Here was the thread showing that Cc causes the stability margin of the amp to decrease:
LM3886: the effect of the compensation network Cc, Rf2 and Cf
LM3886: the effect of the compensation network Cc, Rf2 and Cf
Yes, but they are 50mm apart - adding the distance to the pins is another 2-3cm on each side, so the total is around 10-11cm.
Only one MF decoupling cap is operating at any moment.
When +ve current is flowing to the load, then the +ve decoupling is trying to supply that current.
When negative current is flowing to the load, the -ve decoupling is trying to supply that current.
I suspect you are near my guesstimate of
When +ve current is flowing to the load, then the +ve decoupling is trying to supply that current.
When negative current is flowing to the load, the -ve decoupling is trying to supply that current.
I suspect you are near my guesstimate of
Thanks for the link minifly, I will have to dig up that schematic or post according to which I put that cap there.
So the only difference is that I wont be connecting the MF and HF grounds directly with a trace, but on the board. I'm going to connect the HF ground to pin7 (ground).
Would you please comment on post#40?
PS no luck with that copper shielded satellite cable, all the storekeepers told me it's "old school", so it might be a pain to find again, which is a pitty.
post40 makes sense.
You have arranged for the sqrt(2) factor to be equaled or exceeded in each stage.
It would be worth experimenting with F-3dB =2Hz & =4Hz & =8Hz & =16Hz to "hear" what difference the amp bandwidth makes to the way your speakers perform.
Help,
I typed equalled and my spell checker told me to change it to equaled. This does not look right !
You have arranged for the sqrt(2) factor to be equaled or exceeded in each stage.
It would be worth experimenting with F-3dB =2Hz & =4Hz & =8Hz & =16Hz to "hear" what difference the amp bandwidth makes to the way your speakers perform.
Help,
I typed equalled and my spell checker told me to change it to equaled. This does not look right !
Interesting thread. Someone who can hear a difference between amps getting advice from others who can't. Curious how this will end - will science finally prevail?
I don't tolerate this kind of posts, and I think a lot of members would agree.
Do you have something to add to the discussion? If not, then leave.
Yes, that cap will go from IN+ to GND. I'll just have to see if it fits across to pin7 (gnd) because I will solder the HF decoupling ground to pin7, so it could be crowded.
Think about what you are proposing.
You are adding an RF attenuation to the Hot/Signal input.
Where is that coming from?
I reckon it is coming from the input cable. That cable connection is a two wire connection.
If you need to attenuate RF on that two wire connection then the Capacitor MUST go across Hot?signal to Cold/Return. Not to power ground.
If you are concerned about common mode interference, then both wires in the two wire connection must be cap connected to Chassis. Not to Power Ground.
You are adding an RF attenuation to the Hot/Signal input.
Where is that coming from?
I reckon it is coming from the input cable. That cable connection is a two wire connection.
If you need to attenuate RF on that two wire connection then the Capacitor MUST go across Hot?signal to Cold/Return. Not to power ground.
If you are concerned about common mode interference, then both wires in the two wire connection must be cap connected to Chassis. Not to Power Ground.
Check. And again, I shouldn't post when I'm tired
I guess the optimal place for that RF cap is across IN and SG (return) on the PCB underside.
Will do that as soon as I solder in a new chip, as one of them died.
As far as my initial problem goes - the healthy right channel seems not to be "bloated" any more, but more on that soon...
I guess the optimal place for that RF cap is across IN and SG (return) on the PCB underside.
Will do that as soon as I solder in a new chip, as one of them died.
As far as my initial problem goes - the healthy right channel seems not to be "bloated" any more, but more on that soon...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.