

Iain

EDIT - fixed - it was using the internal clock on the Atmel chip. Changed the fuse settings and now works fine.