For a clock this might be an acceptable solution, but for a music visualizer, it depends on the source of the audio. If there is a headphone jack or the like for audio input, a grounded source will fry the rectifier circuit!
On Thursday, March 22, 2018 at 3:13:23 PM UTC-4, gregebert wrote: > > Or, you can just deal with the AC mains directly. I have 6 clocks of 3 > different designs that run hot on the AC line to varying degrees, but in > all cases the nixie anode supply is not isolated. Most have been running > for 6 years now, with no issues. All told, there's almost 30 years of > runtime (250,000 hours) with zero circuit issues. > > If you are going to go this route, it is absolutely essential that you > provide redundant fuses, have good surge-protection, proper PCB trace-width > and separation, and noise-filtering. And yes, be sure to use an isolation > transformer during testing/debug. Being extra careful will go a long way; > it's been more than 30 years since I've been shocked, had something > smoke/overheat, or explode. > > There's nothing inherently "bad" about the AC mains; the real issue is > that they are typically protected by a 20amp circuit breaker, so it takes a > lot of power to trip the breaker. When all of that power is running thru > your gizmo, it will cause a lot of damage. In other words, NEVER rely on > the circuit breaker in your house to provide safety for your project. > > For starters, use a UL-listed (or equivalent) power-entry module. They > have a socket for an IEC power cord, a fuse, and a power switch, all in one > nice compact unit. So no matter what happens with your gizmo, a > *properly-sized* fuse in the power entry module will blow before damage > occurs. I suggest the smallest slow-blow fuse that is rated for your > typical operating current. That way, any power-on transients to charge > power-supply filter caps will not blow your fuse. If you are extra > diligent, run a simulation of the peak AC line current during power-on, and > use the fuse-manufacturers datasheet curve to see if it will withstand the > transient. If you're lucky, a fast-blow fuse might even work. > > Next, when your AC power comes onto your PC board, have *both* legs (hot > and neutral) run thru their own fuses; PCB-mounted fuses are inexpensive. > Now you technically have triple-redundant fusing; if any of the 3 fuses > (power-entry, hot-lead, neutral lead) blows, you will not have a complete > circuit for the AC line. After the fuses, add a varistor for surge > protection; a nasty surge will blow a fuse though the rest of your gizmo > should survive. Then place a 0.01uF 1kV cap across the varistor to absorb > fast transients. > > If you run SPICE simulations on your gizmo, you can compute the RMS > current, and use that to select your fuses. Otherwise, use a 'Kill-a-watt' > device to measure the actual current consumption. > > > > -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- > In case you wondered, it's pretty easy to find out what time it is in my > house....total of 61 nixies are on-the-job, and I've only had 2 fail (both > were Burroughs 6091) > -- 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 [email protected]. To post to this group, send an email to [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/d8a98142-14c0-40ed-8e87-7a6d3d7d96eb%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
