:-) :-) :-) *** Hello diyAudio members, I need some help in many questions about LM3886 *** :-) :-) :-)

Polypropylene is the best, but it isn't really necessary. Ceramic is also good for filters and such. I usually just use electrolytic.
Agree that polypropylene is best for filter capacitors. A good bipolar electrolytic cap can work well too.

You can use ceramic capacitors in filters but you really need to pay attention to the capacitor dielectric if you do. The C0G/NP0 ceramic dielectric is excellent and the performance can meet (and sometimes exceed) that of a polypropylene cap in the same spot. The other ceramic dielectrics are useless for filters but work well for bypass capacitors and other non-signal applications.

BTW: I've written quite a bit about the LM3886. You can find the information here: https://neurochrome.com/pages/taming-the-lm3886-chip-amplifier

Tom
 
Agree that polypropylene is best for filter capacitors. A good bipolar electrolytic cap can work well too.

You can use ceramic capacitors in filters but you really need to pay attention to the capacitor dielectric if you do. The C0G/NP0 ceramic dielectric is excellent and the performance can meet (and sometimes exceed) that of a polypropylene cap in the same spot. The other ceramic dielectrics are useless for filters but work well for bypass capacitors and other non-signal applications.

BTW: I've written quite a bit about the LM3886. You can find the information here: https://neurochrome.com/pages/taming-the-lm3886-chip-amplifier

Tom

... so using the break for some sensible Topics.
🙂 Biology is so boring 🙂

Dear tomchr,
you wrote a little bit 🙂 ???
Only one Word ... WOW !!! It looks like you are the Amp specialist ... WOW !
Your work could be now a good exemplary what can been reached after hard working in topic. Fine to get to know you.

🙂

So you think i can use COG ( in the Range 33pf up to 470nF ) also in audio path,
I see you use ceramic also in your Moduls 186er Version. No foils are seeing 🙂
So , after your Specs it must be working fine 🙂

So it is much cheaper for me and better for the PCB....because of ( ESL )...
--> Sorry about before calling it LC , of course I meand ESL of the C Componets.

What you are thinking, are one is better as the other : TDK or Kemet ?

For bigger caps like 100µF i found the bipolar nichicon MUSE types ...

Can I please ask you something tomchr ?

Manual Soldering in 1206 or 0603 also in SMD is no problem for build few homemade boards.
Why are the most "maker" PCBs are in wired resistors ?

SMD in Metall and 1% are so much better for a PCB...I think so. Also it gives 0,1% with good temp classes and so on.
I was wodering why all the most selfmade boards are in wired Versions ?

Is there a technical ( audiophile ) reason ? I think not.

have a great day,
 
So you think i can use COG ( in the Range 33pf up to 470nF ) also in audio path,
As long as you stay with C0G/NP0, yes. Also choose a larger package (like 1206) if you can.

What you are thinking, are one is better as the other : TDK or Kemet ?
The one that's in stock. Especially these days... 😉

Manual Soldering in 1206 or 0603 also in SMD is no problem for build few homemade boards.
Why are the most "maker" PCBs are in wired resistors ?
Many shy away from SMD components because they're smaller and can be annoying to work with.

I've written quite a bit on why I don't support DIY SMD builds here: https://www.diyaudio.com/community/threads/modulus-86-build-thread.267802/post-5738981.
Some will argue that a mix of SMD and PTH (leaded) parts is the cat's meow. That hasn't been my experience. You can see my view here: https://www.diyaudio.com/community/threads/modulus-86-build-thread.267802/post-6839949

Tom
 
So , on all many thanks,
actually i start my layout.

What you are thinking is better in my Case ?

Left on my PCB is all with Power & PowerGND and the right are all with Amp & AudioGND

I separate the 2 GNDs and lead at one point on Top with few 0 Ohm.

I make a 4 Layer PCB:

1 Top PowerGND & AudioGND & most Signals inner AudioGND shielded.

2 Flood GND --->> ??? but which one is better --->AudioGND or PowerGND

3 Feedback of Composite Signals and +-15V OPAs, also the rest in AudioGND or PowerGND.

4 Bottom PowerGND and sperate AudioGND ,also lead at one point on bottom with few 0 Ohms

So what is your Opinion about floated Layer 2:
AudioGND ??
or
PowerGND
 
Good morning members 🙂

I want to ask were I must place the topic if
I want to discuss a active high pass active filter. I do not found a fitting Forum place.

I want to add a high pass Filter seperate in front of my design to make it adjustable for the LM 3886 Amp.
So preamp and highpass in one.

I found a active high pass with multiple negative feedback. It should be better than a Sallen-Key high-pass, because distortions from a common-mode operation are avoided ?!

See here the original link : https://www.aktivfilter.de/hochpass-mit-mehrfachgegenkopplung.php

But this "preamp module" uses the negative input of the OP.

My Question:
Does it matter if the preamp comes in on negative input and goes into the postive input of composite OP of the LM3886 ?
Is then the Phase 180° changed , or is is the the same Phase ?



