> "bring your board under microscope, cut the 12th 0,0Xmm wide trace out > of 37 running close to each other, then .... " Hmmm, yeah... I can see how that suggestion might not appeal to all users! :) > I have to fiddle somewhere in the middle (of the transmission line). No easy answer, then. Would it make sense it install a noninverting buffer IC and break the signal path into *two* transmission lines? There are some really tiny, single-gate IC's available in 6-pin SMD packages, and the prop delay is generally 3 ns or less. A quick plug re these IC's: I can't promise they're the solution to the issue at hand, but these IC's do offer a *lot* of problem-solving power. Anyone who enjoys hacking hardware should become familiar with them, IMO. For an overview, see my post on anycpu.org "Tiny, superfast gates rival programmable logic" http://anycpu.org/forum/viewtopic.php?f=16&t=270&p=1803 Jeff On 2020-01-02 12:27, silverdr_at_wfmh.org.pl wrote: >> On 2020-01-02, at 17:55, laughton_at_cyg.net wrote: >> >>> maybe someone has a better idea on how to "fix" that clock >> >> My first instinct would be to put a low-value resistor in series with >> the signal, with the resistor physically located near the IC pin which >> is the source of the signal. No capacitor, just a resistor... 47 >> ohms, maybe, but in the range of roughly 22 to 100 ohms. > > Right, this type of "source termination" comes to mind first, as > easiest and quite effective. The problem is that I can't really fix > this on the board[*], where the actual source lies so I have to fiddle > somewhere in the middle (of the transmission line). > > > * - theoretically I could probably do this on my unit, but I surely > can't tell everyone who would be willing to put BeamRacer into > Reloaded board to "bring your board under microscope, cut the 12th > 0,0Xmm wide trace out of 37 running close to each other, then .... "Received on 2020-05-30 00:05:04
Archive generated by hypermail 2.3.0.