Power Supply Filter Anatomy

Started by POTL, January 08, 2017, 05:25:17 PM

Previous topic - Next topic

GibsonGM

Quote from: EBK on July 04, 2017, 12:50:23 PM
Nah, lamps aren't my thing (which means I won't yell at anyone using them :icon_wink:).  I was going to suggest a vibrating motor or piezo buzzer might make a good reverse polarity deterrent.  Perhaps even whatever fog stuff model railroad enthusiasts use for realistic-looking steam engines.  Maybe that is the way to address the problem.   :icon_biggrin:

I like it!

In my world, tho, you only get a 1N400x diode across the power inputs to the board.  That's it..you have 10 seconds to figure it out...   :)
  • SUPPORTER
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

POTL

http://www.smallbear-electronics.mybigcommerce.com/resettable-fuses-2/

What do you think about installing resettable fuse?
If you add it before the diode, and the resistor is removed or left, but after the diode, to prevent it from burning out?
What are the advantages of such a fuse and will there be a voltage drop?

Which parameters should be considered for pedals working at 9-18v and 100ma (or less)

antonis

"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

POTL

Quote from: antonis on September 06, 2017, 06:12:22 AM
RTFM...  :icon_wink:

https://archive.org/details/Raychem-PolySwitchResettableFusesforCircuitProtectionOCR

Thanks, but
I do not even have to copy the text, not all technical terms in English are familiar to me
Translation takes hours = (
you can hear from you the compressed information
I like it when it's short and clear  :D

amptramp

If you are using filtering from an AC supply, this calculator:

http://www.duncanamps.com/psud2/

can be quite useful.  If you are powered by a wall wart, you can sacrifice one (if they are cheap enough) to determine what is inside and fill in the values.  Don't be surprised if a wall wart has an electrolytic capacitor with a resistor in series with the cap.  I did find that in one wall wart where filtering quality was a secondary consideration.

POTL

#85
Quote from: amptramp on September 06, 2017, 04:21:04 PM
If you are using filtering from an AC supply, this calculator:

http://www.duncanamps.com/psud2/

can be quite useful.  If you are powered by a wall wart, you can sacrifice one (if they are cheap enough) to determine what is inside and fill in the values.  Don't be surprised if a wall wart has an electrolytic capacitor with a resistor in series with the cap.  I did find that in one wall wart where filtering quality was a secondary consideration.
Hey.
I'm a beginner pedal builder and do not use AC projects, all I do is work from DC power supplies within 9-18 volts.
In this thread, I was given excellent advice and helped to understand the principle of this filter (I really do not like to do something if I do not understand it)
In general, the power filters and reverse polarity protection look very good.
However, I was recently told that it makes sense to install a fuse instead of / together with a resistor 10R-100R.
I looked through a lot of schemes of boutique and serial manufacturers, but I found fuses only from JHS (now they refused it).
Now I wondered about the advisability of installing a fuse.

Earlier with similar fuses, I had no business (worked only with glass).
I'm interested in the principle of work.
As I understand it in case of wrong polarity it will not burn as a resistor and will not allow to burn the power supply unit (for example if there is no resistor in the circuit).

I wonder how the fuse will work (turn off the power circuit or send current to the power supply), whether it should replace the resistor or the resistor will remain in the circuit (remaining part of the filter).
Does it give a voltage drop or leave me exactly 9-18 volts without falling?
Maybe it's just an extra detail in the effects pedals (after all, almost no one uses it).

Phoenix

All fuses have SOME resistance, so there will always be some voltage drop. One time fuses like glass fuses will have lower voltage drop than a resettable fuse like a polyfuse. You'll need to check datasheets to find out specifics. You COULD put the fuse inside a feedback loop to correct for voltage drop, but that's getting unnecessarily complicated.

It is difficult to fuse anything in the low milliamp range like we deal with in low voltage pedals, so it is much better (and simple, and reliable) to current LIMIT the power supply. Simple 3 pin linear regulators have thermal shutdown limiters which roughly translates to some current limit, which is somewhat useful, but the adjustable types (like the LM317) can also easily be wired in a different configuration for current limiting, for much more precise control - check the datasheets for example circuits and the formulas needed to calculate values.

I'd suggest that the simplest manner for you to proceed is a simple linear power supply, followed by an adjustable linear regulator wired for current regulation, finally followed by a typical 7809 or whatever linear voltage regulator. This means that if too much current is pulled through the 7809, the current regulator that preceeded it will shut off the voltage. Simple.

POTL


7809 does not fit, since I develop effects with multi-power 9-18 volts
since I'm worried about the reverse polarity, and I always put the components with a margin (most of them do not grow anymore, and those that become larger, their size does not change much).
If the fuse will reduce the voltage as 5817, I also do not fit. I prefer to work with effects on JFET / Germanium - they do not really like dropping or changing the bias point  :)

PRR

The chemical fuses may take some Seconds to trip.

Transistor death happens much quicker.

