SX-Key Clock Jitter
pjv
Posts: 1,903
For Those Who Care;
I just got bit again!
In doing some critical timing debugging at 50 MHz using the SX-Key, I noticed up to 200 nanosecond jitter in my interrupt routine when there ought be none.
Reprogramming the device in non-debug mode, and plugging in the 50 MHz. resonator made it rock-stable.
Then of course I remembered I have been tripped up by this more than once before; the SX Key's phase-lock loop oscillater causes the jitter in its clock generation. Perhaps it is (somewhat) power supply induced as I'm using a 5V switcher. I'll have to check that out.
OK, checked the switching power supply, and while there is a 100 mV or so blip at the occurrence of the jitter, this happens roughly every 10 milliSec, and my switcher is much faster than that, and it is fed from a regulated 12 V bench supply.
So the only conclusion I can draw is again the SX Key's circuitry (phase lock loop ?) drawing a power pulse at that rate. If it becomes enough of a problem I will delve moore deeply.
Will I never learn??
Cheers,
Peter (pjv)
Post Edited (pjv) : 3/9/2005 9:39:44 PM GMT
I just got bit again!
In doing some critical timing debugging at 50 MHz using the SX-Key, I noticed up to 200 nanosecond jitter in my interrupt routine when there ought be none.
Reprogramming the device in non-debug mode, and plugging in the 50 MHz. resonator made it rock-stable.
Then of course I remembered I have been tripped up by this more than once before; the SX Key's phase-lock loop oscillater causes the jitter in its clock generation. Perhaps it is (somewhat) power supply induced as I'm using a 5V switcher. I'll have to check that out.
OK, checked the switching power supply, and while there is a 100 mV or so blip at the occurrence of the jitter, this happens roughly every 10 milliSec, and my switcher is much faster than that, and it is fed from a regulated 12 V bench supply.
So the only conclusion I can draw is again the SX Key's circuitry (phase lock loop ?) drawing a power pulse at that rate. If it becomes enough of a problem I will delve moore deeply.
Will I never learn??
Cheers,
Peter (pjv)
Post Edited (pjv) : 3/9/2005 9:39:44 PM GMT
Comments
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out· the "SX-Video Display Module"
www.sxvm.com
·
When I use the SXkey to run the circuit at 50 MHz, an output pin high or low looks considerably·noisier than doing the same thing from a 50 MHz resonator.
something periodically. I've noticed the same issues running with a >1mbps IR data link.
Makes debugging a little more painful, but still possible.
-dan
According to my measurements, the clock jitter is a maximum of 200 nano Seconds, and is not due to debug activity. It behaves the same whether in DEBUG mode using the SXKey, or in RUN mode with the SXKey. Using an external resonator makes the jitter disappear.
During debug activity, that is getting register/memory screen updates, the processor executes debug software that has been loaded into the processor during the DEBUG command, and that software takes many milliseconds to execute; that is not the jitter I was referring to.
Peter (pjv)