Bum 3.6?

Status
Not open for further replies.

ETMoody3

Well-known member
I seem to have a non responsive 3.6. ( PJRC invoice # 77619)

No need to post code as the code in question works fine on another board.

The board in question blinked initially, responded to the loader, but the program never ran. The serial port never returned. ( ubuntu , arduino 1.89, TD 1.47)

I have indeed soldered pins, there are no shorts, and all joints pass continuity check.

I attached it to a windows 10 machine, ( ard 1.89, TD 1.46..." safety machine ") and *briefly* saw a serial port show up in devices, which disappeared in about a second. No repeat of this behavior.

No response to 15 second recovery, and no further enumeration as any sort of device on any coding machine, windows or linux.

The 3.3 v conversion is functional. So power is making it through the usb cable. The usb cable works with the other 3.6 running the code I attempted to flash into the questionable board.

I have tried pressing down on the chip, no avail.

There's no obvious heat while powered. Just nothing happening.

Hope there's some other spell to cast, standing by... can get chicken blood and black candles.
 
Yet another 3.6 soldered up and placed in the project works fine, still would like to know how/if/why before I shell out to replace the one that just wasted most of a day for me. 4F2F8D3D-1880-4846-84A0-318925B2B8A9.jpgD3CA6C52-295A-442F-894C-A1608E059C73.jpg
 
Nothing? Not even bad guesses?

I'm impressed. "Speechless" isn't an attribute of this forum.

My best guess at the moment is roughing up the card getting it out of the low profile sockets between test fit and flashing the program. Moving on.
 
I would recommend to resolder all of your solder joints underneath the Teensy under magnifying glasses or a stereo microscope.

I can see some partly unsoldered or looking "cold", eg. pins 17, 18, 21.

All the best,

Frank DD4WH
 
Hmm... on non responsive 3.6, reset is floating. Program reads 3.298 v

I have read what might be every marginally related post a search turns up, and this is the only test that turned up a probable failure.
 
Nothing? Not even bad guesses?

I'm impressed. "Speechless" isn't an attribute of this forum.

My best guess at the moment is roughing up the card getting it out of the low profile sockets between test fit and flashing the program. Moving on.

I started to reply with the only test that came to mind - beyond soldering issue.

Before soldering was there a series of uploads and button presses done to say if 'normal' behavior was seen from new?

The only faulty T_3.6 I saw - or reported with a 'couple' of others ( but only on early batch and not since? ) - was where tapping program button lightly to click often would not initiate the bootloader - but a firmer press 'beyond clicking' would. When the T_3.6 PCB flexed down in the center the harder press seemed to complete something floating? Beyond that or bad solder I've not seen any problem matching the description presented where here those things seemed to be 'ruled out' by the posting. Problem here was unsoldered board - once soldered it is harder to say that something else wasn't introduced.
 
The first and only time it was recognized and took an upload was after soldering, and continuity/ short testing , a step *always* taken.

Initially powered, it showed the blink program. After first upload it only showed up briefly in the windows device manager, and that only happened once.
.
I *did* bend the whole card getting it out of the socket after test fitting, but the blink and upload happened after that. I have carefully removed the next one ( pictured) from the sides, not the ends...these low profile headers have some vicious grab to them. Be warned.

I don't know what I did... I was nervous to try the last one I had on hand lest there be an issue in the assembly pictured, but that not only passed tests but actually works with the second board.

I can't rule out heat damage, ESD damage, or RFI damage, although from all I have read today the last two don't seem very likely. This isn't my first build, and I can say there's been times a project should have fried but did not. In fact, on another 3.6 project an external ram chip stopped working after a 5v linear converter got shorted on a casing, but the Teensy still works.

BTW, hoping to hear of some esoteric rescue possibilities.
Killing components attends the hobby.
Anybody ever hear of the bootloader getting clobbered by ESD?

Straws. Grasping.
 
Sorry,

I did not answer figuring, you covered most of the standard things I would suggest, like, verifying that you have 3.3v. No shorts...

Suggest the 15 second reset. Try the hold program button down while you insert the board and then release button, as to leave it in program mode, then try to download blink.

Try different USB cable?

And if all else fails, hit it very hard with hammer, then you don't have to worry about what is wrong with it.
 
Yup, tried all the standard things before posting. 15 sec, different usb cable, different computer, different TD version...( always have the last known stable version on another machine )

Finding reset pin floating is no bueno. Should be 3.3.

Percussive maintenance is always good. If I had a reflow setup I'd get it hot...but those just seem like a fine way to burn the house down.

It's already in the recycling box, adding a replacement to next budget-thing.
 
I found the odd unit here when I ordered the KS Bulk pack. Just pulled each from shipping and upload sketch and button pushed in some fashion - one acted odd but I dismissed it as user error/inattention. Then I got back to it and the behavior repeated so I spent time with it - still unsoldered/molested. Found that clicking the button would often do nothing without a firm press. Other than that issue I've not seen any bricking trend indicated on the forum in past years - especially WRT to bootloader MCU.
 
i have a suggestion on the methodology. when using an inverted teensy/low profile headers in the testing phase it's not necessary to fully seat the teensy. [or any setup/microprocessor] i generally do at most half way. i know the teensy will probably need to be removed a few times and no need to stress it. i only fully seat it after it has been verified ready. additionally, using inverted and low profile [or any profile for that matter] it would be very difficult for me to access the button if need be. at first i thought you probably had your project board cut out to provide access then i wasn't sure. in the past i have had similar experience, for whatever reason after a successful load there was nothing and windows lost it. what i did was unplug the teensy and let it sit for a bit, reboot windows, plug in teensy and press and hold the button and it came back. i have found when using teensy, particle, arduino when windows drops any of them there are times they don't come back without being unplugged, reset, and a reboot. so, i don't have a specific suggestion on how to fix this current issue but perhaps by looking at your current methods and tweaking a bit you increase success rate in future. good luck.
 
Thanks.

Not a difficult target to hit: 9040A0C6-E3EC-4B9A-AF4F-B8BE4F4F2194.jpg

I call one dead teensy out of all of them I've bought acceptable success.
 
yeah. i never have done that because always afraid of hitting wrong pad accidentally. that is too close for comfort for me, but props to those who can habitually pull it off. i have hands that can be kinda twitchy at times.
 
Status
Not open for further replies.
Back
Top