Filtering switching noise with ferrite beads and linear regulator

[ramicio]

Why does anyone do anything?  Because they want to.  I want an LED audio spectrum analyzer in a 5.25" drive bay that has fast, fluid response.

[Niklas]

I checked the picture of the PCB available at:
http://www.youritronics.com/forum/download/file.php?id=92&mode=view

There can be stability issues with SMPS circuit, up in the right corner, related to poor layout with long traces. A workaround can be to remove the output inductor L1 and to patch in an LM7805 with proper bypass and buffer capacitors. Ceramic (X7R) or film 100nF in parallell with 100uF electrolytic on both the input and the output of the LM7805. In the attached picture the patch wires are shown but not the capacitors on the voltage regulator.

Perhaps the TO220 package of the LM7805 can be thermally connected to the large copper fill on the LED display board for heat sinking?

[KJDS]

Run it off a battery. If that cleans it up then it's a power supply issue

[ramicio]

I don't own any batteries.

[c4757p]

If you won't even buy a battery then we're back to "string, chewing gum and wishful thinking", and if you can't afford a battery you're not going to make this work no matter how much you'd like to. That's like saying "here's $5, how can I build a spaceship?"

[c4757p]

Why does anyone do anything?  Because they want to.  I want an LED audio spectrum analyzer in a 5.25" drive bay that has fast, fluid response.

Design it yourself. You'll learn a hell of a lot more that way.

[ramicio]

Stop making assumptions.  I don't own any batteries because I find them useless.  The only portable electronics I use is my phone.  Everything else I like to be corded.

[c4757p]

I wasn't making assumptions, I was listing a possibility so I wouldn't have to come back and address it if it turned out correct.

I don't own any batteries because I find them useless.

You've now been proven wrong, as we've found a use for them, so go buy some.

[ramicio]

Why does anyone do anything?  Because they want to.  I want an LED audio spectrum analyzer in a 5.25" drive bay that has fast, fluid response.

Design it yourself. You'll learn a hell of a lot more that way.

I can't learn it, therefore I don't want to learn it.  If I have to ask how to design a simple filter, do you really think I could learn to code some microcontroller?  I bought into the PIC thing thinking there would be tutorials that were easy, but it was beyond me.  I had to pay someone to write me a program to do a simply on/off with a button.

[c4757p]

I can't learn it, therefore I don't want to learn it.

 

If I have to ask how to design a simple filter, do you really think I could learn to code some microcontroller?

 

Do you know what a "beginner" is?? Of course there are things you don't know! Stop being lazy and keep trying! Fuck, do you think I was born knowing about electronics and programming?

You need to pick up a book. That's the only way you're going to solve your problems. And if you had to pay someone to write you a program to "do on/off with a button" you're not even trying.

[c4757p]

These things are hard. I have years of experience with both (more experience with programming, actually) and that's how I can see solutions to problems like this. Go work your ass off and get the experience! No, you're not going to have this done today. Probably not this year. Such is humanity, I'm afraid; you can't just lay your head on a textbook or your computer screen and expect the knowledge to embed itself into your brain through osmosis. There are no 'easy' tutorials because programming isn't easy.

[ramicio]

It's not about being a beginner.  I know the limitations of my intelligence, and this is not something I will ever be able to do.  I am not unrealistic in this regard.  I don't waste my time with things I am naturally not good at.  It's not about knowing something, it's about being able to understand it.  Again, making more assumptions like "you're not even trying."  Some people are born with artistic abilities, and some people are born with technical abilities.  A person who can't draw or paint will never be able to learn to do so.  Their brain is just not wired for it.  I can tell you that I am not intelligent enough to understand how to do this.  I absolutely cannot pick up a book and learn things.  I failed every area in school that required me to read something to learn it.  I only learned anything by hearing it being broken down by another human.  I am not being lazy.

[c4757p]

I know the limitations of my intelligence

You do not.

I don't waste my time with things I am naturally not good at.

Some people would call this "effort".

