Power supply and GND

[aballen]

AC Voltage going in to the rectifier is 11.8V, consistent across all of the secondaries.

[asatbluesboy]

Nice, now I need a PDLXFMR-2, lots of time and patience to study mains wiring and a SpongeBob lunchbox... 

Thank you all so much for this thread.

[aballen]

FYI, I don't recommend a lunchbox(even if you find a cool one). 

It seemed fairly sturdy when closed, and empty, but with the weight of that transformer in it and all the drilled holes it wont hold up.  IT will work as a temporary case while I figure out the voltage problem though.

Based on the feedback a few posts ago, is a potentiometer, really the only way to get this working?  What value?

I'm concerned something might be off in my design for the voltage to be of by this much.

[Jdansti]

You've got a 1.5k resistor now. If you have a 2.5k-10k pot lying around that would work. Remember, this is just a temporary swap to see what resistance you need to achieve ~9v. I recommend removing only one of the R2s, and if possible, use alligator clips to temporarily connect the pot to points on the PCB that are common with the R2 connections (for example, you could use any ground point and the + side of C3).

After you remove the pot and measure the resistance, use the same alligator leads to test the selected resistor and see what your voltage is.

Alternately, if you don't have any spare pots, but you have some resistors, you could sub these in starting around 2k and working upwards in resistance, hopefully find a value that works.

[PRR]

> putting out 14.7-14.9V

I think your LM317s may be in backward?

http://www.electronics-lab.com/articles/LM317/

That pinout picture is with the label facing you, the heat-surface away from you.

[aballen]

I'll flip one and test it out.

PRR, did I fry them?....

updated, Well I don't think they are backwards, and even if they were the diode should have protected them.

I flipped one, and got 14.4V, just to be sure I soldered in a new one, but filpped... still got 14v.4

something is definitely wrong.

[slacker]

Are you using this layout?

I'm afraid it's incorrect, you have the output and adjust pins switched, the middle pin is the output. You can probably get away with twisting the leads on the regulators to make them fit correctly.

[Jdansti]

Are you using this layout?

I'm afraid it's incorrect, you have the output and adjust pins switched, the middle pin is the output. You can probably get away with twisting the leads on the regulators to make them fit correctly.

Good catch Slacker. It's easy to assume that the schematic geometry is the same as the physical geometry.  I've caught myself making the same mistake.

[aballen]

Slacker!  You get the prize... which is not much, just a thank you.  It seems the pinout on the part I used in eagle was wrong... Not sure what I chose, bit I found the LM317 part, replaced it, and the error was verifiable right away.  I had to do some hand alterations of my board to fix it, but here is the updated schematic... with the correct part... and a photo of my hand re-wired board.

I'm getting rock steady voltage on every output, thought they are not all identical.  The voltage varies from 9.14-9.22v, but if one circuit is 9.16 for example, it does not budge from that spot.  I'm thinking the variations are from the changes I just made, the heat put in to the 317s by unsoldering re-soldering, inconsistent "traces" if you can call them that.

Time to box it up and plug it all in!




updated schematic, now verified

[R.G.]

I'm getting rock steady voltage on every output, thought they are not all identical.  The voltage varies from 9.14-9.22v, but if one circuit is 9.16 for example, it does not budge from that spot.  I'm thinking the variations are from the changes I just made, the heat put in to the 317s by unsoldering re-soldering, inconsistent "traces" if you can call them that.

I think it's more likely to be a variation of the reference voltage inside the various LM317's and the tolerance of the voltage setting resistors.

The LM317 datasheet from National says the nominally 1.25V reference can actually be from 1.20 to 1.30 and be within tolerance. That's +/- 4%. If your setting resistors are 1% tolerance, then the voltage they would set varies by a bit less than 1% (there's some math to do here to get the real numbers, so I handwaved this) so +/-5% variation is not unreasonable at all for un-trimmed outputs.

You're seeing 9.22/9.14 = 1.00875 or about 0.875% variation from min to max, about +/- 0.435% from min to max. That's incredibly good, and suggests that all your LM317's came from the same wafer, or were otherwise somehow selected.

[aballen]

Nice!  I definitely used 1% resistors, and all the LM317s were ordered together.  I usually do that because I know its fairly likely all the parts will come from the same lot(or wafer) and reduce variance.

