All excellent points. Having been on the fixing end of this stuff but never design I never really thought about the how's and why's of certain layouts. They just worked or they didn't. Electronic design engineer I am not.
Though I am very well aware of just how dangerous electrolytics are ... having been responsible for letting the smoke out of them on various occasions. Choosing cap's seems to be a black art. I think I need a pro. The link Jonathan F pointed too has opened my eyes. Flyback is needed here I think. On Thursday, February 11, 2016 at 1:46:38 AM UTC+8, gregebert wrote: > > There's nothing inherently unsafe about non-isolated supplies; like any > high-voltage project, you have to be careful about keeping everything > properly insulated and follow minimum-spacing rules carefully. > > The most-dangerous item in a high-current supply is the electrolytic cap. > You *must* handle all of these concerns > > - Extra margin for voltage-rating. I use 450V (500V when possible) > caps for ~200V supplies. Higher voltage ratings also reduce leakage > current. Caps designed for solar-energy inverters are perfect because they > come in high-voltage + high temp + high ripple-current. > - Be very conservative with rms ripple-current, because it directly > affects self-heating. Obviously you dont want any warm components near > your > capacitors. Remember: Lifetime is dramatically impacted by temperature. > - You must have a discharge resistor; I even put a flashing neon bulb > across my large caps to indicate they have dangerous voltage. Larger > resistance values take longer to discharge, but they reduce wasted energy > (heat) > - Series fusing. In case the cap fails, you want to blow a fuse, not > the cap. The fuse must be a small as possible, and dont forget the RMS > charging current is not sinusiodal. BTW, this fuse is for the cap; it's > in-addition to the fuse at the AC input. > - Charging-current needs to be limited during power-up. For the 1.5 > amp supply, a 1500uF cap will have 10 volts of ripple at 50Hz when using a > full-wave rectifier. I'd suggest a series charging resistor of about 200 > ohms to charge at power-on, then 'shorting' the resistor with a relay > after > charging is done. > - Surge protection at the AC input. I always have a fuse on both AC > lines, then a varistor and 0.01uF capacitor (for filtering hash noise if > it's present). If your current is 'low' (which isn't the case in this > design), adding series resistance to create an RC filter is a big help, > and > furthermore the resistors will act as secondary fuses if you pick > low-enough wattage. > - Reverse-polarity protection diode. If you dont use a bridge > rectifier, be sure to put a protection diode across the cap. BTW, a > bridge-rectifier gives you reverse-polarity protection down to 1.4V, > whereas a single diode is 0.7V. Polarized electrolytics caps can be > damaged > with as little as 1 volt of reverse voltage, so you may want to add the > diode even if a bridge rectifier is used. > > -- 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/2ac038a0-aadf-42ca-989c-fb8b8fb93ca6%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