It's not about knowing something, it's about being able to understand it.  Again, making more assumptions like "you're not even trying."  Some people are born with artistic abilities, and some people are born with technical abilities.  A person who can't draw or paint will never be able to learn to do so.  Their brain is just not wired for it.

Prove it. You brain does not have wires. It is conditioned by your prior experiences. You can recondition it, it's just really hard. Doing really hard things when they need to be done is the sign of not being lazy.

I absolutely cannot pick up a book and learn things.  I failed every area in school that required me to read something to learn it.  I only learned anything by hearing it being broken down by another human.  I am not being lazy.

Then the sources you had did not explain it in a way you understood. Find others.

[ramicio]

How do I not know the limitations of my intelligence?  You're getting downright insulting.

I don't waste my time with stuff I am not good at because I know the end result will be far from my standards, so it's not worth doing.  It's all or nothing with me.  I would rather channel my effort into things I am good at.  Some people might call this sound logic.  The world is too full of people lately who think they are good at something but actually such at it because trendy yuppies have taken control of society and criticism is taboo as long as there is sound effort.  Natural talent is now ignored and people are taught that they can be great at something by merely putting in the effort.

I don't need to prove anything to you.  If you can't understand this through observation in every day life, then you just aren't looking.  Some people are good at math, some people suck at it.  Numbers are a certain way of thinking and people who don't understand it cannot have it taught to them no matter how good the teacher is or how they try to alter the interpretation of it to fit the student.  Some people are artists, some are not.  It's the way the brain is wired to be able to coordinate hand movements to what the eye sees.  You cannot be taught how to be an artist.  Some people can remember trivia and other other useless knowledge, and some people can't remember the name of the person they've just met.

I looked at many sources.  It had nothing to do with how it was explained, it's that my brain does not operate in the manner required to be a programmer.

[KJDS]

I wasn't making assumptions, I was listing a possibility so I wouldn't have to come back and address it if it turned out correct.

I don't own any batteries because I find them useless.

You've now been proven wrong, as we've found a use for them, so go buy some.

So I've come up with the perfect way of identifying if it's the power supply that's the issue and the OP won't spend the price of a beer on some batteries, but it's our fault for not designing a circuit for him to do something but we don't know what it needs to do.

[ramicio]

It's already been quoted here that the designer knew it was the power supply on the thing.  I didn't ask anyone to design me a circuit.  If I would have you would see offers being made to pay people to do so.  I wouldn't have to buy just batteries, I would have to also buy something to put them into.  It's not about money, it's about practicality.  I don't own any method transportation, so I would have to order such things and pay for shipping.

[free_electron]

The designer of that board has given a solution. Get a 7805 regulator , wire it up as shown and it will work. A 7805 cost 0.5$.

The designer also gave an explanation. The switching power supply he designed is faulty. The core saturates and this collapses the output voltage for a short moment. Most likely the atmel keeps running but the analog portion goes crazy and this shows up as a glitch. Anyway.

Get a 7805 , wire it up as shown and it will work.
No kind of filtering before the regulator can help as the problem is in the regulator backend.

And dont come with excuses as i dont have money i do t have transport i dont have yadadayadayada. You got this spectrum analyser and you have a computer . So you have money to buy those and you managed to get them at your place which implies a form of transport.

[c4757p]

How do I not know the limitations of my intelligence?  You're getting downright insulting.

I don't waste my time with stuff I am not good at because ...

...it's that my brain does not operate in the manner required to be a programmer.

Every word of that was utter bullshit. Your brain can be trained to operate that way, just as anybody's brain can learn to be an artist. You can teach somebody to be an artist. You cannot teach somebody who does not care, which is, I suspect, the cause of this common misconception. You refuse to learn, so you will not learn. And yes, I am getting insulting, because I'm getting impatient.

You have a problem and a solution, implement it and you're done. But stop expecting everybody to solve your problems. You're not incapable of learning this stuff. If you don't want to bother with it I can understand, but don't try to do it, then refuse to do the actual "work" part of it.