Now I just need to get that 9V ac wired to a jack and I'm good.

When I'm done I'll post up a good quality board for etching so others can make something similar.  Can't thank you all enough... I've made  a handful of pedals, but I had no idea what I was getting into when I started this power supply project.  Switching to the 317 was a learning experience.  It should all be worth it.

[aballen]

I have one extra jack on the opposite side puttinh out 9.6V AC, should be up to 2 amps and will power my only non-diy pedal, my line-6 JM-4 looper.

[Jdansti]

I have one extra jack on the opposite side puttinh out 9.6V AC, should be up to 2 amps and will power my only non-diy pedal, my line-6 JM-4 looper.

...and you can pack a bologna sandwich in it when you go to a gig!

[aballen]

Ok last stupid question.  The center tip in the power supply jacks should be gnd correct?  I know wall my pedals are build with "center ground"  but before I plug something in, I want to be sure its not - to + here.... I've been using batteries for so long.  I wired up all my jacks to be gnd on the tip, which should connect the center on my pedals to gnd.  It feels really obvious... but I'm asking anyway.

Also, I really appreciate all the input, so I redrew the board, thinking someone else my want to use it.  This layout has four circuits.  Etch two side by side for all 8.  This is the final, I would consider it verified, though mine was modified by hand.  If anyone wants eagle files, I'll be happy to share them.  I can upload the files to batchpcb if someone wants to place an order rather than etch them.





Thanks again

[tempus]

Depends on the pedal, but most (including Boss etc) are centre negative. I don't know what pedals you've built or the layouts for them, but it wouldn't hurt to check them closer just to be on the safe side.

[Jdansti]

Depends on the pedal, but most (including Boss etc) are centre negative. I don't know what pedals you've built or the layouts for them, but it wouldn't hurt to check them closer just to be on the safe side.

+1

If the jacks are center negative, then yes, the tips of the jacks would all be "ground". You're using the plastic jacks that insulate both terminals from the case, so you're good.

I assume that you will be using power cables with male plugs on each end that will connect between your new multi power box and your pedals.   If the pedals you have are center negative, then you'd want the jacks on your new multi power box to be center negative. However, if you've built or purchased any pedals that are center positive, then you can't plug them straight into the PS box if all of the jacks are center negative.

I recommend clearly labeling the PS jacks showing the polarity. One option if you have one or two positive center pedals would be to make one or two jacks positive center (clearly labeled) or make an adapter cable(s) that turns center negative into center positive. If you do this, make the special cable a different color and label it as a "cross-polarity adapter".

Finally, try to develop a habit of checking your power supply and pedal for compatible polarity before you insert the plug.  Even if you know it's the same PS and pedal you've used a thousand times, the habit will save you the one time you have a different PS or pedal (kinda like checking to see if a gun is loaded when someone hands it to you, even though you saw the other person unload it).

[artifus]

just to add that you should keep an eye on different pedals in metal enclosures with mixed polarities and stray unconnected daisy chain dc jacks floating around - you don't want the jacks barrel touching a pedal that's grounded to its case +/- ...does that even make sense?! please forgive the word salad - i'm tired.

anyway the tip i wanted to give was that i use cork stoppers on unused but powered dc jacks to insulate them. you can pick up a bag full cheap at your local arts and craft outlet and they can also be used as make shift knobs for pots and have a nice feel. a drill bit slightly smaller than the shaft/barrel can make for a snug fit.

[aballen]

Sorry for pulling up an old topic, but I seem to get a lot of emails on this one.  Since I re-organized my photos, all the links broke so....

The schematic:


The layout:


An etch able layout, at 300 DPI, needs to be mirrored:


You can order a fabbed one here, its pricey, and since I already provided a single sided one, I recommend you etch, but if you have more $$ than time:
http://www.oshpark.com/profiles/aballen

Also I got it in a proper enclosure.  It really is solid now, practically bulletproof.  That spongebob lunchbox was awesome, but Viacom sent me a letter about intended use, not sure I understood most of it....

One side with 8 isolated 9VDC


One side with a lonely 9VAC:


And the power in:


Of course all the standard disclaimers apply.  Mains power is dangerous.  You should never even think about building one of these, etc, etc.