LM3886 - DC offset problem

luis.jimenez · 2010-03-28 4:18 am

Ok, I have the amp completed and no smoke this time (I fried the rectifier earlier; had a short between one of the connections and the heatsink. Point to Point on crowded spaces is not my forte). I test for DC offset with no load and I have 3 mv on one channel (at 0 vol) but have 320 mv on the other channel. With volume up, I have 32 mv on the 'good' channel but the other channel (the one with the short earlier) goes to
(minus) -220 mv. What could cause that?

Thanks,

Luis

Sangram · 2010-03-28 5:25 am

Check connections around the input resistor. Do you have an input capacitor? Feedback capacitor? What values of resistors have you used? Can you post a schematic/pictures?

wwenze · 2010-03-28 6:02 am

Point to Point on crowded spaces is not my forte).

Buy teflon tubing - the insulator tubing that one uses for DIY wires, and use that to cover up all exposed metallic leads leaving enough just for soldered connections.

tomchr · 2010-03-28 9:49 am

3 mV sounds normal. 320 mV sounds high. A few 10's of mV probably won't hurt any. But if it gets beyond that, I'd deal with it.

~Tom

luis.jimenez · 2010-03-28 2:11 pm

sangram said:
Check connections around the input resistor. Do you have an input capacitor? Feedback capacitor? What values of resistors have you used? Can you post a schematic/pictures?

Sorry, no input resistor - will connect a 22k one today. no input capacitor nor any feedback capacitor.

luis.jimenez · 2010-03-28 2:12 pm

tomchr said:
3 mV sounds normal. 320 mV sounds high. A few 10's of mV probably won't hurt any. But if it gets beyond that, I'd deal with it.

~Tom

Sorry, I meant 32mv instead of 320mv. It does go negative with volume full up. Thanks.

luis.jimenez · 2010-03-28 2:15 pm

wwenze said:
Buy teflon tubing - the insulator tubing that one uses for DIY wires, and use that to cover up all exposed metallic leads leaving enough just for soldered connections.

Wwenze, thanks for the tip; I should've used it on the capacitors, but there isn't a lot of exposed wire. Will try to post a picture if you guys promise not to laugh...

luis.jimenez · 2010-03-28 2:54 pm

Bad hum, bad hum..

This morning I was getting ~50mv offset on the amp so decided to connect some Polk Audio bookshelf speakers. With the source (CD plyer) off, I could hear a hum, and also some motorboating from the amp. I realized that my RCAs were not isolated so I moved them off the case and connected the source again. This time the hum was really LOUD at 0 volume. Any help would be appreciated.

Cheers, Luis

bobodioulasso · 2010-03-28 4:07 pm

The inputs have to be shorted to measure offset.
No capacitor in the feedback leg leads to an offset.
No input capacitor can lead to an offset if DC is present at the output of the source.
How many uf ( filtering)in your PSU?
Your schematic, including PSU would be welcome.

luis.jimenez · 2010-03-28 4:59 pm

bobodioulasso said:
The inputs have to be shorted to measure offset.
No capacitor in the feedback leg leads to an offset.
No input capacitor can lead to an offset if DC is present at the output of the source.
How many uf ( filtering)in your PSU?
Your schematic, including PSU would be welcome.

bobodioulasso, I thought volume at 0 shunuted the input to ground.
PS is Toroid, 22v+22v, single rectifier, snubbed.
Here's the ckt schematic. One star ground on the chassis.

Capacitors are 10,000 uf, snubbed with .1uF.

Question: where do I ground the (-) input? Right now it's not grounded - the input ground(s) go to the volume pot and nowhere else. I tried to ground them to the star ground but it made no difference.

luis.jimenez · 2010-03-28 6:05 pm

PS schematic

luis.jimenez · 2010-03-28 8:46 pm

More annoying than the hum, which is bad enough, is the motorboating, really, more like a lawn mower... I have the xfrmr CT on the same ground point as the rest of the ground signals. Is that a problem?

You can hear the motorboating coming from the rectifier. I wonder if it's the snubber caps, maybe got damaged when the other rectifier burned? I'll try to connect without the caps.

Also, no music out when I connect the CD player to the amp. I checked the attenuator and it and all connections worked fine.

AndrewT · 2010-03-28 9:04 pm

that schematic looks like it's from the National datasheet.
That's the stripped down version that is likely to misbehave when beginners and intermediates try to implement it.

You need the AC coupled version from the datasheet with all the optional parts fitted.

luis.jimenez · 2010-03-28 10:06 pm

motoroating

Thanks Andrew, I may indeed change layouts. I notice that when I relocate the rectifier, the motorboating either lessens or totally subsides. Is this a clue? Could the rectifier be too close to the star ground? the output signal wires?

Thanks again,

Luis

AndrewT · 2010-03-29 9:29 am

the motorboating may be an LF instability, but I have a feeling it is a protection circuit re-triggering due to detection of a fault.

luis.jimenez · 2010-03-29 1:47 pm

AndrewT said:
the motorboating may be an LF instability, but I have a feeling it is a protection circuit re-triggering due to detection of a fault.

Andrew,

That makes more sense than anything else, as I have checked connections and moved things and nothing gets rid of it - also may explain why no output. So, if it's a fault, what could cause it on both channels? V levels are within specs.

Thanks for taking time to answer,

Luis

tomchr · 2010-03-30 12:57 am

I wonder if a protection circuit gets cranky because the mute pin gets pulled low too fast. The data sheet calls for a 100 uF cap there. Also, what's the resistance of the mute resistor (Rm)? The current through Rm should be at least 0.5 mA. So calculate Rm from Rm = Vsupply/0.0005. Round down to the nearest standard value.

Your "amplifier board" circuit should work if Cm is fitted. You may have some offset on the output. The max input referred offset of the LM3886 is 10 mV. With a gain of 22k/680 = 32.4, the max voltage offset you should see on the output of your amp is 324 mV.

Note that all the grounds in your schematic should come together at one point - the star ground. The hum you are describing could be caused by a floating input ground.

~Tom

tomchr · 2010-03-30 1:02 am

luis.jimenez said:
Thanks Andrew, I may indeed change layouts. I notice that when I relocate the rectifier, the motorboating either lessens or totally subsides. Is this a clue? Could the rectifier be too close to the star ground? the output signal wires?

What do you mean by "relocate the rectifier"? If you physically move it around without disconnecting any of the wires, then I bet it's a cold solder joint somewhere around the connections to the rectifier.

The rectifier does not connect to the star ground so its location with respect to the star ground should not matter.

~Tom

luis.jimenez · 2010-03-30 1:22 am

Tom, thanks for all the pointers and advise. I will fit CM and will verify Rm value. I will also check all solder connections.

Cheers,

Luis

luis.jimenez · 2010-04-10 1:04 am

Hello again,

I corrected the ckt and got one channel to work, pretty quiet too. But, the other channel is still exhibiting some of the same symptoms, and it gets extremely hot right off the bat - the other channel doesn't. I checked everything and can't figure out what's causing it to get so hot. Could the chip be shorted inside? It amplifies the music, but very noisy and with a cadence, not like the motorboating earlier but like the protection ckt is kicking in and out, in and out, every half a second.

Any help is appreciated.

Thanks,

Luis

LM3886 - DC offset problem

Member

Moderator

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

R.I.P.

Member

R.I.P.

Member

Member

Member

Member

Member