Voltage divider power supply?

[thermionix]

I'm generally a battery guy, but I now have two pedals (Carbon Copy and Chicken Salad) that draw too much current for batteries to be practical.  So I'm thinking about building a power supply for those.  Yes I know there are great products available commercially, but I'm a cheap bastard, and hey...this is a DIY site, right?

I have a transformer handy that was pulled from an old Casio adding machine.  Don't know the specs, but judging by size and weight, I'm guessing that it has more than adequate current rating for the job of powering two pedals.  It turned a motor to feed paper through the adding machine, as well as power the rest of the circuit, including a display, so I'm pretty sure I'm good for current.

But the voltage is too high.  I measured the non-center-tapped secondary at about 21.5VAC unloaded, so I'm just guessing it's an 18V secondary.  Obviously I want a 9VDC output for the pedals.  I suppose most would use a regulator, but I have no experience with those and frankly it's over my head at this point.  I could probably find something online to copy, but I had the idea that it might just be simpler and easier to use a voltage divider to get down to 9V after rectification and filtering.  Also seems the voltage output might remain relatively steady under various current load conditions.

Any problems doing it this way?  Advice?  Thanks.

[bool]

You could try to use a power zener diode (like a 1N5346) and a wirewound resistor ...

Or 2x 1N5337 in series for better heat dissipation. Because such an arrangement would get hot quick.

If I was to go this route, I would make a separate "zener regulator" for each pedal, both fed from the same xfmr/rectifier/reservoir-cap. This should be easy enough to make ...

There are variations on this theme, of course, but these would require more effort than you are willing to put in ...

[EBK]

I suppose most would use a regulator, but I have no experience with those and frankly it's over my head at this point.

Now is probably a good time to learn and gain some experience.   

[R.G.]

Voltage dividers power supplies are only practical for tiny amounts of current because to have any significant output current, you have to waste maybe ten times the output power as heat in the divider. Otherwise, the voltage out varies hugely with any changes in the load.

As EBK says, this is a great time to stretch your mind. First, go read "Power Supplies Basics" at geofex.com.

Just from your description, you're going to have practical issues with the transformer you describe. If it's output is 18-20 volts, the rectified voltage will be 25-28Vdc. A 9V regulator will have to waste 2/3 of that voltage to keep the output to 9V, so you're going to waste 2/3 of its output as heat. If you don't plan for how this heat gets wasted, you're going to have hot and dying parts.

Oh, wait. You've already started learning.     

[armillary]

I'll point to some schematics of DOD 670 and 680 pedals from about 1980 as an example of how this was done in the past. The components used are still available.

http://www.aronnelson.com/DIYFiles/up/DOD_680_schematic.pdf
http://www.aronnelson.com/DIYFiles/up/DOD_670_Schematic.pdf   

These pedals use a transformer with a CT secondary to take 120VAC to 18VAC, then two 1N4001 diodes for rectification, followed by an LM7815 (15 volt) regulator and its associated power supply filter capacitors. The DOD folks used a 470uF electrolytic between the diodes and the LM7815 and a 10uF after. And that's it. I measured 24VDC on the input leg of the LM7815 and 15VDC on the output leg.

There are numerous models in the LM78xx line, including the LM7809 (9 volt). They have three pins: input, ground and output. Input voltage up to 35V. The old ones handle current up to 1 amp, the new ones 1.5 amp or more. DOD used the TO-92 size, but the TO-220 has a tab with a hole for mounting a heat sink. You can socket the TO-92, but the pin spacing for TO-220 also works nicely with stripboard. They cost less than a dollar.     

Datasheet here:
https://www.fairchildsemi.com/datasheets/LM/LM7809.pdf

I was lucky enough to read R.G.'s post while I was previewing mine, so I will now backpedal vigorously and say that these early DOD pedals required only a few mA each, while modern effects require many times that, so given heat and power considerations, what worked back then might not be so practical now. But it's an example.

[GibsonGM]

At a glance, if it's an 18V secondary it's not the ideal transformer for your needs.  Would make a great benchtop supply I bet, tho!! 

If you want 9V from the 18V secondary, you're going to throw away 18V. That's very wasteful, and (just said) often makes a lot of heat you have to deal with.  Too bad no CT!  I'd hang onto it for another project that you can toss 18V at....my 2 cents. 

Like others said before, this is a GREAT time to start learning how this stuff works!    It's actually HARDER to set up a voltage divider for this than to suck it up and get the regulator.   You'd have it working in theory, just fine, then burn your fingers (and your parts) up pretty fast, is my prediction.   Just like we all did.   

Eventually you find yourself looking for transformers that are only a couple volts over your desired voltage, makes things much easier on you, ha ha...

It would be useful to find out what each of those boxes you wanna power draws for current; that can be done off a battery for a brief test...

[thermionix]

Okay, thanks, great replies everybody.  Yes indeed, I do have some learning to do.  I will put this transformer aside for now.  I thought the secondary voltage was too high to be very practical with a 9V regulator (heat), but I didn't know how it might work out with a voltage divider.  Also impractical I see.  So much for going cheap, lol.  I will read up, but in the meantime I'll probably just buy a power supply, might find a good one for less than the cost of parts to build my own.  I don't know the current draw of these pedals, but could easily find out with a 1 ohm resistor.  Also might be printed on the bottom of the pedals, perhaps I should look...

[blackieNYC]

https://www.monoprice.com/product?c_id=115&cp_id=11501&cs_id=1150108&p_id=611690&seq=1&format=2

6 bucks. 500mA, 9v.   DIY would be an admirable pursuit, but you could be building a pedal instead.  One more pedal. 
       (DIYmanhours squared)/(price of store-bought thing minus "DIY-wood") or something.  There's a formula we should work out.

