Ultra Clean 9VDC Power Supply

[akm]

I've built this (http://www.generalguitargadgets.com/effects-projects/power-supplies/ultra-clean-power-supply) and it works well in my home which was built before grounding was wired to wall outlets.

It does not work when I try it at my rehearsal space which has grounded AC outlets.

There are also issues when I plug more than one of my home-made pedals in the same time.

I suspect I have a few things going on; can someone explain a few things about how grounding should be managed in this circuit? I connected all grounds in the circuit to the same (AC) ground. Attaching some photos but it's hard to show everything; can pull it apart and get better photos if it helps.

[Rob Strand]

It's probably because you have grounded the positive rail.   If you are going to ground the DC you need you ground the negative rail.  It would be more normal to wire the DC outlets as centre negative for pedals - look at some commercial pedal power supplies.

Lots of safety issues with your wiring, especially the mains:
- The mains wiring should not be solid core wire and it should not be low voltage wire.
- It would be wise to ground the chassis of the transformer (with thick mains rated wire).
- a loose board inside the enclosure with exposed mains wires is bad.
- The solid wires from the transformer secondary should be flexible, not solid.
  The board flapping around could cause them to fatigue and break-off and then they could
  fall down onto the mains terminal.  That would connect mains to the DC output.

You would be better off buying a DC adaptor and powering the regulator of that.  That way all the mains wiring is in the adaptor.

[akm]

Alright thanks for all the advice Rob, will rethink this thing.

[bluebunny]

Everything that Rob said, x1000.  Your pictures scare me to death for so many reasons (and none of them to do with "will it power my pedals ok?").  Please don't die.

[antonis]

>Off-topic ON<

Is there any reason for so much "dispensable" [(25*1.414 - 2*0.7 - min Vin-out) * I] power dissipation when only aiming at 9V output..??
(transformer secondary could be rated down to 10VAC without issue..)

>off-topic OFF<

[ElectricDruid]

+1 what others have said. That looks seriously dangerous.

Please buy some heatshrink for those mains connections! If that PCB wobbles about you're going to have smoke at best and mains on the outputs at worst. Even if it doesn't kill you, it'll kill all the pedals you've got plugged in.

[antonis]

And with NO mains fusing.. 

It's getting more and more interesting..

[Phoenix]

Andrew, I'll echo other above comments on safety (lack of rigid mounting of board, heatshrink, fusing, appropriately rated wire on the primary side, etc).

>Off-topic ON<

Is there any reason for so much "dispensable" [(25*1.414 - 2*0.7 - min Vin-out) * I] power dissipation when only aiming at 9V output..??
(transformer secondary could be rated down to 10VAC without issue..)

>off-topic OFF<

While I agree that this design has excessive dissipation in the regulator, personally I'd be a bit more conservative than 10VAC, that leaves zero margin.

I'd aim for:
Vout 9VDC (let's say 9.6VDC for "fresh battery")
+3V LM317 dropout voltage (12.6VDC)
+2V ripple voltage (14.6Vpk)
+2x1V Vf - 0.7V is marginal for silicon rectifiers (16.6Vpk)
/Square-root 2 (~11.74VRMS)
/0.9 to allow for 10% mains voltage variation - don't want it dropping out in summer/winter when everyone turns on airconditioning/heating (13VRMS)

A 13V secondary might be a little hard to find, rounding up to the next common value of 15VAC allows for smaller bulk filter cap, or better, an RC pi filter before the regulator (I'd split the dropping resistance between positive and negative rails for better common-mode rejection - make this closer to "ultra-clean" - something I don't think this design has any claim to).

[antonis]

+2V ripple voltage

You're very stingy with smoothing capacitor value, Greg..
(1000μF cap is more than adequate for less than 1% ripple at 10mA load current..)

On the contrary, you're very generous with rectifier diode forward voltage drop..
(never experienced more than 700mV for a couple of tens mA current..)

P.S.
Shall we shake hands at 12Vrms secondary rating..??

[Phoenix]

+2V ripple voltage

You're very stingy with smoothing capacitor value, Greg..
(1000μF cap is more than adequate for less than 1% ripple at 10mA load current..)

On the contrary, you're very generous with rectifier diode forward voltage drop..
(never experienced more than 700mV for a couple of tens mA current..)

P.S.
Shall we shake hands at 12Vrms secondary rating..??

I don't know, say a 10mF bulk filter (10,000uF), 1.5A load, that's 1.5/2fC = 1.5Vpk-pk ripple at 50Hz... I wouldn't call that stingy!
1N400x Vf is 1V at 1A...
Clearly we're making very different assumptions about the load 

(And obviously more than Andrew's aiming for with his choice of transformer)

[Rob Strand]

A 9V regulator fed by a 9V AC transformer or unregulated 9VDC adaptor (ie. 12.4V DC unloaded) will work but what generally happens is the regulator drops out with any appreciable load at which point the dips in the ripple waveform feed through to the output.  A 12VAC / unregulated 12VDC  will fair a lot better but still might not make it to the full rating of the adaptor.   For example a 12VDC adaptor might only put out 10.8VDC (average) at the rated current  and the dips of the ripple with drop below 10.8V for sure.    How much load we can pull gets into specifics of the adaptor regulation, ripple, and the regulator drop-out.

[akm]

Thanks for all the constructive feedback 

I'll use this power supply instead for now:

[antonis]

Clearly we're making very different assumptions about the load 

That's for sure...

P.S.
Just realized we're talking about multiple 9V outputs so I humbly revoke my load current assumption..