I know someone is going to yell at me for building this thing as a first amp 
, but I built the Mick F amp (http://dogbreath.de/Chipamps/ThreeResAmp/ThreeResAmp.html) and included the optional 220r input resistor as well as used a 470r resistor to tie the signal input ground to the ground star.
I hooked it up to my Sandisk Sansa mp3 player and it sounds awesome! Even hooked up to a crappy old 10w shelf speaker!
The problems:
- If I hook my cell phone to it via the headphone jack, it only turns on for about half a second and immediately shuts off, but the volume sounds about right. It will only stay on if I crank up the volume all the way and then it sounds terrible and loud. Amp has to be power cycled to bring it back on.
- It won't work for the green speaker out on my computer (I don't have a preamp output, just the motherboard speaker/mic in/line in, no option I can see to use the line in as line out).
- It will stay on for the mp3 player all the way from 0 volume to max. Even if I pause the song.
- If I power off the mp3 player, the amp will shut off and has to be power cycled.
Additional info:
- Gain is set by 22k/680r to be around 30 from what I can see.
- The mp3 player puts out 1.6v+ at all time, even when paused.
- The phone headphone puts out around 4v when idle, but when playing music reads about 300mv
- The green output jack on the computer puts out anywhere from 200mv to about 800mv when a song is playing.
Is this due to something missing in the circuit? I think I'm flying under the input sensitivity of the chip with everything but the mp3 player, but how do I fix this to allow a greater range of inputs? Should the 220r input resistor be changed depending on what is going into the amp? Why does turning off the mp3 player but leaving it connected still require a power cycle to bring the amp back online?
Sorry for all the questions! I know (again) that someone will chastise me for my build choice, but at this point even if I build something else I HAVE to know why this doesn't work right!
(Edit: I was just reading Mick's page and found out he hates direct linking to his images and such. I'm sure this is more directed toward people trying to pass off his ideas as their own and selling on ebay, but I updated the link above to point to his page and not be a direct link to the schematic)
, but I built the Mick F amp (http://dogbreath.de/Chipamps/ThreeResAmp/ThreeResAmp.html) and included the optional 220r input resistor as well as used a 470r resistor to tie the signal input ground to the ground star. I hooked it up to my Sandisk Sansa mp3 player and it sounds awesome! Even hooked up to a crappy old 10w shelf speaker!
The problems:
- If I hook my cell phone to it via the headphone jack, it only turns on for about half a second and immediately shuts off, but the volume sounds about right. It will only stay on if I crank up the volume all the way and then it sounds terrible and loud. Amp has to be power cycled to bring it back on.
- It won't work for the green speaker out on my computer (I don't have a preamp output, just the motherboard speaker/mic in/line in, no option I can see to use the line in as line out).
- It will stay on for the mp3 player all the way from 0 volume to max. Even if I pause the song.
- If I power off the mp3 player, the amp will shut off and has to be power cycled.
Additional info:
- Gain is set by 22k/680r to be around 30 from what I can see.
- The mp3 player puts out 1.6v+ at all time, even when paused.
- The phone headphone puts out around 4v when idle, but when playing music reads about 300mv
- The green output jack on the computer puts out anywhere from 200mv to about 800mv when a song is playing.
Is this due to something missing in the circuit? I think I'm flying under the input sensitivity of the chip with everything but the mp3 player, but how do I fix this to allow a greater range of inputs? Should the 220r input resistor be changed depending on what is going into the amp? Why does turning off the mp3 player but leaving it connected still require a power cycle to bring the amp back online?
Sorry for all the questions! I know (again) that someone will chastise me for my build choice, but at this point even if I build something else I HAVE to know why this doesn't work right!
(Edit: I was just reading Mick's page and found out he hates direct linking to his images and such. I'm sure this is more directed toward people trying to pass off his ideas as their own and selling on ebay, but I updated the link above to point to his page and not be a direct link to the schematic)
Last edited:
Wondering whether the output of your phone is actually causing the amplifier to shut down? Many phones use filterless class d amplifiers to drive the headphone output and I am wondering if there is a lot of switching artifacts that are tripping the 3886 protection circuits. A possible solution would be to make something a like 25kHz low pass filter to put between the amp and phone.
DC into the input of the LM3886 is a no no. Make sure that you are AC coupled into the amplifier. (Coupling capacitors?)
I'm assuming you do not have access to an oscilloscope?
DC into the input of the LM3886 is a no no. Make sure that you are AC coupled into the amplifier. (Coupling capacitors?)
I'm assuming you do not have access to an oscilloscope?
Your assumption of my lack of access to an oscilloscope is completely correct!
The interesting thing about the phone theory is that if I crank the phone volume to max, the chip will come on and stay on (albeit with terrible sound that could be attributed to running the test with a 10w old speaker). Would the motherboard jack be doing the same thing?
The interesting thing about the phone theory is that if I crank the phone volume to max, the chip will come on and stay on (albeit with terrible sound that could be attributed to running the test with a 10w old speaker). Would the motherboard jack be doing the same thing?
You need an input capacitor between the amplifier and anything you connect to it. I would use a 10K input resistor (RG) and a 10uF coupling capacitor.. Given the large dc voltage present on the outputs of all of your devices orient the + side of cap towards the source device (phone, mp3 player, etc.)
You run the risk of damaging the speaker and potentially frying the LM3886..
You run the risk of damaging the speaker and potentially frying the LM3886..
Right now Rg is a 22k resistor that goes to ground, but there's also the 220r in line with the input voltage (I assume acting as sort of a filter since it really just cut out the clipping?). Do you suggest substituting a 10k resistor for the 22k to ground, or are you thinking the 10k resistor should replace the 220r that's in-line with the input signal? Similarly, I'm assuming the coupling capacitor should be in-line with the signal and not on the path to ground?
Thank you very much for the incredibly help assistance by the way! I was really afraid my question would be discounted to "build a different amp" and then I wouldn't get to use it as a learning opportunity
(Just in case anyone is reading and wonders what the values are on my schematic):
Part Value Rating, type Rg 22k 0.35 W metalfilm Rf 22k 1 W carbon Ri 680R 0.35W metalfilm Rm 10k 0.35 W metalfilm Cs 2200 uF
|| 100 nF
63 V electrolytic;
100 V MKS
Thank you very much for the incredibly help assistance by the way! I was really afraid my question would be discounted to "build a different amp" and then I wouldn't get to use it as a learning opportunity
(Just in case anyone is reading and wonders what the values are on my schematic):
Part Value Rating, type Rg 22k 0.35 W metalfilm Rf 22k 1 W carbon Ri 680R 0.35W metalfilm Rm 10k 0.35 W metalfilm Cs 2200 uF
|| 100 nF
63 V electrolytic;
100 V MKS
You don't specify if those voltages from the source are DC or AC. If DC that is REALLY bad, because you have a fully DC coupled amplifier, and are probably sending input voltage X gain to your speakers, which might be in the 10s of volts DC.
ALWAYS measure the DC voltage on the output before connecting any load, with the source connected. The output DC should be less than 100mV DC.
The amp should not care whether there is an input signal or not, but your sources are likely causing it to go into protection, unless something else is wrong with the circuit. Photos of *your* work, decent size and not blurry, would help tremendously.
ALWAYS measure the DC voltage on the output before connecting any load, with the source connected. The output DC should be less than 100mV DC.
The amp should not care whether there is an input signal or not, but your sources are likely causing it to go into protection, unless something else is wrong with the circuit. Photos of *your* work, decent size and not blurry, would help tremendously.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.