Thinking some more and looking at the datasheets for the XL4014 and the LM2596, I’m wondering if since the only load attached to these buck converters is the Photon, ~100mA is really too low. I’ve read that the LM2596 has a discontinuous mode it enters in low load conditions which can cause problems: https://www.eevblog.com/forum/projects/lm2596-controllable-buck-converter-switching-noise-on-signal/
Perhaps I should remove these buck converters and try with a simple linear regulator (7805) taking the 18V down to 5V. This might need a heatsink in order to dissipate the 1.3W of heat. Though I’m open to alternative ideas.
Just tried a simple LM7805 linear regulator from the 18VDC source, the same problem exists. So the problem isn’t switching noise from the DC-DC converters what I think is the issue is noise being induced from the TB6600 Stepper driver, or the steppers themselves.
Connecting a long cable to the input of DC-DC converter fixes the issue in the same way as connecting one to the output did.
I agree, this would be good information.
This is true, and glad you thought of it and tested that point.
Have you tried adding 47uf, 100nf and 10nf caps to the +3.3v power rail?
Following some further experimentation. I’m now fairly confident that the issue isn’t with the Power Input to the Photon. But more that the stepper isn’t running correctly due to interference/noise on the Step/En/Dir lines to the stepper controller.
I’ve tried with replacing the wires to the stepper motors with twisted pairs (some CAT5 I had lying around) and that has significantly improved things and now have the DC-DC converter running almost reliably even with short wires. I’ve ordered some shielded cable to replace all the runs from the stepper controller to the motors and so the Photon control board.
Thanks for all the suggestions and help. Will report back with my final outcome.
