Power supply and GND

[tempus]

You misread Cliff's post - he suggested the 100uF cap after the regulator (on V_out). And yes, non electros (flim, ceramic, etc) are fine for the smaller value caps. In fact, non electros usually perform better than electros, it's just very expensive to make non-electoros in large values, plus the physical size gets ridiculous. According to PRR, the small value caps are no longer necessary - I'm assuming that larger value electros are made better today and can filter out higher frequency stuff as well as other caps, but if you want to be on the safe side, just throw a 0.1uF (not 1uF) on either side of the regulator.

[aballen]

I"m all for reducing the part count/footprint... and yeah PRR really seems to know what he is talking about.  I'm feeling very dumb right now...

So much input, and I just seem to be going in circles here.  Can you answer 3, 4 and 5 for me? 

Let me try to sum up what should be done.

1) Drop the 100uf cap on the input side, YES
2) Should I put a .1uf on the input side, ??
3) Should I change the value of C4 or drop it?
4) Should I change the value of C5 or drop it?
5) C3 needs to stay, YES
6) I'm not going to change the packages in the drawing, but I will use film caps for the small values, YES

So close yet so far..... if I am going to make this myself though, I want it to be correct.

[tempus]

1. Yes
2. Probably not necessary, but wouldn't hurt.
3. Change C4 to 10 - 100uF. Cliff (also knowledgeable) suggests and uses 100uF.
4. Probably don't need C5 but wouldn't hurt.
5. Yes
6. Yes - film caps are generally superior to electrolytics (except that you can't get them in large values) so use them where you can

You're doing fine with this ab, and you are wise to be thorough and make sure you have everything right. How's the saying go? Measure twice, cut once. Keep asking questions until you understand everything and then build it.

Just to sum up, without going into a whole lot of detail:

- the rectifier converts the AC to DC
- the DC coming out of the rectifier is pretty crappy and has a lot of noise (a lot of hum at 60Hz)
- the 2000uF cap smooths the DC so that it's not crappy anymore and resembles steady state DC and filters out a lot (hopefully nearly all) of the hum
- the 0.1 uF cap filters out higher frequency noise (but is apparently no longer necessary)
- the LM317 sets the voltage to 9v, but introduces noise of its own
- the 10uF cap (C3) filters out some (hopefully almost all) of that noise
- the 10 - 100uF cap (C4) filters out more (hopefully the rest) of that noise and helps keep the regulator stable
- the 0.1 uF cap C5 filter out some (hopefully most) of that noise (but apparently is no longer necessary)

Way more info about caps than you could ever want to know:

http://www.radio-electronics.com/info/data/capacitor/capacitor_types.php

and something a little easier to digest:

http://www.intersil.com/content/dam/Intersil/documents/an13/an1325.pdf

[aballen]

Ok, lower parts count, that is good.  Here is my latest.

1) Dropped
2) Not necessary, not added
3) C4 now 100uf
4) Not necessary, not added
5) C3 kept it
6) I'll use my np box caps where I can.

So if we are all sure the .1uf caps are not necessary, I'm good to go.

[aballen]

I think I just found another flaw in my design... I was planning on using this transformer(which I already bought)
http://taweber.powweb.com/store/pdlxfmr.gif

But according to PRR... that wont work, because its only putting out 11V on each of the secondaries....  what do I do here?  I don't expect to be pulling the full 300 ma on each supply, will the transformer I went with suffice for a simple 40 ma pedal?  If not what should I use?

I was planning on making 8 independent power supplies, in a box.... but I don't know if getting 8 separate transformers is practical.

[tempus]

You are hoping to start with 11VAC and regulate to 9V DC. The 11V AC would give 15.5V with perfect rectifiers and infinite capacitor.

- PRR

11 VAC = 11 x 1.414(RMS) = 15.5VDC. So it's actually putting out 15VDC on each secondary. This is why you need the bigger cap (because caps are neither perfect nor infinite). If you use a smaller cap, the voltage will sag below 9V between peaks, and be insufficient to power the regulator. The bigger cap prevents this. Re-read his post on that and if you're still unsure, ask again.
I'm not sure what the current draw of each of your pedals is, so I can't comment on the 300mA. I think most analog pedals pull less than 40 though. I know I once measured a bunch of mine and they were all less than 20 mA, but my digital multi-fx pedals are 300 - 500.

As for your previous post, it lookd like you're good to go. The 0.1 uF are a judgement call, and according to PRR are no longer necessary.

[aballen]

You are hoping to start with 11VAC and regulate to 9V DC. The 11V AC would give 15.5V with perfect rectifiers and infinite capacitor. Real rectifier loses a couple volts. Practical-size caps sag. The 9V rectifier typically needs 3V extra so it can shave-off to 9V exact.

You need a good 12V DC at the end of each half-cycle when the cap has sagged.

Tempus, thanks for the explanation... I'm interpreting some things incorrectly, and I appreciate you helping me get some clarity.  I really thought he was saying I needed 12V.

