first of all i must admit i am new in the DIY thing (but i love it already). With to much time on my hand i decided to build a GC with the LM3886 dual mono kit from chip amp.com.
because of my little problem with concentrating reading is not a hobby an because of my trembling hands soldering is not either.....
But i managed to put it al together you can see the pics here..https://picasaweb.google.com/Xerxes75/GaincloneLM3886
i must say a few tears where there when it turned on.....it worked. i do however have slight problem.
the DC offset with the volume down is one side -32mV and on the other side -22mV? This DC thing is new for me so if anybody can turn the light on for me..please.
and then there is the massive volume! it plays with the volume slightly on very, VERY loud!. (turn down the volume on my iMac to play at a sensible volume)
the DC offset with the volume complete on is off the scale...
besides the kit there are two 225 2x25v toroids (gives 27.1v) on the rail i measure 37.8v (both sides) the source is my iMac with hrtII streamer.
because of my little problem with concentrating reading is not a hobby an because of my trembling hands soldering is not either.....
But i managed to put it al together you can see the pics here..https://picasaweb.google.com/Xerxes75/GaincloneLM3886
i must say a few tears where there when it turned on.....it worked. i do however have slight problem.
the DC offset with the volume down is one side -32mV and on the other side -22mV? This DC thing is new for me so if anybody can turn the light on for me..please.
and then there is the massive volume! it plays with the volume slightly on very, VERY loud!. (turn down the volume on my iMac to play at a sensible volume)
the DC offset with the volume complete on is off the scale...
besides the kit there are two 225 2x25v toroids (gives 27.1v) on the rail i measure 37.8v (both sides) the source is my iMac with hrtII streamer.
Recheck that you have all components in the right places and that the resistors are values that match the schematics (that I assume were sent with the kit).
If you don't want it so loud you can do one of two things. Put a log scale potentiometer on the input as a volume knob, or reduce the amp's gain by changing one or both of the resistor values shown on the schematic between the chip's output and its inverting input, they form a voltage divider that determines gain. This is detailed in the LM3886 datasheet linked below.
Do you have any DC offset when the audio source (iMac?) is not plugged in? Measure source output DC offset. The kits do not appear to include AC coupling (DC blocking) capacitors on the audio input which would remove DC offset before it is multiplied by the (chipamp) gain stage.
On the LM3886 datasheet you can see an example of placing a POT on the input in the typical application schematic, pg 1., and a 10uF coupling capacitor in the single supply application circuit pg 6.
http://www.ti.com/lit/ds/symlink/lm3886.pdf
If you don't want it so loud you can do one of two things. Put a log scale potentiometer on the input as a volume knob, or reduce the amp's gain by changing one or both of the resistor values shown on the schematic between the chip's output and its inverting input, they form a voltage divider that determines gain. This is detailed in the LM3886 datasheet linked below.
Do you have any DC offset when the audio source (iMac?) is not plugged in? Measure source output DC offset. The kits do not appear to include AC coupling (DC blocking) capacitors on the audio input which would remove DC offset before it is multiplied by the (chipamp) gain stage.
On the LM3886 datasheet you can see an example of placing a POT on the input in the typical application schematic, pg 1., and a 10uF coupling capacitor in the single supply application circuit pg 6.
http://www.ti.com/lit/ds/symlink/lm3886.pdf
I would leave it in!!! just to be sure no dc voltage gets to the inputs especially if you are using a passive preamp.
A passive preamp, if we are thinking about the same thing, will have no way to correct DC offset so if the music source outputs DC offset significant enough that the (power amp gain) * (DC offset) = (too high a DC offset to speakers), then you need an AC coupling cap "somewhere" in the audio chain.
Some people would build as if two subsystems are forever married together and only use one cap/channel, while others would treat the separate stages as separate components which might be used without each other and put a coupling cap on both of them. Generally speaking, since the higher the gain is the more a DC offset will matter, it is safer /more compatible to put the cap at the beginning of the (sub)circuit for your last, highest gain stage in the audio path from source to speaker.
Otherwise, I have not built that kit and don't have the schematic for it, consider contacting chipamp.com
It's sort of a shame chipamp.com doesn't provide enough support to even show good pics of the PCB or links to manuals. Are they THAT worried someone is going to copy their design or something? It's not as though there aren't plenty of other people sharing different but viable designs that if someone were included to use, would be a person wanting something other than a kit to build.
Some people would build as if two subsystems are forever married together and only use one cap/channel, while others would treat the separate stages as separate components which might be used without each other and put a coupling cap on both of them. Generally speaking, since the higher the gain is the more a DC offset will matter, it is safer /more compatible to put the cap at the beginning of the (sub)circuit for your last, highest gain stage in the audio path from source to speaker.
Otherwise, I have not built that kit and don't have the schematic for it, consider contacting chipamp.com
It's sort of a shame chipamp.com doesn't provide enough support to even show good pics of the PCB or links to manuals. Are they THAT worried someone is going to copy their design or something? It's not as though there aren't plenty of other people sharing different but viable designs that if someone were included to use, would be a person wanting something other than a kit to build.
Last edited:
There is a pdf article explaining the project in detail. They are using the default schematic on the national semiconductor website with few modifications.
DC offset upto 100mV positive / negative is fine.
Advice:
1. Use a 20K log potentiometer / stepped attenuator.
2. Use DC blocking input capacitors 1 microfarad and above polypropylenes will do.
3. Add some speaker protection / DC protection circuit. Cost $10 to protect those beautiful bookshelves.
DC offset upto 100mV positive / negative is fine.