......and that it doesn't get boring, my father asked if his old dual record player from the 70s also works with it what I do here....
😱....That will be the next topic...😉
 
So , on all many thanks,
actually i start my layout.

What you are thinking is better in my Case ?

Left on my PCB is all with Power & PowerGND and the right are all with Amp & AudioGND

I separate the 2 GNDs and lead at one point on Top with few 0 Ohm.

I make a 4 Layer PCB:

1 Top PowerGND & AudioGND & most Signals inner AudioGND shielded.

2 Flood GND --->> ??? but which one is better --->AudioGND or PowerGND

3 Feedback of Composite Signals and +-15V OPAs, also the rest in AudioGND or PowerGND.

4 Bottom PowerGND and sperate AudioGND ,also lead at one point on bottom with few 0 Ohms

So what is your Opinion about floated Layer 2:
AudioGND ??
or
PowerGND
so what is the better ground reference for all ? PowerGND or AudioGND.
50% , 50% decision ?! Please tomchr, can you give me a tip please ?
I'm afraid that if I put the wrong inner surface, my whole PCB could be waste.

 
Good morning members 🙂

I want to ask were I must place the topic if
I want to discuss a active high pass active filter. I do not found a fitting Forum place.

I want to add a high pass Filter seperate in front of my design to make it adjustable for the LM 3886 Amp.
So preamp and highpass in one.

I found a active high pass with multiple negative feedback. It should be better than a Sallen-Key high-pass, because distortions from a common-mode operation are avoided ?!

See here the original link : https://www.aktivfilter.de/hochpass-mit-mehrfachgegenkopplung.php

But this "preamp module" uses the negative input of the OP.

My Question:
Does it matter if the preamp comes in on negative input and goes into the postive input of composite OP of the LM3886 ?
Is then the Phase 180° changed , or is is the the same Phase ?



......and that it doesn't get boring, my father asked if his old dual record player from the 70s also works with it what I do here....
😱....That will be the next topic...😉
Hello Micha,
look here:
https://www.diy-hifi-forum.eu/forum/showthread.php?103-2-weg-Aktivweiche-Subsonic-Bassanhebung
 
👍👍:wave2::wave2:🙂🙂
bin auch schon kräftig am LTspice simulieren
so what is the better ground reference for all ? PowerGND or AudioGND.
50% , 50% decision ?! Please tomchr, can you give me a tip please ?
I'm afraid that if I put the wrong inner surface, my whole PCB could be waste.

what a pitty that nobody can help me with the ground :-(

Anyway a near finished my PCB and it looks great, a real Try and Learn PCB in 4 Layer ....
Each ground Perfekt seperated on each layer as ... Power, Audio and also OPA Ground...
So star connection is possible or bridge by 0 Ohm Rs, what I want.

Now I thing to make an active differential RMI Filter in Front of, because of Spice Noises over 22 khz.
So simple to do that, i found so much stuff on TI Homepage and application Notes...
Testing in Spice makes good results...

Then I decide how to make the Innerlayer by myself and then ready to go '🙂
 
I'm a bit sad that no one can help me anymore😢. Did I say something wrong ? 🤔sorry , if so.

Now I don't dare to ask anything anymore because nobody answers anymore. I'm only missing 2 things and then I'll have my first design ready, with everything I've learned so quickly.

Such a pity, but many thanks to those who supported me so well at the beginning and gave me a good start.
 
I think you will find all relevant grounding information on tom's neurochrome site:
https://neurochrome.com/pages/grounding
Hi, many thanks for the link 🙂 exactly what i do , seperate all ground on 4 layers.
1) AUdio GND
2) PowerGND
3) OPA GND
4) Audio IN GND
... also one Case

All each by his own, but a possibility to connect them on PCB by 0 Rs.
Target is the Star connection to Main Power PCB or on PCB direct in a own inner layer.

But I do not found which of the 4 is better to make the common refference allover in the first inner layer.
Please look at #26
 
Last edited:
I think that should be avoided, see tom's article.
Star center connection should rather be the output gnd connection (very simplified).
So you mean the location minus of speaker out is the best star connection point for all, yes ?
could I have overread this in the text of Toms article, I do not found it 🙂
---> do you mean this .....GND_PWR: The "ground star" at the local decoupling on the LM3886."
So on this point i make the Star in one layer for all ?

So i take the POWERGND at the Location which is this Gnd the minus of speaker.
On this location i go in the inner layer a take the hole inner area to connect to all other seperatet Gnds
 
Last edited:
...That what irritate my about the star :

......40+ dB improvement over the star grounding scheme at 10 kHz and 95 dB improvement near DC. The only difference between the star ground and the improved grounding scheme is that the GND_SIG trace was routed to the output connector rather than to the star ground....

I must go study Toms circuits more in details on his explanation. ....and don't be so impatient. ;-)
@ stv, many thanks you put my in the right way now.... kisses and many thansk to you