LM1875 Need help

Status
Not open for further replies.
Jump to Latest
#42
1. When I hold the phone down it sounds better, a lot better but it still distorts.
Imgur: The most awesome images on the Internet
I made an TDA7231 1.6W, and it had the same problem but in the picture I am using that speaker and it sounds amazing?
2. Is the distortion caused by the no shielding on the input? However when using that other amp on that speaker is sounds beautiful and its 1.6W.

2a. The other amplifier still distorts at high volumes but the speaker I am using it with makes it sound so much nicer and louder until it gets to the distortion levels.
 
Last edited:
#43
crisroxmysoc said:
Hi everyone! I have always been interested in audio designs and always wanted to design an amp from the IC up. I finally did it and I bought 2x LM1875 chips.

The problem I am having is that at higher volumes is really distorting especially for bass and to some degree just treble.
Click to expand...
Not sure you actually have a problem, ALL amps distort above certain level.

I am using this schematic from the Manuel itself (Figure 2.) since I don't have a split supply.
Click to expand...
The schematic is fine.

Another issue I am noticing is that at higher VCC voltages it cuts out? I am not sure if the gain is too much or anything.
Click to expand...
Already solved, you had an important thermal issue.

As of volume and Bass when clipping/distorting: you are using tiny 2 3/4" speakers with 150 to 200 Hz resonance ... *forget* about getting any bass from them.
In fact all your speakers are Bass challenged, except subwoofer.

If they have low efficiency, you will reach clipping easily.

To boot your 20V single supply means you are probably getting some 5W RMS per channel ... nice for background music but not much else ... and definitely NO bass.

That said, congratulations on your successful build, which in my opinion is working fine; if any I suspect the speakers.

Which are not *bad* either, just that they are *designed* to be used with that woofer.

I bet that your amp driving a couple "simple" car speakers, say from 6" to 6x9" to 8" will work fine.

And you can rise power output quite higher, say 20 to 25W RMS, check the datasheet.

So you are taking the right steps in the right direction 🙂
 
#44
JMFahey said:
Not sure you actually have a problem, ALL amps distort above certain level.


The schematic is fine.


Already solved, you had an important thermal issue.

As of volume and Bass when clipping/distorting: you are using tiny 2 3/4" speakers with 150 to 200 Hz resonance ... *forget* about getting any bass from them.
In fact all your speakers are Bass challenged, except subwoofer.

If they have low efficiency, you will reach clipping easily.

To boot your 20V single supply means you are probably getting some 5W RMS per channel ... nice for background music but not much else ... and definitely NO bass.

That said, congratulations on your successful build, which in my opinion is working fine; if any I suspect the speakers.

Which are not *bad* either, just that they are *designed* to be used with that woofer.

I bet that your amp driving a couple "simple" car speakers, say from 6" to 6x9" to 8" will work fine.

And you can rise power output quite higher, say 20 to 25W RMS, check the datasheet.

So you are taking the right steps in the right direction 🙂
Click to expand...

I have this pre-built amplifier https://www.amazon.ca/gp/product/B01IXV4JSQ/ref=od_aui_detailpages00?ie=UTF8&psc=1 and I tested it with those speakers I mentioned and it sounds so good, it even does the bass pretty well? I dont know whats going on🙁
 
#47
Mark Whitney said:
Your amp is oscillating and picking up interference. Understand that what you have built is not the schematic. If you add loops and position the components badly the parasitics dominate the circuit.

https://www.youtube.com/watch?v=xbinXh1ZSqk
Click to expand...

If I didn't build the schematic then what did I build? I don't know why is oscillating the gain is greater than > 10 thus should be in the stable region. I get that parasites can throw the poles to the far right and cause instabilities, but is that the symptoms of distortion at high volumes?
 
#49
Let us look at R6. It is connected between the output and the negative input. It is drawn as a resistor in the schematic. In reality, it also has an inductance that will influence the circuit. Increasing the lead length adds to the total inductance of R6. You can use this to estimate the inductance of each of your leads and components in your layout.

Another problem is the large loop areas in your layout. Loops pickup and transmit. To reduce loop area you can twist the leads together or use a coax cable. An easy way to check this is to trace the current/signal paths in the circuit. The input return (ground) must follow the input all the way back to the source. For the same reason, the + and - from the PSU to the amp should be twisted together to reduce loop area.

"Grounding" is always a problem. Even if you can get all the connections in the right order you can still have ground loops causing hum. Again, tracing the current/signal paths will help you understand how things should be connected. PSU+ to C7 means that there is an exact same current path from C7 to PSU-. As an extreme example, you would not want to connect R4 to PSU- and C3 to C7.
 
#50
Mark Whitney said:
Let us look at R6. It is connected between the output and the negative input. It is drawn as a resistor in the schematic. In reality, it also has an inductance that will influence the circuit. Increasing the lead length adds to the total inductance of R6. You can use this to estimate the inductance of each of your leads and components in your layout.

Another problem is the large loop areas in your layout. Loops pickup and transmit. To reduce loop area you can twist the leads together or use a coax cable. An easy way to check this is to trace the current/signal paths in the circuit. The input return (ground) must follow the input all the way back to the source. For the same reason, the + and - from the PSU to the amp should be twisted together to reduce loop area.

