we can't do the sums or carry out the experiments to hear what you like.
You have the equipment and should know their specifications.
From these you determine the likely value of the components needed and try them. If you find too much treble or bass missing you change the appropriate component values. If on the other hand you hear no audio signal degradation or not sufficient attenuation of RF interference, then you may want to try component values that narrow the passband.
At some point you have to do it yourself.
That oftens means spending more time and money than buying pre-built equipment.
It certainly means having a stock of surplus components that proved not to be suitable, but may get used in the next project.
I can give you one pointer that may well start an argument with other builders.
The 47pF across the input socket can be a 2.5mm pin pitch C0G (=NP0) 50V ceramic.
You have the equipment and should know their specifications.
From these you determine the likely value of the components needed and try them. If you find too much treble or bass missing you change the appropriate component values. If on the other hand you hear no audio signal degradation or not sufficient attenuation of RF interference, then you may want to try component values that narrow the passband.
At some point you have to do it yourself.
That oftens means spending more time and money than buying pre-built equipment.
It certainly means having a stock of surplus components that proved not to be suitable, but may get used in the next project.
I can give you one pointer that may well start an argument with other builders.
The 47pF across the input socket can be a 2.5mm pin pitch C0G (=NP0) 50V ceramic.
Last edited:
as per your statement" I have my signal and power ground stars separated so I'm a little confused as to what grounds bypass to ground." i was a little confused .thank you for clarifying this.now on to resolve the rf issue :first is project in metal box of some sort,is it compleatly enclosed,are you close by the radio station you are receiving,if so how close and what frequency? also can you post a picture as it sits on your bench.as being a t.v tech what we would do trace such things is to take a cap(in this case .001 to .01)or resistor and clip lead and attach one end to ground and the other to the cap and poke around with the unconnected end of the cap on various pionts of the circuit but only do this on lower voltages say < 100v unless your confiedent and famliar with higer voltages and ALWAYS REMEMBER to keep to one hand in your pocket! yes im serious.also discharge the cap to ground before testing the next piont. even a 10pf cap charged to 30 or 40 volts can blow out a (the input to the)chip. meanwhile i will think of some more things to try.your circuit shouldn't be doing this but it is and i know how frustrating it can be.so hang in there!like an ol'boy once told me "it's easy when you know how"as for me ,i love a challenge.so keep posting and we'll figure it out. jer p.s. somtimes it takes a small cap from chassis to the neutral side of the line this done in most commercial stuff like stereo's and stuff
You would typically want to put an RF filter on the input, and the output, and the power connections, for an audio amplifier. Since RF wavelengths are relatively short, meaning that almost any bit of wire or PCB trace could act as an antenna, you would want to place the "RF-draining" component(s) of any filter (such as the capacitor in a simple RC low-pass filter) as close as possible to what you want to protect from the RF, such as the input pin of an opamp amplifier.
You will want to use "low-pass filters", which will tend to attenuate (reduce) signals from slightly below to everywhere above their "cutoff frequencies". There are very simple formulas for their cutoff frequencies.
A resistance, or an inductance (and a small series resistance, typically), in series with your signal, followed by a capacitance to ground, will act as a low-pass filter. (Sometimes, such as for a negative opamp input pin, with an RC low-pass filter, you would need to use a resistance on both sides of the capacitance to ground, to isolate the capacitance from the negative input, or, to make the filter bi-directional.)
I suggest, if you're up for it, looking at Walt Jung's book, "Op Amp Applications Handbook", which is available free, on line, in its entirety, at:
ADI - Analog Dialogue | Op Amp Applications Handbook
First you can find the bits related to your immediate questions. Then later you can try to absorb more.
To answer your specific questions about filter component values, you could start by downloading LTspice, from linear.com, for free, and simulate the low-pass RF filters' responses. After you get familiar with using it, it's quicker and easier than plugging component values into equations. And for L(R)C second-order filters, it's much faster and easier, especially when finding the R that smoothes-out the resonant peak inherent in simple LC low-pass filters.
Have fun,
Tom
You will want to use "low-pass filters", which will tend to attenuate (reduce) signals from slightly below to everywhere above their "cutoff frequencies". There are very simple formulas for their cutoff frequencies.
A resistance, or an inductance (and a small series resistance, typically), in series with your signal, followed by a capacitance to ground, will act as a low-pass filter. (Sometimes, such as for a negative opamp input pin, with an RC low-pass filter, you would need to use a resistance on both sides of the capacitance to ground, to isolate the capacitance from the negative input, or, to make the filter bi-directional.)
I suggest, if you're up for it, looking at Walt Jung's book, "Op Amp Applications Handbook", which is available free, on line, in its entirety, at:
ADI - Analog Dialogue | Op Amp Applications Handbook
First you can find the bits related to your immediate questions. Then later you can try to absorb more.
To answer your specific questions about filter component values, you could start by downloading LTspice, from linear.com, for free, and simulate the low-pass RF filters' responses. After you get familiar with using it, it's quicker and easier than plugging component values into equations. And for L(R)C second-order filters, it's much faster and easier, especially when finding the R that smoothes-out the resonant peak inherent in simple LC low-pass filters.
Have fun,
Tom
Yeah it's on a test flat sheet of aluminum. I pretty much got it to sound right with the sound card, ground was loose somewhere and killed the sound. So now I'm taking the amp off the test board and once it's soldered to the prototype board with all the components, it's going into a typical aluminum enclosure. I'll see if I'm still getting AM at that point.
Thank you so much for the post, I'll be sure to read the material and download the simulator. Currently I'm using multisim 10.1 and it's not exactly the best program to use for this kind of stuff.
thank you goootee.very well put!that is what i had suggested in the first place .then i got confused with the two seperated star ground systems.a simple small cap from the input to ground should do the trick.thats what i had to do on all 12 inputs on my mixer.it was 1990 when i did that. i'm sorry but i couldn't remember all of the details.speaking of details i was not informed of the other thread on this project and i had just now found it .i am now not blind anymore.
Great.
I forgot to mention that there is an LTspice discussion group, at yahoogroups.com, that is unbelieveably great. The message archive and file library are treasures. And there are tens of thousands of compatible device models that can be downloaded free. I have some links to them at Spice Component and Circuit Modeling and Simulation .
Tom
I forgot to mention that there is an LTspice discussion group, at yahoogroups.com, that is unbelieveably great. The message archive and file library are treasures. And there are tens of thousands of compatible device models that can be downloaded free. I have some links to them at Spice Component and Circuit Modeling and Simulation .
Tom
Actually yes, I sure did. I had low hertz hum that was fixed by tracking down a ground loop and radio along with the hum were completely eliminated by placing a 220pf cap between the + and - input pins.
Updated my build thread:
http://www.diyaudio.com/forums/chip-amps/158716-my-first-step-step-lm3886-illustrated-build.html
Hi
Thats good to know.
I recently built the t-network LM3875 circuit with a ground loop breaker. I don't have any hum but I do pick up interference when I move around over the top of the amp case. I tried installing 220pf and then a 330pf cap on the chip input pins but it doesn't make any difference. I still need to tidy up the wiring so the problem may disappear when done.
Nice pics by the way
Thanks man, I'm amazed by how good such simple setup sounds.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.