Advice:
1. Use a 20K log potentiometer / stepped attenuator.
2. Use DC blocking input capacitors 1 microfarad and above polypropylenes will do.
3. Add some speaker protection / DC protection circuit. Cost $10 to protect those beautiful bookshelves.
Thanks for your reply. I must admit the support and THE manual could be better. But since i have the kit already thats à passend station for me. I am planning à second kit and thinking about à different manufacturer audiosector came to mind.
In THE manual http://chipamp.com/docs/lm3886-manual.pdf they are speaking of placing the feedback cap and placing THE resistor R3 differently.
Is this the Same principal as the 10uf cap?
About THE passive preamp, i want to keep THE amp part seperatly from it so i indeed am free to move about with the amps.
In THE manual http://chipamp.com/docs/lm3886-manual.pdf they are speaking of placing the feedback cap and placing THE resistor R3 differently.
Is this the Same principal as the 10uf cap?
About THE passive preamp, i want to keep THE amp part seperatly from it so i indeed am free to move about with the amps.
Can you tell us which is the exact circuit for your amp?
At least that way we can know what might be the problem, and not just guess it.
From the photos I can see you already have a motorized pot (value?), so that is half a passive-preamp. The other half is the input switch.
Another question: the 3886 is totally isolated from the heatsink? It should be.
At least that way we can know what might be the problem, and not just guess it.
From the photos I can see you already have a motorized pot (value?), so that is half a passive-preamp. The other half is the input switch.
Another question: the 3886 is totally isolated from the heatsink? It should be.
You have two issues. One is that the amps input bias current is connected to the potentiometer. As any current through a resistance produces a voltage drop across it, you get varying DC voltage as you turn the potentiometer. The solution for this is to use the input cap between potentiometer and amp, and to use an input resistor to ground that provides a stable path for the bias current to ground.
Now your non-inverting input has a fixed DC offset and you need to make sure that this is not amplified. For that purpose you need another capacitor in the feedback loop, which is called Ci in the datasheet.
Now your non-inverting input has a fixed DC offset and you need to make sure that this is not amplified. For that purpose you need another capacitor in the feedback loop, which is called Ci in the datasheet.
I'll do the maths for you.
32mVdc into 6r0 (assumed DCR) is ~0.17mW
Maximum amplifier output is assumed to be 28.3Vac (100W into 8r0).
Normal average listening level is ~1W allowing for a 20dB overhead for unclipped transient peaks.
A quiet passage during that listening session could be 20dB below your 1W average and the S/N ratio down there would be just 0.17mW/10mW = 17.7dB
The 32mV offset, when thought of as a noise source, is only 17dB below a quiet passage on a normal recording.
Very quiet passages or lower than 1W average listening could easily bring the wanted music signal to a level that is lower than the offset voltage.
Yes, the amplitude of the quieter passages in the music can be considerably lower than the displacement due to a 32mVdc output offset.
32mVdc into 6r0 (assumed DCR) is ~0.17mW
Maximum amplifier output is assumed to be 28.3Vac (100W into 8r0).
Normal average listening level is ~1W allowing for a 20dB overhead for unclipped transient peaks.
A quiet passage during that listening session could be 20dB below your 1W average and the S/N ratio down there would be just 0.17mW/10mW = 17.7dB
The 32mV offset, when thought of as a noise source, is only 17dB below a quiet passage on a normal recording.
Very quiet passages or lower than 1W average listening could easily bring the wanted music signal to a level that is lower than the offset voltage.
Yes, the amplitude of the quieter passages in the music can be considerably lower than the displacement due to a 32mVdc output offset.
Last edited:
Will it destroy the amp or the speaker? No.
Will the DC offset be audible? No.
Will it have audible effects? Yes.
It will displace the woofer or full-range speaker from its center. Excursion will not be symmetric anymore, leading to higher distortion from the speakers.
It will not displace the other drivers in a passive multi-way speaker, because the crossover will filter DC components out for them. The woofer will be misaligned in relation to the other drivers, producing phase shifts, leading to a more rippled frequency response around the crossing frequency, if you don't redesign the crossover. The smaller the woofer and the higher the crossing frequency, the worse is the effect.
It heats up the woofer coil, leading to power compression. As a result the woofer will produce less SPL for the same voltage, making it quieter than the other drivers in a multi-way system. Again the crossover needs to be adapted to the new situation.
So until now if have the following:
Is everything in the right place
Yes
Place a log potentiometer
An Alps 50k is going to be placed
Increase the resistor R3 (680=original)
Trying different once starting at 1k
Place a coupling cap
Laying around are some 2mfd ones
Buy speaker protection
Ordered the ones from EBay
Place the feedback cap at the Ci position
These came with the kit I will put them in place
From one of the technicians off a high end audio store I got a tip to place to Allen and Bradley resistors to middle the input...
A circuit is coming
Is everything in the right place
Yes
Place a log potentiometer
An Alps 50k is going to be placed
Increase the resistor R3 (680=original)
Trying different once starting at 1k
Place a coupling cap
Laying around are some 2mfd ones
Buy speaker protection
Ordered the ones from EBay
Place the feedback cap at the Ci position
These came with the kit I will put them in place
From one of the technicians off a high end audio store I got a tip to place to Allen and Bradley resistors to middle the input...
A circuit is coming
I am getting there...
I dismantled the hole thing and put I all back together with the writing above.
And it works!!
The dc offset is now 0.1 that is mV... The problems with the gain are also sort off over.
The problem now is that the amp has is bit to low on gain and it sounds a bit dull. But it is all detailed. Mmmhh perhaps a few tweaks will do....
I dismantled the hole thing and put I all back together with the writing above.
And it works!!
The dc offset is now 0.1 that is mV... The problems with the gain are also sort off over.
The problem now is that the amp has is bit to low on gain and it sounds a bit dull. But it is all detailed. Mmmhh perhaps a few tweaks will do....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.