Glad to hear you're making progress. Interesting that moving the HV psu away before did not make a difference, but in the new position it did. Something must have changed with your electrical connections as well which resulted in that improvement.
Grounding is crucial to signal integrity no matter the circuit. Keeping the ground path as short as possible with large, low impedance connections should definitely improve the situation. On Sat, Nov 23, 2019, 12:38 AM Richard Scales <rich...@scalesweb.co.uk> wrote: > I put the 7805 in place but it made no difference whatsoever. I decided to > change the position of things so that the HV supply was at the 'other end' > of the clock and the CPU's were at the end where the signal goes to the > Smart Sockets. There seemed to be some improvement, at one point i thought > it was completely fixed though staring at the thing periodically I could > see that it was not not totally 'fixed'. Reading your comment about ground > wires I will beef-up the 0v and 5v connections to the logic boards as they > are indeed skinny at the moment. > > On Wednesday, 20 November 2019 16:59:26 UTC, Kevin A. wrote: >> >> Does it seem like the degree of glitching has been reduced, is the same, >> or worse with those 2 changes? >> >> An LM7805 could support your 5 volt rail depending on how you implement >> it. The higher the input voltage, the hotter it gets. If you can adjust >> your buck module to 8 volts, then put the 7805 after the buck module and >> see if that works. If your buck module will not go up to 8 volts, you could >> try using the 7805 directly from 12 to 5 volts if you're not pulling more >> than around 0.75 amps (power dissipation (heat) is increased because of the >> greater drop from 12 to 5 as opposed to 8 to 5). >> >> If that doesn't solve the problem then improving grounding is the next >> way to go. Ground paths ideally are short and use larger conductors. Having >> long, narrow ground wires strewn about may certainly be contributing to >> glitches from noise and undesirable ground loops. >> >> On Wed, Nov 20, 2019, 11:35 AM Richard Scales <ric...@scalesweb.co.uk> >> wrote: >> >>> I tried the first two suggestions and thought (initially) that all was >>> good but upon watching in detail it would seem that it is the same. >>> >>> I have a L78M05 to hand - is that going to do the job? >>> >>> Otherwise I can get something like this >>> https://uk.rs-online.com/web/p/low-dropout-voltage-regulators/1246447/ very >>> quickly. >>> >>> >>> >>> On Wednesday, 20 November 2019 15:33:20 UTC, Kevin A. wrote: >>>> >>>> Here would be my first 2 suggestions in changes to your circuit: >>>> >>>> #1: I would definitely place the ground from my logic circuits straight >>>> to the main power ground, instead of routing it though the 5V buck module. >>>> The buck converter is also a switching power supply and this could very >>>> well be coupling noise through the rest of your circuit, especially if it >>>> is the only ground path for the logic. >>>> >>>> #2: If #1 does not fully alleviate the problem, I would try placing an >>>> electrolytic cap of between 100-470uF on the 5 volt rail close to the logic >>>> circuits. This could help decouple the logic from any noise entering >>>> through the 5V rail. >>>> >>>> Finally, if the above two do not solve the problem, I would use a >>>> linear regulator after the 5V module to provide a clean 5 volts to your >>>> logic circuits. It looks like the 5V module has a trim pot on it, so I >>>> would increase the voltage to 6 or 7, and then use a 5 volt LDO (low >>>> dropout regulator) to provide the final 5 volts. This certainly will >>>> alleviate a large amount of noise on this line that you might not even be >>>> able to observe on your scope, unless you zoom in quite a bit on the >>>> vertical (voltage) scale (so that you're looking at hundreds or even tens >>>> of millivolts per division). >>>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "neonixie-l" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to neoni...@googlegroups.com. >>> To view this discussion on the web, visit >>> https://groups.google.com/d/msgid/neonixie-l/7e183735-7ca9-4287-8cc8-51e22a65d43f%40googlegroups.com >>> <https://groups.google.com/d/msgid/neonixie-l/7e183735-7ca9-4287-8cc8-51e22a65d43f%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> >> -- > You received this message because you are subscribed to the Google Groups > "neonixie-l" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to neonixie-l+unsubscr...@googlegroups.com. > To view this discussion on the web, visit > https://groups.google.com/d/msgid/neonixie-l/c427b9f9-a8f4-4acb-994b-345063756793%40googlegroups.com > <https://groups.google.com/d/msgid/neonixie-l/c427b9f9-a8f4-4acb-994b-345063756793%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+unsubscr...@googlegroups.com. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CALcVLKJFD_7d4USGx-A-Tjm%2B5mDLO3Ad%3DnbmR6JW6r9S4wpMYg%40mail.gmail.com.
