I know that my Search button is broken but even if it wasn't this has got to be too stoopid to look up 
Why do some GC kits have a second littler non-polarized capacitor piggybacked in parallel with the main big decoupling caps? The LM3875 "Typical Application" from the datasheet doesn't show this second cap on each side of the supply.
C5 and C6 in the attached GC kit schematic for example.
http://tech-diy.com/Chipamps/chipamp.jpg
Why do some GC kits have a second littler non-polarized capacitor piggybacked in parallel with the main big decoupling caps? The LM3875 "Typical Application" from the datasheet doesn't show this second cap on each side of the supply.
C5 and C6 in the attached GC kit schematic for example.
http://tech-diy.com/Chipamps/chipamp.jpg
To my understanding, the big caps provide the power needed for the amplifier to work.
The small caps clean the powersupply-lines by providing a short for high frequency-garbage to audio-ground.
Careless use of the small caps can decrease performance , by interaction with pcb-traces/big caps, creating a resonant circuit.
Good (low esr) elektrolyts of smallish value (like in a gainclone) can perform both tasks.
With kind regards,
Klaas
The small caps clean the powersupply-lines by providing a short for high frequency-garbage to audio-ground.
Careless use of the small caps can decrease performance , by interaction with pcb-traces/big caps, creating a resonant circuit.
Good (low esr) elektrolyts of smallish value (like in a gainclone) can perform both tasks.
With kind regards,
Klaas
Jade Factor
You divide the number of views over the number of replies to any thread.
If you look at something like Peter Daniel's showcase thread for his latest GC, the ratio is way out there like 205:1 which is OK. After all, Peter is not asking for help and it would be kinda kissa** for everyone to respond with adulations.
Now look at your standard plead for help though. The ratio ranges from like 20:1 to about 40:1...the more obscure or stoopid or poorly written, the higher the ratio. I'd expect that any thread I start from now on will easily garner 50:1 or higher
AndrewT is the only stalwart here willing to answer and continue the stoopidest threads for the good of furthering audio amp knowledge
You divide the number of views over the number of replies to any thread.
If you look at something like Peter Daniel's showcase thread for his latest GC, the ratio is way out there like 205:1 which is OK. After all, Peter is not asking for help and it would be kinda kissa** for everyone to respond with adulations.
Now look at your standard plead for help though. The ratio ranges from like 20:1 to about 40:1...the more obscure or stoopid or poorly written, the higher the ratio. I'd expect that any thread I start from now on will easily garner 50:1 or higher
AndrewT is the only stalwart here willing to answer and continue the stoopidest threads for the good of furthering audio amp knowledge
Carlos I fear most of us know that most of your findings are not yet the total answer, most of us know that we don't have the total answer to it, nor that we might ever. We also know that we all asked the same question, and are still pursueing this as should you. Giving you an abridged "answer" which might even convince you for the time being, stopping your search, would be a disservice.
In searching for more info I suggest you look up decoupling and bypassing on google maybe, always helps to have a handle... piggyback is not going to get you far.
Some of the concepts you will run into would be, that large caps (values) suck, and large electrolitycs suck even more. By bypassing them with a better quality cap like polypropylene or mica (which is only available in small values) will provide some of the qualities of the smaller cap to the combined one...
One of the qualities that one looks for is linearity, in laymans terms for capacitors this would mean, it would retain the same value regardles of the frequency it enqounters... I.e. a 1pf cap that would be a 1 pf cap at 1Hz and still at 10mHz would be extremely linear.
This is were electolytics and ceramics tend to fall out of hte tree badly.
Second thing you will see is impendance and ESR
Another thing you will see is that it is not only about keeping noise from the power supply out of your circuits , but keeping the noise from your circuits from getting to your clean power rails.
even the space between traces is of importance, with capacitive loads always finding the shortest path.
Also large caps are less effective at dealing with small high frequency ripples where the oposite is true for smaller values.
High frequency noise is more dense around it source and gets weaker as it gets furter away... visualise a bright light or candle, at the centre of the flame there is so much noise you can only see bright light, while at the edges there is less energy that made it this farand you can see through the flame or light's corona.
SO when people say to bypass as close as possible to power pins etc, that is exactly what they mean, the closer you get in the more of the nasty stuff you get to.
When you want to fill a baloon with water, you hold it to the tap, you don't open the tap up wide and hope to catch the water with the baloon - regardless how flexible your baloon is.
Hope that is worth keeping with you on the first stretch of the persuit to more knowledge.
In searching for more info I suggest you look up decoupling and bypassing on google maybe, always helps to have a handle... piggyback is not going to get you far.
Some of the concepts you will run into would be, that large caps (values) suck, and large electrolitycs suck even more. By bypassing them with a better quality cap like polypropylene or mica (which is only available in small values) will provide some of the qualities of the smaller cap to the combined one...
One of the qualities that one looks for is linearity, in laymans terms for capacitors this would mean, it would retain the same value regardles of the frequency it enqounters... I.e. a 1pf cap that would be a 1 pf cap at 1Hz and still at 10mHz would be extremely linear.
This is were electolytics and ceramics tend to fall out of hte tree badly.
Second thing you will see is impendance and ESR
Another thing you will see is that it is not only about keeping noise from the power supply out of your circuits , but keeping the noise from your circuits from getting to your clean power rails.
even the space between traces is of importance, with capacitive loads always finding the shortest path.
Also large caps are less effective at dealing with small high frequency ripples where the oposite is true for smaller values.
High frequency noise is more dense around it source and gets weaker as it gets furter away... visualise a bright light or candle, at the centre of the flame there is so much noise you can only see bright light, while at the edges there is less energy that made it this farand you can see through the flame or light's corona.
SO when people say to bypass as close as possible to power pins etc, that is exactly what they mean, the closer you get in the more of the nasty stuff you get to.
When you want to fill a baloon with water, you hold it to the tap, you don't open the tap up wide and hope to catch the water with the baloon - regardless how flexible your baloon is.
Hope that is worth keeping with you on the first stretch of the persuit to more knowledge.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.