Why the small resistor in series with power?

[bluelagoon]

Ah, so i am guessing the 50 ohm resistor is the winner there in that last graph response.
All said and done, I have a particular power supply, if some of you more knowledgeable folk wouldn't mind please commenting on its
strengths and weaknesses, and suitability. I were mostly wondering if C1 Filter cap ought to perhaps be higher at 220uF , and if a second smaller 100nF filter cap  would help in paralell with C1? Thanks

[antonis]

I'd make C1 470μF/16V and place 100nF ceramic cap in parallel (as close to R1 right leg as physically possible..)

[Clint Eastwood]

D1 is for reverse polarity protection I guess? you lose about 0.3 volts here, something you don't seem to like. Maybe opt for no protection and just not reverse polarity, or choose a method without voltage loss?

[FSFX]

D1 is for reverse polarity protection I guess? you lose about 0.3 volts here, something you don't seem to like. Maybe opt for no protection and just not reverse polarity, or choose a method without voltage loss?

D1 is not needed for reverse polarity as long as the Zener diode and resistor R1 are high enough wattage. The Zener will provide both overvoltage and polarity protection with R1 providing the current limiting and noise filtering in conjunction with the decoupling capacitors.   

[FSFX]

I'd make C1 470μF/16V and place 100nF ceramic cap in parallel (as close to R1 right leg as physically possible..)

Actually, for better noise suppression, the 100nF would be better placed as close as possible to pin 1 and 8 of the charge pump IC, not R1.

[antonis]

2W for R1 should be the minimum power rating..
As for the Zener, any DO-41 package (like BZV85) should easily handle more than 212mA (the current through Zener and R1 in case of power reverse polarity conection..)

[bluelagoon]

Thanks all for the advice, and Clint, I'm not totally against voltage drops, just would prefer not if there were ways to avoid. I know  if you get that desperate you can put two 5817 diodes in parallel, then that effectively halves the voltage loss of just a single diode on its own, Sometimes if you are making the pedal as optionally battery powered you hope to save on voltage drops, but like RG and others keep commenting, the battery powered effect pedals have pretty much gone the way of the dinosaur!

The idea to drop the 1N5817 sounds good. Just wondering what size voltage rating would be needed for the zener and the 39 ohm resistor? Thanks

[antonis]

Actually, for better noise suppression, the 100nF would be better placed as close as possible to pin 1 and 8 of the charge pump IC, not R1.

In such a case, Vref shouldn't be RF (or any HF) filtered..
(of course, they all depend on layout configuration..)

[bluelagoon]

Hi Antonis, its like you read my mind on that last question I had, posted before I asked. Cheers.

[antonis]

I know  if you get that desperate you can put two 5817 diodes in parallel, then that effectively halves the voltage loss of just a single diode on its own,

They should share CURRENT, not voltage..

P.S.
Actually, diode forward voltage drop DOES lower with smaller currents but in a, IMHO, consideration unworthy manner..



You can presicely estimate the forward voltage drop difference by solving Shokley diode equation for  different given currents..
(which, in your case of 2 parallel diodes is just 17.3mV..)

[FSFX]

Actually, for better noise suppression, the 100nF would be better placed as close as possible to pin 1 and 8 of the charge pump IC, not R1.

In such a case, Vref shouldn't be RF (or any HF) filtered..
(of course, they all depend on layout configuration..)

I really don't think you understand this at all. That is a charge pump and so needs the 100nF as close to it as possible. Please read all of the stuff published by Maxim and others to understand the issues with charge pump input noise..

[antonis]

I really don't think you understand this at all.

I think we're dealing with entirely different purposes cap implementation..

[FSFX]

I really don't think you understand this at all.

I think we're dealing with entirely different purposes cap implementation..

Not really, it is all about the correct way and best practice for reducing the noise on the power to the circuit.

[ElectricDruid]

I'd make C1 470μF/16V and place 100nF ceramic cap in parallel (as close to R1 right leg as physically possible..)

+1 agree. 39R/100u is a cutoff at 40Hz. Since this is only a -6dB/oct filter, that's barely any cut for 50/60Hz mains hum.
Changing to 470u would push that cutoff down another couple of octaves, and that improves the filtering quite a bit.
100n ceramic cap in parallel is always a good idea for the amount that it costs.

[Rob Strand]

I'd make C1 470μF/16V and place 100nF ceramic cap in parallel (as close to R1 right leg as physically possible..)

+1 agree. 39R/100u is a cutoff at 40Hz. Since this is only a -6dB/oct filter, that's barely any cut for 50/60Hz mains hum.
Changing to 470u would push that cutoff down another couple of octaves, and that improves the filtering quite a bit.
100n ceramic cap in parallel is always a good idea for the amount that it costs.

In practice it will be better that that because the hum from a rectified output is actually 100Hz/120Hz - I used loose 50Hz/60Hz numbers earlier on as well.

A big cap will help stop noise from the circuit itself getting in the audio.  That's more to do with current pulses on the circuit side and the size of the cap.  Layout can make a big difference as well, especially with DC/DC converters.