Clean power from any source? (Battery, unreg., reg. power supplies)

[earthtonesaudio]

Here's my dream:
An efficient way of providing clean, quiet power to my effects without dropping voltage.  Reverse polarity protection too.

Possible solutions:
-On-board regulator. Pro: works great with unregulated supplies  Con: drops voltage when using a battery (or a regulated supply)
-Zener-based regulation.  Pro: Zero voltage drop when using batteries or regulated supply (9.1V zener)  Con: inefficient, sometimes noisy, subject to temperature-induced instability
-Current-limited zener regulation. Pro: Same as simple zener regulator, but wastes less power; might add some additional filtering action.  Con: more complex circuitry, slight voltage drop due to the current limiting.
-Switching regulation.  Pro: Can be made to work with about any supply voltages, and output basically any voltage you want.  Con: cost, complexity, possibly adds switching noise.
-Low-dropout regulators: I don't know much about these... maybe this is what I want?

[slacker]

Here's a completely over the top way.
Full wave rectify so you get polarity protection plus the ability to use a AC supply if you want. Then regulate down to 5 volts using a low drop regulator then voltage multiply back up to 9ish volts with a MAX1044 type chip.

I cant' find the post but I think it was from Paul Perry about a synth manufacturer that basically did that or something similar. 

[earthtonesaudio]

I had almost the exact same thing written (minus the FWR bit) under the switching regulator section before my session timed out and I had to re-type the whole thing.   

I guess if you used a Schottky FWR and some sort of efficient buck/boost regulator, then the only downsides would be complexity, cost (?), board space (unless surface mount), and switching frequency noise. 

[earthtonesaudio]

For the particular effect I'm working on, I think I'll end up doing the "down, then up" trick suggested by Ian.  Think I'll design the voltage multiplier from scratch though.

[R.G.]

Be aware that if you FWR, both the positive and negative side of the power available inside the effect is one diode-drop removed from the external power supply. So you can't hook the external power supply directly to other effects with one side grounded without interference and possible power supply issues.

If the external power supply is to be merely an external power source, not connected to ground, then you're doing almost all the work of rectifying/filtering/regulating inside each pedal. Space, cost, complexity...

[Sir H C]

LDOs could get you there, but reverse voltage protection is not happening, but a series shottkey diode could get there with minimal drop (.3 - .4 volts).

[earthtonesaudio]

I think I'll do this:

Series diode--> 5V regulator--> voltage multiplier--> effect circuit
The 5V output can also be used for LEDs, TTL, CMOS, or other requirements.  The multiplier output would feed low-power pedal circuitry.
The low input voltage requirement might help extend battery life, if the rest of the system is efficient enough.

This is overkill for most things, of course.  But the pedal I'm working on needs at least 8.5V, or it just plain stops working.

[darron]

i'd definitely agree that i think there should be a rectifier in there for the reasons that slacker mentions.

funny you guys keep talking about regulated then voltage multiply. whenever i thought about this question i always thought voltage multiply then regulated drop. that seems to give a cleaner and more accurate response in my mind. The only thing I don't like is the regulator is almost halving the voltage. If you wanted to be really anal you could have two regulators to step down smoother, but I'm not sure if it would really make a difference?

take a look at the data sheets for some regulators. the efficiency rates of some are only around 70-80%. The better ones seem to give above 90%. The trend seems that the more current they pass the less efficient they get. Also, for boost regulators, the higher voltage you want and from a lesser voltage makes less efficient.

I considered making something like this to put into all my builds, but decided it's better just to use good filtering practices etc. and assume a good power supply from AC/DC convertor or battery.

[earthtonesaudio]

I think most professional and DIY builders do just that, assume a good power supply and provide adequate filtering for that.  And that's totally fine.  In a perfect world this is all you'd ever need.

I have 2 reasons for wanting to go beyond this.  First, I use unregulated power supplies because I get them for free (my family knows to check with me before disposing of anything electronic) and to keep them out of landfills.  Second, I try to design a circuit as if it's going to be for sale later, because I might eventually do this for a living (hey, it could happen).

You're right about the efficiency part though.  It'll be interesting to see what the final power requirements are if I get around to implementing all these ideas.

[earthtonesaudio]

Just found this really cool javascript calculator for %^&*roft-Walton multipliers:
http://www.cosmicrays.org/muon-%^&*roft-walton-calculation.php

[jakehop]

Am I the only one who thinks the 5V regulator WON'T extend battery life?

It's not like the battery's got the a lot of energy left, when it's what we call depleted (7V or so). And doesn't the regulator dissipate everything over 5V as heat? That can't be efficient?!

Kind regards, Jake

[R.G.]

Am I the only one who thinks the 5V regulator WON'T extend battery life? It's not like the battery's got the a lot of energy left, when it's what we call depleted (7V or so). And doesn't the regulator dissipate everything over 5V as heat? That can't be efficient?!

Like everything involved with power supplies, you need to know the details to tell.

