possible hardware problem???

Status
Not open for further replies.
M

Marcus LaGrone

Member
Pardon me as I can't post my code, but maybe someone can direct me where to poke the Teensy:

I have code that runs fine on a Teensy 3.6. It is running Optimised Fastest with LTO and is overclocked to 240MHz.
Everything is awesome (cue music).

However... I have three of these on the same board. all running the exact same code and on one, the I2S audio is jittering.
Same batch of Teensy 3.6, all programmed within minutes of each other yet one is always a problem.
I'm powering off of the 5V side with the 3.3V pins dangling, unconnected.
I'm showing 4.96V on the 5V rail and 3.28V on the 3.3V lines on the teensy 3.6.

If i run just a simple test code on the recalcitrant Teensy 3.6, it works fine overclocked to 240MHz.
With the real code: jitter

How dicey are the Teensy 3.6 running them overclocked? I've been doing this for five months now and this is the first time I've seen one misbehave.

Grounds look good, power is as above...

I've jumped extra grounds, I've jumped extra power and this one Teensy 3.6 just doesn't want to play nice.

I've washed both the Teensy and the board with alcohol and it made no difference.

Any suggestions on things to check or things that might help with an overclocked Teensy would be helpful.


<addition>

the trace for 5V power coming from the USB to the Vin was cut...
is it possible the cut was too deep and damaged something below?



--mjlg
 
Last edited:
PaulStoffregen said:
Does cooling it help?

Or can your application run at the normal 180 MHz speed? Might be useful for testing...
Click to expand...


Haven't tried cooling yet...
this is how things look with more samples:
3 of 9 Teensy 3.6 @ 240 MHz jitter
1 of 9 Teensy 3.1 @ 216 MHz jitter
none are a problem @ 192 MHz and lower... I CAN run all the code (still double checking) fast enough at 192 Mhz to make this work (thank you FASTRUN!!!!) , so I have a solution...

on the plus side, making it work at 192 ought to draw less power! :p

Might I be looking at tolerances in clocks?
the "SFG30A" is a crystal, yes?
I noticed that on all the one that were misbehaving that this little guy was left of center compared to the ones that were playing nice.
HOWEVER, I tried moving one of them back toward the middle and overshot to the right and it still misbehaves... need to let my coworkers do the delicate solder work, not me!
This is statistics on a very small; sample (N=9) so any other things to try (besides cooling) that come to mind, please yell.

I have a solution for my current project and I may just have to hand sort for the other project...

--mjlg
 
Status
Not open for further replies.
Back
Top