Electronics > Beginners
Need help with noise on headphone amp! (op-amp issue?)
<< < (3/6) > >>
magic:
If it's an NJM4558 from China then it's not an NJM4558, although probably a very similar Chinese chip ;)
But I think I see an NJM4556 on your fist picture and it looks like it might be genuine. NJM4556 is beefy enough to drive most headphones, that's why Grado used it. NE5532 might work too if output current isn't too high, but it is more prone to oscillation (which would result in overheating, among other problems).
Using too weak of an opamp will not make it overheat. Quite the contrary, a too powerful one may overheat if it is driving too much power into the load. Heat is (12V-output voltage)·(output current).

Now grounding: this circuit is clearly made for a bipolar PSU and the midpoint of the PSU must be grounded. This means a PSU with three connections: ground, + and -.
If you simply apply voltage between ground and + or ground and -, it will not work.
If you simply apply voltage between + and -, the two electrolytic capacitors form a divider which splits the voltage in half and provides a ground for the circuit. But it only works for a while, because current consumption from each rail isn't 100% equal. So what happens is that the midpoint between the capacitors (ground) charges up or down until it reaches one of the rails.

Do an experiment: connect power and measure voltage from ground to + or from ground to -. You will see that one raises and the other falls.

Solutions:
- use a proper bipolar PSU
- use a "rail splitter" like TLE2426 to generate ground
- add a resistive divider (like 1kΩ/1kΩ) between the rails and connect its midpoint to ground
Mp3:

--- Quote from: magic on September 12, 2019, 08:11:24 am ---Do an experiment: connect power and measure voltage from ground to + or from ground to -. You will see that one raises and the other falls.

Solutions:
- use a proper bipolar PSU
- use a "rail splitter" like TLE2426 to generate ground
- add a resistive divider (like 1kΩ/1kΩ) between the rails and connect its midpoint to ground

--- End quote ---

Aha..... you rock!! I think you have solved the problem.

Unfortunately the instructions that came with it were only in Chinese so that is probably why I had to come here for help  ^-^

I'm going to look for a bipolar PSU online, i think 8)

If i have a strong enough resistor around, i will try doing what you say!

Thank you so much!
magic:
12V applied to 1kΩ is 144mW. And that's the absolutely pessimistic limit, because normally there will be about ±6V and hence 36mW per resistor. Just about any THT resistor will do.

CMoy amplifiers use 4k7/4k7 and that seems to be enough for them, but lower is better.

edit
Bipolar PSUs have their own problems. When one rail gets damaged or disconnected, the opamp may end up applying voltage very close or equal to the opposite rail to the load. You can try it using your unipolar PSU now. The exact behavior will also depend on the type of opamp used.
Mp3:

--- Quote from: magic on September 12, 2019, 08:24:55 am ---12V applied to 1kΩ is 144mW. And that's the absolutely pessimistic limit, because normally there will be about ±6V and hence 36mW per resistor. Just about any THT resistor will do.

CMoy amplifiers use 4k7/4k7 and that seems to be enough for them, but lower is better.

edit
Bipolar PSUs have their own problems. When one rail gets damaged or disconnected, the opamp may end up applying voltage very close or equal to the opposite rail to the load. You can try it using your unipolar PSU now. The exact behavior will also depend on the type of opamp used.

--- End quote ---

Thank you again! I found this and it seems to be what i want.
https://www.amazon.com/DROK-Converter-Non-isolated-Regulator-Transformer/dp/B0752TRXDC/ref=sr_1_1?keywords=negative+voltage+12v+amplifier&qid=1568277742&sr=8-1


In the meantime, i would like to learn how to implement the resistive divider, but i don't think i properly understand it. In fact, i am almost certain it is wrong. I would like to wait for advice before going further. as you can probably see, i have a hard time implementing EE theory in real life  :)

Whales:

--- Quote from: Mp3 on September 12, 2019, 08:09:58 am ---Ok, so i found an AC-AC adapter, plugged it in, and instantly blew up the OP-AMP  ^-^ ^-^ ^-^
--- End quote ---

Yeah, you don't want AC :P



--- Quote ---+12V & Ground - 0.8-0.9v
-12V & Ground - -11.9V constant

--- End quote ---

Yep, that's a 12VDC power supply, not a +/-12V (bipolar) supply. 

This will still work fine, IF you are using  an isolated power supply OR the amplifier is designed for use with arbitrary ground-level devices.  Isolated power supplies tend to be the old-fashioned heavy (transformer) wall-worts, not the newer lightweight ones (switchmode power supplies).  I've blown up amplifiers before making this mistake - more than just the voltage matters.

I presume you are feeding audio into your amplifier from a grounded device (eg a computer)?  Try a non-grounded  (battery powered) sound source like a phone or laptop (not plugged into the charger) instead, see if that fixes some of the issues.

Edit: magic has spotted the triple-input and has realised there is no spare op-amp to make a virtual ground if this is a stereo amp.    Woops :P  Follow magic's advice.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod