Lack of bass with LM3886

[maillard]

Hello everybody,

I have built a complete "chipamp" (amp + Psu) kit, double mono, powered by 2X300VA transformers.

Sounds good, but compared to my small "NAD 3020" and to a big "Kenwood" L1000M, using the same loudspeakers, I feel a lake of bass, which is very frustrating.

I was about to upgrade the caps of the PSU to Munsdorf 15000 microF/18mohm instead of the Panasonic 10000 microF/36mhom.

Should I go this way or may be the limitation is part of the LM3886 itself ? In that case would be better to change for another type of amp (4981x or discrete)

Any experiences will be appreciated.

Thank You for the help.

 

[tinitus]

hi

lake ?
do you mean lack of bass

would you want me to correct it ?


moderation

 

[ficklecycler]

How deep is the lake?

 

[infinia]

lake of bass fishing?

 

[N Brock]

do you mean lack of bass

would you want me to correct it ?

reminds me of this guy

 

[ficklecycler]

I actually swam on Lake Bass as a child. Areas of it are muddy.

 

[Conrad Hoffman]

I don't think changing the caps will make much difference. The problem probably lies elsewhere. Can you measure the frequency response? Sometimes a bit of a tilt upwards will sound like a lack of bass. There's also the gain issue. Whatever's louder sounds better. When you listen, is the level the same? Is there any subtle oscillation going on? If the chip amp runs warm that's a possibility, and the tiniest amount of oscillation, IMHO, affects the sound. Where do the speaker returns physically connect in the amp in relation to the circuit grounds? It matters!

 

[infinia]

sometimes old receivers have a bit of bass EQ even when tone controls are set flat. If they have a loop connection you can plug the chip amp up , bypassing the built-in amp and check difference there.

 

[abraxalito]

It would help if you said what kit you've bought. I agree with Conrad, the lack of bass could easily be an overly bright sound. Brightness that I've experienced with chipamps I've fixed with relatively minor circuit and layout changes.

 

[gootee]

To get any real help you would almost have to post or link to the schematic and photos of your built unit.

 

[digits]

What is the size of the input cap?

 

[wintermute]

What digits said + what is the size of the NFB cap (if fitted). and also as gootee said please post a schematic (which should answer both those questions).

Tony.

 

[Vix]

Hi,

I don't know if I can really help, but I had the same problem with my GC, compared to NAD 3020 amp. LM3886 sounded anemic compared to NAD. The schematic was a non-inverted one, almost identical to the one as presented in the datasheet. NAD always sounded better with a more fuller sound. Note that I never used the loudness and the tone controls were set to neutral.

Long story short, LM 3886 sounds good until you compare it to something better. I played with it for a while, and now it is collecting dust. I would be also interested if someone comes up with a solution for this problem, but I am afraid that it comes down to topology and the limitation of the chip itself.

 

[digits]

the way I see it, the best way to apply the national chips is in a 3 way active system. LM1875 for tweeter duty, LM3875 (or again 1875 for efficient speakers) for mids and the lm3886 for subwoofer duty. I do agree just about any decent discrete topology will beat the standard datasheet or gainclone topology with chipamps.

I have a softspot for Carlos' old discretes the original DX amp, not so refined but it was steel balls, or the HRII. I have not realy been keeping up with developments these last 4 years or so, call me Rip van Winkel.

 

[analog_sa]

Long story short, LM 3886 sounds good until you compare it to something better.

Goodness. And the "something better" is the godawful, as Mr Kessler used to call it, 3020??? And it feels it was only yesterday that chipamps were simply the best amps in existence and people in their droves were busy junking their 300B SETs for a 3875 ultra-platinum kit. Were they all deaf? Or just hated bass?

 

[Vix]

Goodness. And the "something better" is the godawful, as Mr Kessler used to call it, 3020???

I don't want to start any flames here, but, imho, NAD just sounds better.

 

[UrSv]

I don't want to start any flames here, but, imho, NAD just sounds better.

And obviously, under extreme circumstances and with a faulty LM3886 based amp, there is no reason why it couldn't.

 

[maillard]

Thank You for your interest,

(Sorry about the English)

I am talking about a double mono Brian GT kit, schematic and bom can be found at www.chipamp.com/docs/lm3886-manual.pdf.
It's a non inverted type, I have used the components of the kit (no change).
The transfos are 2X24v AC which gives +/- 35 V DC. After one hour, the heatsinks became a little hot (I will says 40 °C max).
There is no input cap (direct coupling).
I have measured about 45 mv offset at the output, that's all I can do, I do not have a scope.
I use the variable RCA output of a Sony CDP 559 ES (no pre-amp).

(a picture will follow tomorrow).

Bye.

 

[AndrewT]

Hi,
is the optional Ci=47uF fitted?

I would recommend you convert this amp to AC coupled, until you learn a lot more about how amplifiers operate.

 

[wintermute]

My chipamp is based on the BrianGT schematic and has the optional 47uF feedback capacitor fitted. Attached is the RMAA test of the frequency response. I'd have to look up the datasheet to see what the calc is for determining the -3db frequency (due to the feedback cap) is, but as can be seen the FR is down around 0.6db at 20Hz. Whether this is enough to be noticeable or not I don't know.

When I did power tests I did get quite significant droop in the PS at frequencies below 30Hz. I'm running a single 300VA dual 20V toroid, giving around 28V rails unloaded.

edit: ok for 680 ohm Ri and 47uF ci forumla is fc = 1/(2πRi Ci) which works out at about 5Hz. Increasing the size of Ci will lower the -3db point.

Tony.