LM3886 internal Schematic build or reverse engineering circuit

  • Thread starter Thread starter Deleted member 550749
  • Start date Start date

Suggest better output transistor, using 2SC5200

  • If anyone interested plz make proper pcb share plz

    Votes: 2 66.7%
  • Any improvements?

    Votes: 2 66.7%

  • Total voters
    3
Sorry for not updating the final schematic, this the sch which m using and made-
AddText_09-01-05.40.05~3.jpg
 
  • Like
Reactions: Zohaib Ahmad
The LM3886 equivalent circuit has a phase inversion problem on the negative clipping because the VAS current is not limited, but I would be very surprised if that problem actually exists on real chips. It can be fixed with a Baker clamp or a collector resistor on the Darlington transistor, or a transistor or diode current limit using the 800 Ohm Re. If anyone has an LM3886, I'd like to see the clipping waveform? Also, has TI fixed their spice model? It has/had serious problems. Others here in DIYA have made models based on the "equivalent circuit", ie in the LTC Spice thread.
 
Fun project but I can't help but wonder why it exists. The LM3886 is a pretty standard Linn architecture with a few tweaks and NPN-only output stage (because PNPs in the process are awful). You could design a better amp if you used a complimentary Class AB output stage. I'd also be a bit careful with the diode biasing of the output stage in a discrete circuit. You do want reasonably good thermal tracking to prevent thermal runaway. That said, it looks like a fun learning exercise.

The LM3886 has a few protection circuits built in that aren't shown in the equivalent circuit schematic, including:
  • Over-temperature protection
  • Over-current protection
  • SOA protection
and probably others that don't immediately come to mind for me right now.

You also seem to have dropped the mute circuit. 😉

I have never seen phase reversal on an LM3886. It does suffer from a bit of rail sticking and spurious oscillation when it exits rail sticking on the positive rail. That can be mitigated by the Rf2, Cf, and Cc shown in AC Test Circuit #2 in the LM3886 data sheet. If you treat it right, the LM3886 works very well.

Tom
 
Yes, i just made for fun, removed mute cirquit to make it more simple , when I first made with bootstrap version, there was very bad clipping distortion happened that was very bad, after coming here someguys r helped to improve the circuit, made this just because in India u can't find a original ic and if u find that's double or triple price, and diy is best it can take huge place on PCB but repairable, and I can tweak, modify for high voltage for more power 👍 fun project, working satisfied 😉
 
I did remember I had analyzed the LM3886 once upon a time. Beginners find this circuit noting but ordinary quasi complimentary without noticing the extra NPN driver in negative pole.
The DS if someone reads carefully says, if amplifiers are concidered to be class B, this amp is class AB. Here is why look bellow the output transistors currents, the positive rail one is non switching.

lm3886 output currents.JPG

By then I had found that adding 680K resistor in parallel to 10pF makes it better.I don't remember why.
 

Attachments

  • LM3886 CLONE.JPG
    LM3886 CLONE.JPG
    144 KB · Views: 233
Last edited by a moderator:
Last used 2or 3 month ago, only tried to fix negetive clipping, fixed but that diod remove is wrong dicision reverted working, but I got better circuit of this same IC that's without diod version, currently using honeybadger sounds good or better than this 🤣🤣 LM
 

Attachments

  • Schematic.png
    Schematic.png
    17.4 KB · Views: 111