|22nd November 2011, 07:56 PM||#11|
Join Date: Nov 2007
1). Television and movies? Adapt speaker tone to suit chipamp.com amplifier
2). Peaceful quiet analytical listening with exaggerated detail? Audiosector LM4780 copy, using exactly their specifications at their support forum.
3). Enjoyable loud house party? Decibel Dungeon Inverting LM3886 for great quality jamming out and the cool running of a stable amplifier.
4). Powerful High Fidelity amp? MyRef design.
47uF Ci (NFB cap) electrolytic is correct size to put with 6,800 ohm, but the feedback shunt resistor aboard the kit is 680 ohm. It looks like a decimal error since its 10 times wrong. You can try 470uF (a low impedance model like those Panasonic, but you can use a little size 16v from a recycled computer motherboard) // 10uF or 22uF ordinary power cap // 10nf (0.01uF) polyester dip cap (the "//" says these are parallel to each other, creating a larger value cap). This figure is grossly approximate "in the ballpark" which is a lot better than the 10 times wrong 47uF (bass blocker) value. The sound is a little better when the values aren't far off.
The amplifier will have a much more useful tone if you get exactly the same part # as Audiosector uses for the two power caps at the amplifier board rather than the ineffective 100uF. I believe that the Audiosector LM4780 (internally 2 of LM3886 cores) uses something like a 1500uF Panasonic (actually a low impedance cap like Nichicon PW, FW as well--whatever you do, use a low impedance part, preferably the exact same model as the Audiosector kits use since that is a known good). This well working part is 15 times different than the chipamp.com kit. That will sound different for sure.
The board shown has metal resistors setting on a mess of metal traces where power crosses small signal. That's not the least bit stable. Please see the National Semiconductor LM3886 datasheet and add the little picofareds RF Blocking cap, because it also helps to promote stability. You could choose to re-build small signal UP above the board a bit so that you can also incorporate a Signal Star instead of that bus rail design. As I remember, setting 1/2w carbon film resistors somewhat upright (dogleg), was fairly effective at leaping over the problematic board layout, but it was a minor difference. Audiosector support forum here at diyaudio.com illustrates applying 330pF (picofareds) in parallel to the feedback resistor of LM4780/LM3886. Similar values for RF blocking cap (see real datasheet) will probably help promote stability.
|Thread Tools||Search this Thread|
|Thread||Thread Starter||Forum||Replies||Last Post|
|Chipamp 3886 Mono Beta Build||JL889||Chipamp||4||21st November 2011 02:31 AM|
|Dual-mono chipamp (3886) problem - possible grounding issue||phresh||Chip Amps||31||29th February 2008 04:50 PM|
|PS for Brian's chipamp 3886 and 4780||rgrayton||Chip Amps||0||21st April 2005 12:39 AM|
|New To Site?||Need Help?|