"Grounding" is always a problem. Even if you can get all the connections in the right order you can still have ground loops causing hum. Again, tracing the current/signal paths will help you understand how things should be connected. PSU+ to C7 means that there is an exact same current path from C7 to PSU-. As an extreme example, you would not want to connect R4 to PSU- and C3 to C7.
Click to expand...

and I get that you don't want power signals and input signals to flow in the same ground ( Simple terms). I don't have a picture right now but Ill try to explain what I did with the grounding. I have Pin 3 as my ground so what I did was essentially just connected everything into that pin (obviously splitting it off, I labeled my input signals as 1, Power signals as 2 and output signals as 3). I grouped them to different regions of the breadboard but they all still lead to PIN 3 is this correct or? I am sure nH won't influenced the stability of the amp (I get that they all add up).
 
#54
Confusion reigns.
It all stems from the schematic that misuses ground symbols because of lazyness in drawing the diagram.
Rayma pointed out the problem.

I'll lead you through it:
The bottom of R4 is not ground.
The end of C3 is not ground.
Connect C3 to R4. That new junction is Signal Input.

Your input is shown as one wire. That is wrong.
Your input needs TWO WIRES.
Signal Hot/Flow and Signal Cold/Return.
Signal Hot goes to the top of R4.
Signal Cold goes to the bottom of R4.

The output is not a single wire. It too is two wires, Signal Hot and Return.
The speaker is connected to these two nodes. The return wire comes from the Power Ground at the MF+HF decoupling.
Now use a resistor to provide a voltage reference between Signal Input Return and Speaker Output Return.

Connect the Zobel capacitor, C5 to the decoupling ground.
Connect Pin3 to decoupling ground.
Connect your PSU zero volts to this node.
You now have 6 power circuits meeting at a node.
And one low level reference taken from that node via the extra resistor.
Try using a value from 1r0 to 22r. Lower is usually better, but may leave some hum in your output. Higher reduces the hum.
Add two power diodes across this resistor. 1N4001 in inverse parallel will do.

You have two further connections, R1 and C2.
These contain some power line ripple and so can't be connected to the Signal Return. Instead you have to take them to the decoupling ground.

There is an input filter missing.
Add a series resistor in the line between C1 and R3/Pin1
Add a pF capacitor from Pin1 to the Signal ground. Keep the loop area of this new wiring as small as possible. It is intended to attenuate interference, but if you make the new wiring big in loop area it becomes an aerial picking up interference, the opposite of what is intended.
 
Last edited:
#55
AndrewT said:
Confusion reigns.
It all stems from the schematic that misuses ground symbols because of lazyness in drawing the diagram.
Rayma pointed out the problem.

I'll lead you through it:
The bottom of R4 is not ground.
The end of C3 is not ground.
Connect C3 to R4. That new junction is Signal Input.

Your input is shown as one wire. That is wrong.
Your input needs TWO WIRES.
Signal Hot/Flow and Signal Cold/Return.
Signal Hot goes to the top of R4.
Signal Cold goes to the bottom of R4.

The output is not a single wire. It too is two wires, Signal Hot and Return.
The speaker is connected to these two nodes. The return wire comes from the Power Ground at the MF+HF decoupling.
Now use a resistor to provide a voltage reference between Signal Input Return and Speaker Output Return.

Connect the Zobel capacitor, C5 to the decoupling ground.
Connect Pin3 to decoupling ground.
Connect your PSU zero volts to this node.
You now have 6 power circuits meeting at a node.
And one low level reference taken from that node via the extra resistor.
Try using a value from 1r0 to 22r. Lower is usually better, but may leave some hum in your output. Higher reduces the hum.
Add two power diodes across this resistor. 1N4001 in inverse parallel will do.

You have two further connections, R1 and C2.
These contain some power line ripple and so can't be connected to the Signal Return. Instead you have to take them to the decoupling ground.

There is an input filter missing.
Add a series resistor in the line between C1 and R3/Pin1
Add a pF capacitor from Pin1 to the Signal ground. Keep the loop area of this new wiring as small as possible. It is intended to attenuate interference, but if you make the new wiring big in loop area it becomes an aerial picking up interference, the opposite of what is intended.
Click to expand...

So this is what I understood from what you said. Let me know if its right New setup - Album on Imgur
 
#58
Sorry for stealing the thread, but I just don't have the heart to start another LM1875 thread.
After a few class D amps and then a JLH1969 which is my current daily driver, I stumbled across this article:
6moons audioreviews: PTP Audio Blok20
and
Mono and Stereo High-End Audio Magazine: PTP AUDIO BLOK20 INTEGRATED SOLID STATE AMPLIFIER

It said something about the amp working with only 3 resistors and that's it, but there are definitely some caps also. Anyway, it got me wishing to try it, so I already ordered a few chips.
So what is the absolute bare minimum for this amp? I want to use a SMPS. Is it even worth trying, coming from a real class A?

Build a Bare-Bones “Skeleton” Amplifier — Music | Make: - here is a pretty good step by step explanation of the circuit I'm using as a "reference", being an electronics noob. 🙂
 
Status
Not open for further replies.
Top Bottom