Your intuition is correct, series regulating down to 5V, then up-vertng and re-regulating is a lossy way to go.  However, it does allow you to eat the last remaing crumbs out of the battery as its voltage falls. Maybe it helps, maybe not.

I've kind of been staying away from this, as the discussion has been directed at what's expedient, not what's elegant practice. If you want good practice, high efficiency and all that, you make up a small switching converter to chop the input and then regulate the output. That lets you wring all the power out of the battery, and do it at about 90+% efficiency. There are small flyback switching converter chips that would do this with a single chip. The only issues are
(a) do you need ground isolation from the power source? If so, that forces you to use a transformer, not simply an inductor.
(b) what is the range of input voltages you will accept? It's fairly easy to make an input range of 2.5:1 on a flyback converter, so you could could accept inputs from about 7Vdc to maybe 18Vvdc.

I've always avoided telling people to go make switching converters because it's picky work getting the magnetics right and it involves high frequences that you can't see on a DMM (!  ) and so the instrumentation is something people here generally don't have.

I guess I need to go post a design at GEO.

[MarcoMike]

I'm not sure about the "eat the last remaining crumbs out of a dead battery"... batteries are not like tanks... I try to say it better: a 9V batteries doen't drain as a 9L tank The content of energy is not directly proportional to the voltage (see Nernst equation, which anyway is not for real-life devices as batteries). my guess is that if a 9V battery ever reaches 7V, its energy content is likely to be few % points (I'd say 1%...). so the energy you wasted during all its life with the voltage doubling circuitry is hardly recovered in the 8V-"0V" range.
As usual, I have to admit I'm just a newbie in electronics... but here is the way I did it:

input-FWR-7812-7809(and boosted 7809) this is integrated in the pedalboard. I don't even carry my PS anymore as I can plug everything in it, as far as the plug matches AC, DC(+in), DC(+out). if you really need the 9V battery plug, it's easy enough to connect one on the daisy chain coming out after regulators, even if you are loosing the 12V and 9,6V, which are not mandatory at all... just a nice feature...
last thing: a 9V battery is usually able to provide <500mA with <500mAh capacity, the whole doubling/rectifying/filtering will even reduce that... that battery will not last long... it may just not work at all if you're asking more than its max current...

[R.G.]

I'm not sure about the "eat the last remaining crumbs out of a dead battery"... batteries are not like tanks... I try to say it better: a 9V batteries doen't drain as a 9L tank The content of energy is not directly proportional to the voltage (see Nernst equation, which anyway is not for real-life devices as batteries). my guess is that if a 9V battery ever reaches 7V, its energy content is likely to be few % points (I'd say 1%...). so the energy you wasted during all its life with the voltage doubling circuitry is hardly recovered in the 8V-"0V" range.

I wasn't clear. You are correct that there is very little energy left in a 9V battery when it's below 7V. This is why battery makers set 7V as a nominal "battery dead" point. And yes, it is wasteful of the power in the good range of the battery to do all the up/downverting.

As I said, the right thing to do from an elegance and economy view is to go design a small switching front end that has a FWR, filter cap, and flyback switching converter with isolated output.

In the minutes since I posted that other note, I've looked up some things. It's probably possible to get one of the UC38xx power controller chips to do this application. The design work exclusive of the transformer is right out of the app notes, but the transformer design will need a bit of work and luck to find parts that are both effective in the application, as well as easy to find for the amateur and easy to put together and have work. Ferrite switching transformers are much easier to wind than output transformers, for example, but most amateurs will never wind a transformer at all. For that matter, the vast majority of EE professionals will never wind a transformer either. 

[earthtonesaudio]

Looking for a more elegant and efficient solution, maybe something built around this:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MAX631ACPA%2B-ND

No info on the output ripple though.

[slacker]

funny you guys keep talking about regulated then voltage multiply. whenever i thought about this question i always thought voltage multiply then regulated drop.

The reason I suggested regulate down and then multiply is that regulators are pretty tough, for example if you used a 7805 it will work with an input of up to 35volts DC. With rectification/polarity protection that means you can plug in just about any wallwart you've got lying around and you won't kill it.

Am I the only one who thinks the 5V regulator WON'T extend battery life?

I realise this is an inefficient way of doing it but I was really thinking about using this with wallwarts not batteries. Like R.G has pointed out there's probably loads of better ways of doing this, looks like the thread has turned into a discussion about what I suggested though rather than any of the other ideas Alex posted

[darron]

that regulator looks kinda cool


i remember i was looking into making a power supply that could charge FAST using super caps. when looking up one of the possible boost regulators i found a forum where people were wiring them up to their PSPs to catch those final voltage dropoffs and extend their play time. they believed it was well worth while... but batteries are going OT now.

[birt]

maybe you should look at Aphex pedals. they are designed to be used with a 9V battery, any power supply from 5 to 32VAC and 7 to 48VDC or phantom power.

[earthtonesaudio]

I found their site but I couldn't find any info other than what you said.  I wonder if they're using a switching regulator or something like Ian's suggestion...?

[birt]

i could open one?