I've been looking into off-line regulators. I'm curious if anyone has a recommendation for a circuit. I'm sure that there are circuits that are better than this.
I'm looking at it for an embedded application. Those 1/2 watt resistors aren't low power friendly.
I'm sure a high side driver could be used as a buck circuit. But I haven't found a PWM drive chip that I'm happy with. Any one have any suggestions? Isolated vs non-isolated is less important, but I would prefer isolated if reasonably possible. My desired output voltage is around 3.3 to 5V, for MCU usage, so probably less than 100ma.
thanks for the info, I've never heard of 'off-line regulation'
Is your goal to go from the 'wall plug to the device' in the least amount of steps ? to me, safety trumps all - isolation would be a hard requirement for me wrt to anything that plugs into the wall.
My direct need is completely internal to the device. Not external wires, or external connections what so ever. It will be a self contained potted device with one wire in, and one wire out. If it fails internally, there won't be any abnormal paths for electricity to flow.
My plan is that most of the time, it will function as high impedance device, with a small leakage current to power the MCU. This is where this off-line regulator comes into play. In this case the application allows for the leakage. Goals for now include low cost which likely means the least number of steps. Sorry for some of the ambiguity, this project probably shouldn't be too public at the moment.
There are also switching variations, these allow for both isolated, and non-isolated variants. If I can get the isolated for under say $2 usd, I could probably justify the extra safety. However, in this design, it would be potted, so if it does fail, and puts a high voltage to the MCU, it still has a second safety that would also have to fail.
Also note, that isolated doesn't specifically mean safer. I once had to diagnose an issue with a vendors UL listed battery charger, where the battery leads would float and cause a notable shock. It passed UL's safety standards because it was completely isolated. I would prefer to have seen at least one lead from the output tied to ground. So I feel the safer approach is mostly isolated, vs completely isolated.
Claims up to 700ma includes the normal protections.
I'm a bit concerned about the 125V note. If one rectifies 120 and adds a cap, that would be around 170V. It appears that to work with this voltage, one would want to include a second current limiting stage as noted in the suggested schematics. That puts the cost of this regulator around $4 usd.
Seems to me that a switched variant would waste less heat, and could likely cost less $.
