Hacker News Reader

Comment by hw-guy | Hacker News Reader

Just a guess...maybe the 5V is supplied from a linear regulator with a maximum dropout of 1.2V, so the board is specified to require at least 6.2V from the wall wart.

View on HN

That sounds entirely plausible. Many of the hobby electronics projects I make require at least a 6V supply because of exactly this. I usually use a 9V wall wart, though, because those are very commonly available (so are easy to scrounge up).

For what it's worth in hobby projects it's time to throw linear regulators into the dustbin. There's 7805 pin-compatible switching modules now!

Stuff like this. [1]

[1] https://www.digikey.com/en/products/detail/cui-inc/VXO7805-5...

Depends on what you're doing. For a lot of sensors and things you might want the power-supply ripple rejection (PSRR) you get through an LDO. I'm not doing hobbyist things, but in our designs we have a lot of places where we will have a switcher that generates an intermediate voltage to feed a bunch of ultra-low noise, high-PSRR LDOs for different components (isolates them from each other too). This is RF though, where that is especially important.

I just meant as a replacement for your main input power regulator - for the 7805 way too many people still use.

You are absolutely right that there is a role for LDOs in modern devices.

There's no reason for most hobbyists to be using super hot, super inefficient and super constrained high-dropout linear regulators from 1972 when there's a pin-compatible all-in-one switching module for roughly the same price. To pull more than a couple hundred mA at the lowest voltage differential you need a heatsink!

Trying to use a switching regulator for everything becomes a complete nightmare when you have to debug a circuit. You scope something that should be a square wave or similar and it has the switching frequency of the regulator superimposed upon it everywhere.

Most people don't even use input power regulators. 5V wall warts are dirt cheap due to USB charging and provide the exact voltage you need.

Why mess around with a 7805 when you already have a dozen 5V power supplies lying around the house?

Unless it's a USB power adapter - which is a good choice, btw - I'd be very wary about what's coming out of that 5V no-name wall wart. That power can look like ass and not be close to 5V. If you're doing that an LDO isn't a bad choice.

90% efficiency for $2.50ish. Pretty cool, thanks for making me aware of these.

IIRC, you can get them for ~$1 on aliexpress. They're handy little devices.

Linear regulators are still preferable when very clean supply and no spurious RF radiation are needed. Filters and screens can be easily added so that one gets both high efficiency and clean output, but they add to the cost and size.

That's exactly it. Any less than the regulator won't regulate properly. More than that it'll dissipate too much heat.

Combine this with the sibling thread about lead acid battery voltages (2.1V per cell, 3 cells gives 6.3V), can we then say that this is the reason 5V circuitry is so common? 6.3V from battery ancestry plus a linear regulator?

Yeah I think even earlier 6.3 volts was also the voltage used for the heating element on vacuum tubes. I think because very early radio's used 6V batteries for the coils. And a separate high voltage battery for the plates.

Would not surprise me if there are roots going back to telegraph.

Yep, modern LDOs can maintain 0.2-0.3V with good regulation, but older cheaper ones are a diode drop (~0.6V) or two.

Older cheaper voltage regulators are typically more like 2V drop and the optimum input voltage is somewhere around 7-9V for a 5V output. In practice a lot of devices used 9V power supplies because that was the lowest widely-available voltage that met the requirement. More modern LDO regulators would have plenty of headroom on plain 6V. It'd be kind of unusual for any electronic device to need the extra .2V in this particular case.

I think you just described the 7805?

https://www.tutorialspoint.com/what-is-voltage-regulator-780...

It's a great old school design, but there are somewhat newer capless designs that are about 2 drops like the LM2940.

There are a lot of other designs between a 0.2V current feedback pFET design, but not all of them have been so popular that their costs are close to packaging.

That extra 0.2V is to go capless, which saves a few cents and some area on the PCB.

That doesn't really make sense, though. If you need 5V, why use a 6.2V wall wart and a LDO instead of just using a 5V wall wart in the first place?

That was my immediate thought too.

I think you nailed it.

Reply on news.ycombinator.com