[EBK]

https://www.monoprice.com/product?c_id=115&cp_id=11501&cs_id=1150108&p_id=611690&seq=1&format=2

6 bucks. 500mA, 9v.   DIY would be an admirable pursuit, but you could be building a pedal instead.  One more pedal. 
       (DIYmanhours squared)/(price of store-bought thing minus "DIY-wood") or something.  There's a formula we should work out.

There are some purchases that just require you to repeat to yourself, "I am serious about this." A decent soldering station and a quality multi-isolated-output power supply (I'll make an unbiased, uncompensated plug for Voodoo Labs here) are two important ones.  I will say that you should make your own pedalboard though.  My favorite one that I saw gigged (not by me) was merely a pig-shaped wooden cutting board with some velcro. 

[thermionix]

https://www.monoprice.com/product?c_id=115&cp_id=11501&cs_id=1150108&p_id=611690&seq=1&format=2

That is a good price, and more than enough current for the two pedals (MXR claims the Carbon Copy draws 26mA, sounds kinda low to me).  BUT...would have to be modified to run two pedals at once, and I tend to despise wall warts, they take up too much real estate on power strips and fall out of some looser wall sockets.  What I would build, if I went that route, would have a regular AC cord going into a box containing the transformer etc with multiple 2.1mm sockets and some patch cords.  (I would also filter the livin' snot out of it.)  I'm sure there are commercial products available like that, but probably not for 6 bucks, lol.

I'll be researching...

[maiko]

18v is to high to regulate down to 9v.

But . . . and forgive me thread owner(starter) no intention to hijack.       cant we just cascade multiple regulators to get down to 9v say start with a 7815 then 7812 then finally to 9v.   Is this a practical idea or waste of time. haha.  im not an engineer but i understand how regulators work.  that it needs just to be 3-4 v max to be efficient so thats why i got that idea.   

Ive also always wanted to ask but afraid i might get bashed because it might be too dumb an idea.

[EBK]

18v is to high to regulate down to 9v.

But . . . and forgive me thread owner(starter) no intention to hijack.       cant we just cascade multiple regulators to get down to 9v say start with a 7815 then 7812 then finally to 9v.   Is this a practical idea or waste of time. haha.  im not an engineer but i understand how regulators work.  that it needs just to be 3-4 v max to be efficient so thats why i got that idea.   

You absolutely could do that, and it would work, but it would not be very practical due to the wasted power.  A better (perhaps) solution would be a buck converter.

Ive also always wanted to ask but afraid i might get bashed because it might be too dumb an idea.

In that case, thanks for being bold.  That's how we learn.  I've started entire threads with dumb ideas (and I am an engineer, on paper anyway) just to sample the brilliant (and patient) minds around here. 

[GibsonGM]

Yes, we've all had those ideas! I've done things just like that, and then found that I've made the thing so complicated that I can't debug it, work on it....understand it... 

The first circuit I ever 'designed' on my own involved MANY opamps (6?) to make an octaver...splitting, using 2 opamps to invert phase, recombining....and someone pointed out that it's already been invented, that there are much simpler ways to do it, such as using a transformer like the Octavia.   It was a good thought exercise, but it was no better than something already invented that was far simpler and more efficient.  But I DID get to use theory I had learned, and it WORKED!  Just as your regulator idea WOULD work.

So you learn and move on - ideas aren't dumb, they're worthy of discussion, and once in 100,000 times, a new thing is invented!

[maiko]

EBK

thanks for the kind words.

What do you mean by not practical.   would the regulators get still get hot.??

[GibsonGM]

They probably would get "less hot", because all 3 would divide the load - it WOULD work.....but it is not practical to do this because it adds complexity, and you still totally waste HALF of your supply's voltage...that is very inefficient, and is a design practice that would be very much frowned upon!

Part of the effort is is to make things the most efficient we can, elegant in design...WELL-designed....it is inelegant to use a sledge hammer to arrive at a place that a more streamlined method would take you to.   If we were in a plane crash, and you NEEDED to use a method like that to get 9V to run a radio for rescue, that would be great!  But if we are starting from nothing, we wouldn't do it this way. 

They'd also take up valuable space in your build! 

Adding complexity leads to more places for failure, like I mentioned...more places to have to search for a problem.  It was a neat idea, exploring new things, but really not a good thing to really do unless you had ZERO alternatives...

[EBK]

EBK

thanks for the kind words.

What do you mean by not practical.   would the regulators get still get hot.??

Mike's answer is quite good and picks up the virtue of your creative solution: you have improved the heat problem by spreading the heat over a larger surface area (of course, you'd still need to get it out of the box). 
What we really need is a Mark Hammer analogy here, although I do like Mike's plane crash survival scenario.
The best I can come up with is driving a car down the highway with the brakes on 100% of the time.  It's an efficiency problem.

[thermionix]

The best I can come up with is driving a car down the highway with the brakes on 100% of the time.

I wind up behind those people from time to time.  The lights give them away.  I can't imagine the thought process involved.  I guess they just mash both pedals down and hope for the best!

[J0K3RX]

Unwind the secondary until you get into the 9V range..

[robthequiet]

Hm, I'm taking a different approach here... what about looking at pedal voltage requirements and see if they could work at 18v? People who make 18v pedals go through the complexity of DC-DC conversion going up. I would probably find an 18v application, such as ROGs Thunderbird or an MXR 10-band EQ, and take advantage.

[bool]

For a DIY approach, you know, zeners and wirewounds (cca 82-100R, 3W) "just work".

Can be cooked in less than half-hour - if you know what you're doing. Cost effective, too. Can be wired p-t-p, and gets warm quick for ... warmer tone.