It looks like I'm good to go.  I'm going to keep the parts count down, I do have to make 8 of these

Time finalize my board, and order some parts, yay!  It may be a little time, but I'll post some pictures as I go.

[aballen]

Here we go, all 8, mirrored for etching

[tempus]

I really thought he was saying I needed 12V.

He was, and with the 2000uF cap you do. See, what happens is that at the start of the cycle, you've got a solid 15v, but the cap discharges over the course of the cycle. This is the sag he was referring to. If that cap discharges below about 12v (which it will if the cap is only 1000uF unless conditions are ideal - this is demonstrated in the 2nd diagram - notice the pink underlined 11.811? That's what the voltage will be at the lowest part of the cap's sag), the regulator isn't going to have enough voltage to keep regulating. It'll still work, it'll just be unregulated during that part of the cycle, maybe longer if it takes a bit for the regulator to bounce back. In the 3rd diagram, the lowest that the 2000uF cap will allow the voltage to drop to is 12.332. This is because larger caps take longer to discharge. So with a 2000uF cap the voltage at the regulator's input will never dip below 12.332V, and so your regulation will be stable and continuous.

[aballen]

Cool.  I was thinking if I started with 11v, there was no way to get a higher voltage out of it.  I guess that really depends on the available current/voltage vs the load current/voltage, and the size of the cap.

I should order some parts tonight!

[aballen]

Just thought I would post on my progress.  I redrew the circuit, and managed to get 8 of them on a 3x5 board.

This leaves me with one question.  For the 2200uf caps, is 16V enough?  They should only see 11v from the transformer and the 16v ones I bought are significantly smaller than the 50v ones.

Here is a pic... now a lot of drilling!

[PRR]

> thinking if I started with 11v, there was no way to get a higher voltage out of it
> 16V enough?  They should only see 11v from the transformer

You are missing the fact that the "11V" is an RMS ("average") reading, and the rectifier/capacitor thing is a PEAK-catching system. Assuming sine-wave power, the Peak is 1.414 times the RMS. 11 Volts (RMS) of AC makes 15.5 Volts DC.

Minus losses. There's always losses.

Plus/minus wall-voltage variation.

Plus transformer un-sag on light load.

> 16V enough?

I'd do one or two 16V caps off 11VAC for experimentation. I've done riskier things (like 36V on 35V caps for years.) But I would not like to build eight barely-sufficient caps into a small box especially for gigging. The transformer's "11V" is probably rated for 115V or 120V walls. I have 125V at my house. I hear of 129V 130V in many many places. (I have plugged-into 160V and not known it: big electro caps make large mushroom clouds.)

I'd prefer to use 25V.

[aballen]

I have 2200uf 50V caps... I'll use them if I can get them to fit... it will be very tight though.

[aballen]

Ok I got the 50v caps in, it was that or 16v.  I don't think there should be any harm in using the big caps, other than the space.  They are packed pretty damn tight though.



I'm also thinking squeezing this all on a 3x5 board I neglected considering how I'm going to mount this board or the transformer.

I'm thinking I should re-name this thread too, its pretty much a build thread at this point.

[tempus]

Looks great ab (way better than any of my builds that's for sure). Those 50v caps won't cause any problems, so don't worry about that. The extra voltage will give you a little more relability, if anything.

[aballen]

Thanks.

I really appreciate all the feedback I've been getting.  Reducing the parts count helped a lot too, no way it would fit on a 3x5 board if there were any additional components.

Now I have to figure out how to box it up.  Any suggestions for mounting a board without any space for drilling?  I suppose I could use a whole lot of double sided tape, or just go with a plastic box.

[bhill]

Velcro

[tempus]

I don't see what you mean - it looks like there's lots of space to drill a couple of holes there.

[aballen]

I finally got this assembled, and the power is rock solid, but I have a problem.  It is putting out 14.7-14.9V.  The output is consistent across every supply, so that is good, but the voltage is way to high.

The only variable I see here is the resistors.  I'm using a 240 and a 1.5K resistors, which should give me 9.06V

Did I mess up the design somehow?  I know a lot of people verified it for me, but it seems in practice something is not correct.

There is no load on the board while I test, not sure if that has an affect... but 14v seems way too high.

How do I trouble shoot this?  Any suggestions are appreciated.  I really want this working... I even cannibalized my kids Spongebob lunchbox for a makeshift case.

[Jdansti]

It seems that the +/- tolerance of the resistors shouldn't be enough to cause this. You might want to measure the Vref and use that value when calculating the theoretical Vo or R2.



Unless there's a mistake somewhere, you might just have to temporarily swap a potentiometer for one of the R2s, adjust it until you get the desired voltage, remove the pot without changing the setting, and measure the resistance. Choose a resistor as close as possible to the adjusted resistance of the pot.

EDIT: BTW, what is your AC voltage going into the rectifier?