lm3886 hum

[Mnvizb]

Hello, Ive just finished building amp using lm3886. The problem is - when I apply voltage ican hear hum both through speaker and my toroidal transformer. Chip is insulated. Schematic is from National Semiconductor. Any ideas?

Edit: I've just tried the same circuit with another chip (this time uninsulated, but I insulated it properly)

[johnr66]

Sounds like something is miswired and causing heavy current to flow. Recheck you wiring. Post clear pix of your circuit so people can help.

This sounds like a case of first power-up without a dim bulb tester. Your money goes up in smoke.

[minsik]

I am newbie to these forums but however I have 30 years plus in consumer electronic service before this. In TV/VCR/DVD/amps etc.

1) First don't turn anything on. Physically check and double check all connections, eyeball measure all connections and check for s/c and bridging on all components. Assuming its not a kit then recheck the wiring and tracks against the cct. Assume nothing! Make sure all tracks and connections match the cct. Do use an ohmmeter and verify all components/tracks and connections are as expected. That means looking at the cct, deciding a worthwhile measurement then deciding if its in line with what you expected and if not why now. 1 good measurement is worth 100 bad ones (I suggest removing the amp ic and do not insert till all checks/cct verify is done)

2) The previous poster means to use a current limit light bulb in series with the dc or low ac secondary supply to limit the current (prevents smoke escaping while faultfinding) However this assumes step one 1 is fully followed. Sometimes its a low- medium value resistor ie 10-100 ohms in both supply rails to limit the current. I suggest removing the main amp IC and checking carefully all DC rails to all pins and check they are as expected. ie +-40Vdc depends on the Transformer and caps and bridge/diodes (after step 1).

3) As said b4 there is a lack of any information of real value supplied. Like you are a doctor and I visit. "Something is wrong I say, you=doctor, you fix" Now we both know this is bull... and no doctor can do anything but give you a bill and say have a nice day! We need to know what measurements/steps you have taken to find/fix the fault/problem so far. Good details please would be of help.
ie have you measured the dc rails. see note 1+2. What is the cct. specifically. Did you add/change anything. Can you print the cct, measure the dc and then anotate the printed cct, scan and upload with your reply post. or cut the cct to a jpg and add points and voltages you have measured. (without the IC first)
Once again the supplied information is worse than sketchy! Hum in amp and Transformer loud or soft? does ic get warm in 10 seconds? Dont run it too long of poof!

Many repairs are performed long before any power is applied to any cct. Observation by eye/smell/touch is very valuable as well.

Love to help when we get some real information. Also it looks like you have limited/no faultfinding experience from the info supplied. What experience do you have? What test gear do you have? How good can you solder? Have you built ccts before?

Can you imagine a doctor going to a doctor, going armed with supplying the xrays, the samples and a written record history so far, internet references, specialist quotes etc. Then the diagnosing doctor has stuff to get his teeth into is likely to come back with a fair and reasonable diagnosis.

Lots of this can be time consuming. Thats why my most freq customer argument used to be the $60 labour against the 10c resistor problem. I make no excuse for that and an explanation of experience and method versus guesswork and a big bin usually made sense.

Minsik

[AndrewT]

Don't ever connect a speaker to an unproved project!!!!! never ever.
Measure the project. Prove it is working properly. Prove it is safe.

Attach a dummy load and repeat the testing measuring to prove that under load the project is still safe.

Only when you are convinced that the project cannot misbehave do you attach the correct impedance speaker.

[AndrewT]

Did you test and measure the transformer before attaching the PSU?
Did you test and measure the transformer after attaching the PSU?
Did you test and measure the PSU before attaching the amplifier?
Did you test and measure the PSU after attaching one channel of amplifier?
Did the bulb tester light up during any of these stages of test and measurement?

[Mnvizb]

everything with my supply is alright. I'll try to buy new parts and wire it using same schematic.

[djoffe]

This link covers some of the more subtle things...

http://www.updatemydynaco.com/docume...lemsRev1p4.pdf

Some things that might help remote trouble-shooting...
How many mV of hum?
Send pictures of your wiring.
Post a schematic (amp and power supply).

With some care, an LM3886 amp should be dead quiet.

Good luck.

[Mnvizb]

Yup, lm3886 is dead quiet. I made one channel which performed flawlessly, then, I made another one using same pcb, just different chip(insulated). Pcb was computer made (transferring printer toner) so it was flawless. I used the same supply which I used powering the other chip. When i turned it on, it wasn't hum you could hear from some lower quality amplifiers, but it was buzz and my toroid transformer started shaking and it's temperature started rising quickly. Note: if I disconnected ground the noise stopped, but my speaker still reacted to any signal (if u transferred tone or even touched input with your finger)

[!]

I'd wonder if the chip is damaged.

[CanAm Man]

Originally Posted by AndrewT

Don't ever connect a speaker to an unproved project!!!!! never ever.

(Actually, it's OK.....as long as you use BORROWED speakers--from a friend you don't value very much.......)

Honestly.....as always, Andrew T is "on the mark...... Large wattage resistors make great dummy loads. I was able to pick up a pair of 6 ohm, 100-watt ones for $5.00 at my local electronics surplus store..... (if at all possible, get the non-inductive resistors....).