All old transistor workers KNOW that transistors always blow to protect the fuses.
  • SUPPORTER

Phoenix

Quote from: POTL on September 06, 2017, 09:38:57 PM7809 does not fit, since I develop effects with multi-power 9-18 volts

So what are we talking about? Power supplies that are voltage regulated and have a current limit? Or reverse polarity protection (and over current protection???) to be built into pedals? These are two very different topics, and the topic title is Power Supply Filter Anatomy... If the discussion has moved on, you should probably start a new topic if you want pertinent responses.

If you want reverse polarity protection for pedals that is better than the frankly useless reverse biased diode, but you can't afford even 0.3V drop from a schottky, then check out RG's polarity protection circuits Advanced Power Switching and Polarity Protection for Effects and A cheap - and good - polarity protector.

As for over-current protection/fusing of pedals, that shouldn't be necessary if you design it so that it should not fail under any reasonable eventuality - protect the input and the output. Series resistance on both input and output, clamping diodes/leds/zeners/transorbs/whatever on the input if you're paranoid. Nothing in a normal pedal should be operating anywhere near the SOA of any of the devices, so over-current is not something you should ever need to worry about.

If you also want to protect against over-voltage, RG's cheap and good circuit can be easily modified to handle that too in addition to the reverse polarity protection.

POTL

#90
on the power filter I was perfectly consulted and there are no questions with him.



Now I again began to worry about protection from reverse polarity.
Earlier we discussed protection with Schottky diode 1N5817 - its main disadvantage was the voltage drop (something about 0.5V) that is not suitable for me, since my germanium and fet prototypes start to work incorrectly under the voltage below 8.7-8.6)
circuit with a rectifier diode 1N4001 and a resistor in front of him I like, but many write that it is unreliable.
As I understood that if the Schottky does not cope, then it burns and the current does not go to the circuit (hence it remains alive).
And if 1n4001 burns, then the current goes farther and the entire circuit burns, for this purpose a 10-100 ohm resistor is placed before the diode, which acts as a fuse, it burns down and the current beyond the residuals of the resistor does not pass.
In general, my question is as follows.
I do not want to add any other components (in the form of transistors)
I do not want the voltage drop.
I'm interested.
1) If there is a resistor in front of the diode in the circuit with 1N4001, it 100% works as a fuse and the circuit remains 100% alive (only the resistor requires replacement)
2) Does it make sense to put the resettable fuse in addition to the resistor (or replace the resistor with a fuse)?
Will this method be more reliable and will I not lose tension?
Let me remind you that all the circuits with which I work have a voltage of 9-18v and do not exceed 100mA

bluebunny

Check this out if you're concerned with polarity protection and worried about diode losses and the possibility of the sacrificial diode shorting when dead.
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

Phoenix

#92
Quote from: POTL on September 07, 2017, 09:40:46 AM
1) If there is a resistor in front of the diode in the circuit with 1N4001, it 100% works as a fuse and the circuit remains 100% alive (only the resistor requires replacement)
NO. Like PRR said above, the transistors will blow to protect the fuse. And a resistor is not even intended as a fuse! It's delay characteristics are far worse, so that whole concept is useless. If you're lucky, the resistor will burn before the diode does, but that's certainly not garuanteed. And do you really want to be relying on a resistor burning inside your enclosure for protection? It'll get soot and ash everywhere, and it'll burn any circuitboard it's mounted to, or melt any plastic devices it's near, it's just an all around ugly solution, even if it did work 100% of the time.

Quote from: POTL on September 07, 2017, 09:40:46 AM
2) Does it make sense to put the resettable fuse in addition to the resistor (or replace the resistor with a fuse)?
Will this method be more reliable and will I not lose tension?
Let me remind you that all the circuits with which I work have a voltage of 9-18v and do not exceed 100mA
Good luck even FINDING a polyfuse (resettable fuse) below 100mA (and a fuse must be rated only marginally above nominal current requirements in order to function properly - putting a 100mA fuse would be useless for protecting a circuit that draws 10mA), and those will have quite a high resistance. You're very worried about voltage drop, so that will be an unacceptable solution.

Why are you even worried about fusing your pedals? Who told you you need to?

Also, if your circuits stop performing correctly at 8.6V, then something is very wrong. That might be fine if they will always be powered by a power supply, but if they're running from batteries, you'd only be using about 5% of the capacity of an alkaline battery. Ideally 9V battery powered items should run down to at least 7V to get 80% of the capacity of the battery, or 6V to get 95%.

You seem to want to get all these features for free: "I do not want to add any other components" "I do not want the voltage drop"
Well, simply put, you CAN'T get something for nothing (though sometimes the chicks are free).

EDIT:
Haha Marc, I posted that a couple of messages above, he dismissed it out of hand unfortunately.

bluebunny

Quote from: Phoenix on September 07, 2017, 10:15:39 AM
Haha Marc, I posted that a couple of messages above, he dismissed it out of hand unfortunately.

