Author Options:

LM1857T circuit help? Answered

I have been wanting to build a good amp for a while now and i just got a LM1875T. so i put together this circuit diagram: http://imgur.com/ov3Ikd6

also i am running the audio signal thru an LM386 as a pre-amp/buffer before it hits this. i am using a speaker rated  at "4ohm 80W"
and i am using a 30V 2A power supply.

my questions:

#1 the circuit does not run at all if R2 (22k between pin 1 and ground) is there so i removed it for testing. why is it there if it keeps circuit from working?

#2 C8 (100uf capacitor between pin 3 and ground) is reversed? i thought it might have been a typo on the Texas instruments site so i put it normally. nothing happened. so i put it the way they showed it and it worked! but my cap blew 5 min later... (i used a 50v 100uf) so i replaced it with a 200V 147uf capacitor. it started heating up after 1 minute so i disconnected it to come ask. should i use a tantalum 50V capacitor?


That circuit is using dual supplies, you have only a single supply. I've attached the correct application circuit.

You may well have destroyed your speaker. If the sound is distorted, try another one.


I was using a speaker at the time rated for over 100W so i doubt i destroyed it ;) and it was not a distortion it was more of a loud oscillation.

Thank you i will attempt this one... why can't they give more detail in the datasheets?

No speaker is designed to accept DC, which is likely what you have given it :-) "DC" is anything less than the minimum response frequency of the speaker.

The circuits in a datasheet are not for beginners, but for those skilled in the art, they are not an attempt to teach amplifier design.

Also i was wondering what -VEE was... i thought it was the - input not another power supply -_- and the speaker had a capacitor built in so it was only getting AC.

-Vee is the negative supply rail.

I've not seen a speaker with a "built in" capacitor, ever.

It is a car subwoofer :P it has a 50V 220uf capacitor on it i can trace the wires :)

Ok built the circuit based on steveastouk's picture. worked perfectly
until a capacitor blew (why did it have to blow right as my head was
next to it... that was loud) i thought using 2 220uf 16v in series would
give 220uf 32volts but they both blew. now i am using about 300uf of
other capacitors in parallel but the output is very noise (needs a
higher capacitance but i am out of capacitors rated over 30v)


4 years ago

A capacitor blows up because your power amplifier put over 16 peak volts to the capacitor and wanted to make sure you listen because it could have been your precious eye.

A capacitor also heats up due to RMS ( RIPPLE ) current.

There is something strange here. There is no way your speaker should be using enough power too overheat an electrolytic capacitor.

I used 2 in series so it should have been able to handle 32v. still odd...

How did you guarantee there was no possibility of a negative DC voltage applied to the capacitor.


Yes, in series the DC voltages do add up and the capacity goes down decreases.

But the signal going to the speaker cycles goes Positive and Negative and the capacitors will see a reverse DAMAGING voltage about half the time !!!

There are special 20 uF to 300 uF AC 400vac capacitors but you didn't know that before.

My above schematic only allows each capacitor to see a DC voltage.

And acts like an equivalent AC capacitor for a speaker signal waveform.


Granted the LM1875 may not blow the 2200uF electrolytic because it biases the output in plus DC, But don't try it on a commercial Hi-Fi amplifier that really does put out an unbiased AC signal.

Now that Iv confused everyone, an oscilloscope will show what a biased signal looks like when AC is added to DC.

Ahhh now i understand what you meant. didn't know it went into negative voltage (it really should have clicked in my brain sooner. i do a bit of audio editing in audacity.) I thank you for explaining it very well :)

sadly i don't have an oscilloscope ;( would be useful for many of my projects...

R2 shouldn't be there at that value try a larger ohm resistor like one meg you do need the resistor to drain the capacitor.

As for capacitor 8 it is capacitor 6 and it is in the right way try the tantalum.

This is the problem with circuit simulator programs ideal models don't always fit the real world so you need to tweak the circuit in the real world.


I attempted a few resistors (2M 1M 500K 470K) all seem to make a loud "WOUB WOUB" about 3 times a second out of speaker when added to circuit.

Well if the only way the amp works is if the resistor is out of the circuit then you need the resistor out.

Constructed like your schematic the amp is on full volume and that can add to your capacitor troubles.


Thank you for the help. how would i change the volume? can i use a 5K pot?

I need more details if a 5K will work but yes.

Put the input to one end of the 5K pot and the other to ground to the center connect the input to the amp.

I would up load a pic but the server is bunged.


That is what i thought but i am not at my board so i thought i would ask. that is why i uploaded my pic somewhere else :)


4 years ago

As Steve said, a speaker needs AC Alternating Current to provide energy to the speaker cone to push and pull with the tiny wire coil which is the most easily damaged ( burned out wire path broken ) by passing too much power to the unit.

Ballistic means simply the fact that an electric sine wave speaker input, does not generate a true audio pressure sine wave output. This means sound reproduction is distorted for the serious audiophile. Ergo an amplifier that takes into account the dynamic speaker cone inconsistencies is expen$ive !!