[ramicio]

I'm most likely just done with this project.  The fact that this thing does not have fine bands puts me off.  It basically just goes up and down with beats.  It's basically a VU meter repeated to 20 columns.  Look how the thing acts with a steady 1 kHz tone played to it.  Not only does it register on all bands, but it bounces from a steady tone.

http://youtu.be/FqhcEvGjE84

I would take it upon myself to make something better myself, but as I said before I am just not capable of it and it would cost me too much money.  It would only make sense for me to make something to produce and to profit from, but it's such a niche and expensive thing that it's not worth pursuing.  I have a feeling this guy just did this for some school project and decided to make some extras and sell them.  If it was a serious product he wouldn't make statements indicating that it's not worth his time and money to improve it.

[c4757p]

The core saturates and this collapses the output voltage for a short moment.

Only slightly off topic - any idea why this would happen every few seconds? I don't think I've ever seen a DC-DC converter that saturates its core just occasionally on that sort of time scale.

[c4757p]

I'm most likely just done with this project.  The fact that this thing does not have fine bands puts me off.  It basically just goes up and down with beats.  It's basically a VU meter repeated to 20 columns.  Look how the thing acts with a steady 1 kHz tone played to it.  Not only does it register on all bands, but it bounces from a steady tone.

http://youtu.be/FqhcEvGjE84

That's bad.

I would take it upon myself to make something better myself, but as I said before I am just not capable of it and it would cost me too much money.

I'm not going to keep telling you you can do it, because you don't seem to believe me, but if you change your mind, I think your idea of doing it with FFT was solid. A dsPIC microcontroller has all the facilities you need to do this at audio frequency. It will not be an easy project for a beginner, but it is definitely doable.

[oPossum]

The core saturates and this collapses the output voltage for a short moment.

Only slightly off topic - any idea why this would happen every few seconds? I don't think I've ever seen a DC-DC converter that saturates its core just occasionally on that sort of time scale.

Possibly beat frequency between regulator glitch and ADC sampling. The ADC would have to sample at just the right time to see the power supply glitch that is likely quite regular.

[c4757p]

I don't own any method transportation, so I would have to order such things and pay for shipping.

How does food get to your kitchen? I'm going to assume you don't slaughter your own cows? Use that method to acquire batteries.

[c4757p]

Possibly beat frequency between regulator glitch and ADC sampling. The ADC would have to sample at just the right time to see the power supply glitch that is likely quite regular.

Ah - good point. I forgot there's no oscilloscope here to show the true glitch frequency.

[ramicio]

Every word of that was utter bullshit. Your brain can be trained to operate that way, just as anybody's brain can learn to be an artist. You can teach somebody to be an artist. You cannot teach somebody who does not care, which is, I suspect, the cause of this common misconception. You refuse to learn, so you will not learn. And yes, I am getting insulting, because I'm getting impatient.

You have a problem and a solution, implement it and you're done. But stop expecting everybody to solve your problems. You're not incapable of learning this stuff. If you don't want to bother with it I can understand, but don't try to do it, then refuse to do the actual "work" part of it.

You're just spouting a bunch of trendy new age nonsense.  Every person is born with strengths and weaknesses.  You cannot teach a person to be an artist, but a person can be educated on forms of art.  It's not even really education, it's just a person seeking what previous artists they find inspires them.  No amount of teaching is going to make a person who can only draw stick figures into some kind of Michelangelo.  Your kind of thinking is why art is turning into crap.  We have people who suck at what they do because their mommies and daddies never told them that they have no talent.  They are just taught that all that matters in life is being happy.  It's blind optimism and it's a mental illness.  Well, spending time of stuff I do not understand and therefore can't perfect sure doesn't make me happy, so I spend time doing things I do understand, and therefore enjoy.  I am logical.  I don't purposely go around frustrating myself.  Only masochists constantly invite pain and frustration into their lives.  I have been on this earth long enough to know what I am good at and what I suck at.