Ack!  So you did!  I wasn't wearing my glasses...   ;D
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

POTL

Quote from: Phoenix on September 07, 2017, 10:15:39 AM
Quote from: POTL on September 07, 2017, 09:40:46 AM
1) If there is a resistor in front of the diode in the circuit with 1N4001, it 100% works as a fuse and the circuit remains 100% alive (only the resistor requires replacement)
NO. Like PRR said above, the transistors will blow to protect the fuse. And a resistor is not even intended as a fuse! It's delay characteristics are far worse, so that whole concept is useless. If you're lucky, the resistor will burn before the diode does, but that's certainly not garuanteed. And do you really want to be relying on a resistor burning inside your enclosure for protection? It'll get soot and ash everywhere, and it'll burn any circuitboard it's mounted to, or melt any plastic devices it's near, it's just an all around ugly solution, even if it did work 100% of the time.

Quote from: POTL on September 07, 2017, 09:40:46 AM
2) Does it make sense to put the resettable fuse in addition to the resistor (or replace the resistor with a fuse)?
Will this method be more reliable and will I not lose tension?
Let me remind you that all the circuits with which I work have a voltage of 9-18v and do not exceed 100mA
Good luck even FINDING a polyfuse (resettable fuse) below 100mA (and a fuse must be rated only marginally above nominal current requirements in order to function properly - putting a 100mA fuse would be useless for protecting a circuit that draws 10mA), and those will have quite a high resistance. You're very worried about voltage drop, so that will be an unacceptable solution.

Why are you even worried about fusing your pedals? Who told you you need to?

Also, if your circuits stop performing correctly at 8.6V, then something is very wrong. That might be fine if they will always be powered by a power supply, but if they're running from batteries, you'd only be using about 5% of the capacity of an alkaline battery. Ideally 9V battery powered items should run down to at least 7V to get 80% of the capacity of the battery, or 6V to get 95%.

You seem to want to get all these features for free: "I do not want to add any other components" "I do not want the voltage drop"
Well, simply put, you CAN'T get something for nothing (though sometimes the chicks are free).

EDIT:
Haha Marc, I posted that a couple of messages above, he dismissed it out of hand unfortunately.

I tried the tone bender - it oscillated when the battery was sitting down.
I also tried the Marvel Drive - it seemed to me that the sound was somewhat loose and not as pleasant as the full 9v.
I'll try to try Marvel today with 1n5817.
The circuit with transistors is certainly interesting, I will take it into service, but this is a large number of details, and I try to make the case as small as possible and such a large protection will take up a lot of space, in comparison with the classical methods.
Please, once again explain about its pluses and minuses (except for voltage drop).

Phoenix

Quote from: POTL on September 07, 2017, 11:07:54 AMI tried the tone bender - it oscillated when the battery was sitting down.
If your tone bender was oscillating with a battery, but not with a power supply, then there was something wrong with the circuit.

Quote from: POTL on September 07, 2017, 11:07:54 AMI also tried the Marvel Drive - it seemed to me that the sound was somewhat loose and not as pleasant as the full 9v.
Look up psychoacoustics and Fletcher Munsen (equal loudness) curves. The only thing a couple of volts less supply voltage should do to this circuit is limit headroom which can mean you're getting more distortion with the gain knobs set at the same level as they were for higher supply voltage (clipping distortion is a type of compression, fix this by turning down the gain controls to achieve the same level of distortion, don't fixate on what the dial says), and less availale output volume (which can be perceived as less attack - but you can compensate for that by simply turning up your amp to make up for the reduced output from the circuit).

Quote from: POTL on September 07, 2017, 11:07:54 AMThe circuit with transistors is certainly interesting, I will take it into service, but this is a large number of details, and I try to make the case as small as possible and such a large protection will take up a lot of space, in comparison with the classical methods.
Please, once again explain about its pluses and minuses (except for voltage drop).
It really doesn't take up much space at all, it's only 6 extra components over any other polarity protection, and if you're really stretchhed for space, do it with surface mount components - using SOT23 or SOT323 transistors is easy to hand solder, works nicely on perf, and then you could fit the whole circuit in a 5x5mm space easily.

Pluses: Extremely good reverse polarity protection, even against high voltage reverse polarity.
Low voltage drop, down to sub 10mV if you design accordingly.

Minuses: A few extra parts.

EBK

I still think reverse polarity deterrence may be more effective than reverse polarity protection.   :icon_twisted:
  • SUPPORTER
Technical difficulties.  Please stand by.

POTL

The best protection is a symbol with polarity

And indicating the permissible voltage and type of current (AC / DC)
on the front of the pedal
But we know that now no one likes to read  ;D

EBK

Just put something like this near the polarity marking.  :icon_razz:
  • SUPPORTER
Technical difficulties.  Please stand by.

POTL

Everyone will be afraid of these boxes  ;D
In general, I read a few more sources, if 5817 does not spoil the sound and will not burn, I'll redo the schemes for it :)