Jax, how much offset did you get with your LM3886 design?
http://jax.myphotos.cc/gallery/album02
Jax is http://www.diyaudio.com/forums/member.php?s=&action=getinfo&userid=9528
http://jax.myphotos.cc/gallery/album02
Jax is http://www.diyaudio.com/forums/member.php?s=&action=getinfo&userid=9528
With 1k from NI to gound it's a question of luck.
In my experience these chips have a huge production drift.
1k may be good for one chip, and in another one you may need 20k.
Some chips don't even need the resistor.
BTW, 1500uf per rail * 5 chips gives 7500uf, and that's too much capacitance.
Bass will suffer.
That's my experience with the National Overture series power op-amps, be it LM 1875, LM3875, LM3886, LM4780 (I've tested all these).
I would use 2x1000uf on a central position and 2x220uf on each chip.
Very nice amp.
In my experience these chips have a huge production drift.
1k may be good for one chip, and in another one you may need 20k.
Some chips don't even need the resistor.
BTW, 1500uf per rail * 5 chips gives 7500uf, and that's too much capacitance.
Bass will suffer.
That's my experience with the National Overture series power op-amps, be it LM 1875, LM3875, LM3886, LM4780 (I've tested all these).
I would use 2x1000uf on a central position and 2x220uf on each chip.

Very nice amp.

BTW, 1500uf per rail * 5 chips gives 7500uf, and that's too much capacitance.
Carlos, why do you think this is?
I have read many posts preferring just 1000uF per rail.
I started out with 1000uF / rail at the chip but then added 10,00uF per rail near the rectifiers for better smoothing and liked the sound. It seemed to carry more weight and authority.........
....4 x lm3886 in parallel, each with its own set of regs - 10.000 uF cerafine and 0,1uF fkp3 film at the reg-input...and 470uF at the reg-output + 0,1uF fkp3 at the power in on the lm3886 chip.....the hole thing is bridged with a similar arrangement....2 x 500va incapsulated talemas....and its noninverted with servos........and, no i don't have a digital camera
... you don't have a friend which has...?tbla said:......and, no i don't have a digital camera
don't think so.....but i've heard there is one at work - never used a digital camera before, kind of embarresing

The complete guide to the GC PSU capacitance
Pay attention to the whole thing: bass, mids, treble.
With my tests, more than around 3000uf per rail TOTAL (one PSU for two channels) the bass becomes more prominent but slower, midband gets "thicker" and with less datail, and the treble suffers a little too.
The bass has more "body", no doubt, but it doesn't have detail.
I have recordings with great bass, but even then I have to distinguish between an earthquake and real bass, that is, I have to listen to every string that the guy on the bass is playing.
It's the difference between quality and quantity.
I like a fast and detailed sound.
The ideal capacitance, for me, is around 2,000uf TOTAL.
You have that if you have a GC with 1,000uf per rail, two chips and one PSU.😱
NealG said:
Carlos, why do you think this is?
I have read many posts preferring just 1000uF per rail.
I started out with 1000uF / rail at the chip but then added 10,00uF per rail near the rectifiers for better smoothing and liked the sound. It seemed to carry more weight and authority.........
Pay attention to the whole thing: bass, mids, treble.
With my tests, more than around 3000uf per rail TOTAL (one PSU for two channels) the bass becomes more prominent but slower, midband gets "thicker" and with less datail, and the treble suffers a little too.
The bass has more "body", no doubt, but it doesn't have detail.
I have recordings with great bass, but even then I have to distinguish between an earthquake and real bass, that is, I have to listen to every string that the guy on the bass is playing.
It's the difference between quality and quantity.
I like a fast and detailed sound.

The ideal capacitance, for me, is around 2,000uf TOTAL.
You have that if you have a GC with 1,000uf per rail, two chips and one PSU.😱
My 4 year old daughter can take perfect pictures... not so hard if you put the dial at "Auto"!tbla said:don't think so.....but i've heard there is one at work - never used a digital camera before, kind of embarresing![]()
peranders said:
My 4 year old daughter can take perfect pictures... not so hard if you put the dial at "Auto"!
And then?
Does she take the picture out of the camera?😀
I still think that the easyest thing to operate is a Polaroid.🙂
I don't want to teach her that. She has learned herself how she can put in a video tape and use the remote control, put a CD on, switch all my hifi gear since I have the TV sound through my electrostatic speakers. She has also learned herself how she switch on my Mac and use MS Word and also print the document. My goal all the time has been to keep her off my toys.... which is impossible.carlosfm said:
And then?
Does she take the picture out of the camera?😀
Offset was some +20mV if I remember it correctly. I have not experimented with the resistor at the + input but increasing it should reduce the offset. Ideal should be 330k but that would add some noise to the input as well. 1k is a compromise, I don't like to connect inputs to ground directly.
About the capacitors...
Each amp is fed from a common point with a short piece of 22AWG wire. That would reduce the contribution from neighbour capacitors some.
The amp works just fine but there is no real "bottom" in the bass. It does not matter since it is intended to be the 5 in a 5.1 system. The sub has its own amp.
I have since removed the pots, I got a new sony str-sl5 receiver cheap and I will later rip the power amps out from it and modify it to a preamp. It has individual level settings for all channels via the remote, no need for the pots anymore.
The amp is rock stable and neither hum or hiss can be heard (90 dB speakers).
I will later build a pair of monoblocks with 3875 or 3886 just to experiment with. But not before my triode connected 813 SE is ready.
😉
About the capacitors...
Each amp is fed from a common point with a short piece of 22AWG wire. That would reduce the contribution from neighbour capacitors some.
The amp works just fine but there is no real "bottom" in the bass. It does not matter since it is intended to be the 5 in a 5.1 system. The sub has its own amp.
I have since removed the pots, I got a new sony str-sl5 receiver cheap and I will later rip the power amps out from it and modify it to a preamp. It has individual level settings for all channels via the remote, no need for the pots anymore.
The amp is rock stable and neither hum or hiss can be heard (90 dB speakers).
I will later build a pair of monoblocks with 3875 or 3886 just to experiment with. But not before my triode connected 813 SE is ready.
😉
- Status
- Not open for further replies.
- Home
- Amplifiers
- Chip Amps
- LM3886, Jax how about offset