jens .....
thanks for your input but this guy allready re arranged his pcb hopefully for the best and also your post doesnt answer my question .....
i was actually thinking that lets suppose for argument's sake that our friend managed to debug this circuit and no longer creating oscilation at 1MHZ or lower or higer by adding this capacitor or anything else .....
can anybody guarantee that when the amp is connected to the speakers via this or that quality cable will not pick up anything from this cables since most of these cables work as AM antenas ????
and lets suppose that after all no matter the quality of cables that actually they do pick up something in rf frequency will this be audible ???? and/or effect the amplifier's operation ????
thank you
thanks for your input but this guy allready re arranged his pcb hopefully for the best and also your post doesnt answer my question .....
i was actually thinking that lets suppose for argument's sake that our friend managed to debug this circuit and no longer creating oscilation at 1MHZ or lower or higer by adding this capacitor or anything else .....
can anybody guarantee that when the amp is connected to the speakers via this or that quality cable will not pick up anything from this cables since most of these cables work as AM antenas ????
and lets suppose that after all no matter the quality of cables that actually they do pick up something in rf frequency will this be audible ???? and/or effect the amplifier's operation ????
thank you
hmmmm
thread is quiet .....
is everybody gone on vocation ???? or what i ask is too strange to answer ????
lets see
thread is quiet .....
is everybody gone on vocation ???? or what i ask is too strange to answer ????
lets see
revised layout
Nordic, Jaycee, thanks for the comments.
I had placed the bias transistor so it could be coupled to the heatsink based on my experience with another design, which (now that I go back and look) used a darlington configuration for the drivers/outputs.
Here's a revised layout. I did not use roederstein caps as Nordic suggested as they would be pretty big.
Nordic, Jaycee, thanks for the comments.
I had placed the bias transistor so it could be coupled to the heatsink based on my experience with another design, which (now that I go back and look) used a darlington configuration for the drivers/outputs.
Here's a revised layout. I did not use roederstein caps as Nordic suggested as they would be pretty big.
Re: Re: Re: WAS THAT CORRECT ?????
I'll take a stab at it...
The setup - put a load on the output, drive the input with a test signal, and observe the output with a scope. RF could be picked up and introduced into the amp under test by the test setup, but in my experience (mostly very high frequency digital) this is not an issue with attention to proper grounding of the scope probes or other test connections. I have chased ghosts before but every time it has turned out to be lack of proper ground or otherwise poor technique in connecting to the signal of interest. Probing a signal can make things appear that aren't normally there, or make things that are there disappear. One has to think about how introducing the measurement connection into the circuit will change the circuit. Fortunately with modern test equipment (high impedance, very low capacitance scope probes for example) this is not such a big issue any more.
A tougher thing to quantify is the possibility of the parasitic r-l-c elements that are a result of any layout, and how they might contribute to pickup and amplification of noise or cause oscillation. Here is where good design practice (proper bypassing and grounding for example) eliminates most of the problems. And I would expect a low pass filter at the feedback point to be a necessity. But I am not an expert.
Is 1MHz noise riding on top of an audio signal audible? I don't know, I've never tried to make a comparison. Certainly some of the energy at frequencies outside the audible region will make it down the cable to the speaker, and to the tweeter.
sakis said:
sakis said:
I'll take a stab at it...
The setup - put a load on the output, drive the input with a test signal, and observe the output with a scope. RF could be picked up and introduced into the amp under test by the test setup, but in my experience (mostly very high frequency digital) this is not an issue with attention to proper grounding of the scope probes or other test connections. I have chased ghosts before but every time it has turned out to be lack of proper ground or otherwise poor technique in connecting to the signal of interest. Probing a signal can make things appear that aren't normally there, or make things that are there disappear. One has to think about how introducing the measurement connection into the circuit will change the circuit. Fortunately with modern test equipment (high impedance, very low capacitance scope probes for example) this is not such a big issue any more.
A tougher thing to quantify is the possibility of the parasitic r-l-c elements that are a result of any layout, and how they might contribute to pickup and amplification of noise or cause oscillation. Here is where good design practice (proper bypassing and grounding for example) eliminates most of the problems. And I would expect a low pass filter at the feedback point to be a necessity. But I am not an expert.
Is 1MHz noise riding on top of an audio signal audible? I don't know, I've never tried to make a comparison. Certainly some of the energy at frequencies outside the audible region will make it down the cable to the speaker, and to the tweeter.
i am not really with you
the truth is no matter your detailed explanation i still dont get this ......
and even if this is truth i need more facts to translate this to audible diference ....
i am going on vocation sometime arround the 23 august but when i am back i will look really deep in to this ....
thanks ....i am not solved yet
the truth is no matter your detailed explanation i still dont get this ......
and even if this is truth i need more facts to translate this to audible diference ....
i am going on vocation sometime arround the 23 august but when i am back i will look really deep in to this ....
thanks ....i am not solved yet
Small distance between copper lines behaves alike capacitors
So... there are hundreds of 5 to 10 pf capacitors into this board...as two metal plates placed near one each other represents an air core capacitor.... the biggest the copper line
length and the closest they are , the bigger will be the capacitor value in picofarads.
The copper lines have also inductances..... hundreds of capacitors of low value, if added, may represent a value that will work together with the inductances to create a "tuned circuit"... and this Radio Frequency oscilating system may create 1 Megahertz (or any other frequency) oscilations.
This is the basic work of a transmitter.... even cheap transistors, slow in speed units, alike 2N3055 are able to oscilate (operate) into 3 Megahertz..... so...if you have a tune circuit into this frequency, connected into one transistor base...then you may have the oscilation.
This board seems no good... distance between copper lines seems to be very small....maybe the defect is beeing generated in other part then the board...but this board is very suspectious.
Some of those capacitances will be good... will stabilize your circuit or will represent filters to ground, sending high frequencies to ground and creating high stability....., but... depending were the capacitor is, and how big it is... also can UNstabilize the amplifier.
Carlos
So... there are hundreds of 5 to 10 pf capacitors into this board...as two metal plates placed near one each other represents an air core capacitor.... the biggest the copper line
length and the closest they are , the bigger will be the capacitor value in picofarads.
The copper lines have also inductances..... hundreds of capacitors of low value, if added, may represent a value that will work together with the inductances to create a "tuned circuit"... and this Radio Frequency oscilating system may create 1 Megahertz (or any other frequency) oscilations.
This is the basic work of a transmitter.... even cheap transistors, slow in speed units, alike 2N3055 are able to oscilate (operate) into 3 Megahertz..... so...if you have a tune circuit into this frequency, connected into one transistor base...then you may have the oscilation.
This board seems no good... distance between copper lines seems to be very small....maybe the defect is beeing generated in other part then the board...but this board is very suspectious.
Some of those capacitances will be good... will stabilize your circuit or will represent filters to ground, sending high frequencies to ground and creating high stability....., but... depending were the capacitor is, and how big it is... also can UNstabilize the amplifier.
Carlos
Attachments
- Status
- Not open for further replies.