Hey there,
I want to replace the LM358 in this circuit with an AD8656. As I am a beginner: Can you please help me and say whether this will possibly work? The LM358 is hissing.
It's SOIC8 right?
See a picture of the circuit a bit farther down.
Best regards
I want to replace the LM358 in this circuit with an AD8656. As I am a beginner: Can you please help me and say whether this will possibly work? The LM358 is hissing.
It's SOIC8 right?
See a picture of the circuit a bit farther down.
Best regards
Last edited:
Give it a measure and be sure. It looks bigger than SOIC in the picture:
Will the AD8656 work? One vital difference is the 5 volt maximum supply voltage for the AD device. You need to check it runs on no more than this as measured between pins 4 and 8. Even if that si OK there are no guarantees because the part is a far faster device and you might hit issues with instability or oscillation.
You really also need to know exactly what the LM358 does in your circuit. Is it even used for audio processing for example.
What exactly is the board? Is it a DAC of some sort? I see the 8416 chip.
Will the AD8656 work? One vital difference is the 5 volt maximum supply voltage for the AD device. You need to check it runs on no more than this as measured between pins 4 and 8. Even if that si OK there are no guarantees because the part is a far faster device and you might hit issues with instability or oscillation.
You really also need to know exactly what the LM358 does in your circuit. Is it even used for audio processing for example.
What exactly is the board? Is it a DAC of some sort? I see the 8416 chip.
So that does look like its used in the audio path as the components duplicate each side of the opamp. Probably a line out buffer/filter.
I wouldn't like to suggest anything for this without testing it first. The very low supply voltage is a dominant concern and the LM358 excels on very low voltage operation. Any replacements would have to be tested for real looking at for example whether it could deliver a clean undistorted output at maximum level.
Although this relates to more 'normal' usage of opamps this is also what can and does sometimes happen when just swapping for something that initially might seem an improvement.
I wouldn't like to suggest anything for this without testing it first. The very low supply voltage is a dominant concern and the LM358 excels on very low voltage operation. Any replacements would have to be tested for real looking at for example whether it could deliver a clean undistorted output at maximum level.
Although this relates to more 'normal' usage of opamps this is also what can and does sometimes happen when just swapping for something that initially might seem an improvement.
But SGM358 is not a LM358, neither an audio dedicated product :
https://www.alldatasheet.com/datasheet-pdf/view/223967/SGMICRO/SGM358.html
https://www.alldatasheet.com/datasheet-pdf/view/223967/SGMICRO/SGM358.html
Well spotted. That actually makes a lot more sense for a single rail 5 volt application.
Wow! Thank you, both! That helps a lot.
But I think it is used for audio in this case... If I follow the rales, the audio out leads to the left and right RCA Connectors.
But I think it is used for audio in this case... If I follow the rales, the audio out leads to the left and right RCA Connectors.
Your close up picture clearly shows the circuitry is the same on both sides of the chip meaning it is as you say, audio L and R channels.
Thanks mooly.
I think I found myself a new hobby and a friendly forum for it. I ordered a AD8656 and a MUSE8832e to try.
Looking so much forward to do experiments with the DAC. The sgm358 is a crappy choice here IMHO.
I will report my findings.
I will also make shure to check for stability. Hope my small RIGOL scope will be enough to test.
So far, I just "played" with microcontrollers.
I think I found myself a new hobby and a friendly forum for it. I ordered a AD8656 and a MUSE8832e to try.
Looking so much forward to do experiments with the DAC. The sgm358 is a crappy choice here IMHO.
I will report my findings.
I will also make shure to check for stability. Hope my small RIGOL scope will be enough to test.
So far, I just "played" with microcontrollers.
Hmmm... its great that you are experimenting 🙂
I don't want to add any negativity here 😱 but I'm reading the data sheet on the Muse and see this:
I'm assuming you mean this one:
MUSES8832
This is a typical DAC filter circuit (from the data sheet) and the two resistors set the gain. You might be OK because I see a 12k and 10k which at face value give a gain of just under 7db
I don't want to add any negativity here 😱 but I'm reading the data sheet on the Muse and see this:
I'm assuming you mean this one:
MUSES8832
This is a typical DAC filter circuit (from the data sheet) and the two resistors set the gain. You might be OK because I see a 12k and 10k which at face value give a gain of just under 7db
Ideally we need to see a circuit diagram 🙂
Resistors configured like those set the gain (standard opamp theory) but if you alter the overall gain you will run into problems of distortion and clipping because of the low supply voltage. There are other tricks we can use in some situations such as altering the 'noise gain' but lets not go there yet. That technique adds a resistor between the two inputs. The overall signal gain stays the same however.
Resistors configured like those set the gain (standard opamp theory) but if you alter the overall gain you will run into problems of distortion and clipping because of the low supply voltage. There are other tricks we can use in some situations such as altering the 'noise gain' but lets not go there yet. That technique adds a resistor between the two inputs. The overall signal gain stays the same however.
Gain for those resistors is the 12k feedback resistor and the 10k (which without seeing a circuit we assume goes to a low impedance point... which won't be ground in a single rail design)
So we have gain = (R1/R2)+1 which is (12000/10000) + 1 which is 2.2
So it has a voltage gain of 2.2 numerically,
1 volt in gives 2.2 volts out.
In db that becomes:
db= 20log(V1/V2) which is 20log(2.2/1) which is 6.848db
(you can pick any voltages you want to work the gain out, even crazy ones. Ten million volts in would give 10,000,000 * 2.2 volts output which is 22,000,000 volts. Apply the same formula and you get the same answer of 6.8db. Same for applying 23.6 micro volts at the input... anything you want... multiply that by 2.2 to get the output voltage and the db gain always comes out the same using those numbers)
So we have gain = (R1/R2)+1 which is (12000/10000) + 1 which is 2.2
So it has a voltage gain of 2.2 numerically,
1 volt in gives 2.2 volts out.
In db that becomes:
db= 20log(V1/V2) which is 20log(2.2/1) which is 6.848db
(you can pick any voltages you want to work the gain out, even crazy ones. Ten million volts in would give 10,000,000 * 2.2 volts output which is 22,000,000 volts. Apply the same formula and you get the same answer of 6.8db. Same for applying 23.6 micro volts at the input... anything you want... multiply that by 2.2 to get the output voltage and the db gain always comes out the same using those numbers)
The MUSES8832 is in. 🙂 I just risked it and tried it out. I really like the sound. As I have two of these DAC, I can do comparisons. For me, personally, it is enough to hear it sound better. To my liking, even WORLDS better. But if anyone wants me to do some measurements, please let me know. I will sit down now and enjoy some music.
Attachments
Thanks for the good vibes, @Mooly, I was actually quite nervous. Glad the result does not show that.
I know, perception of audio is very personal. Nevertheless, I want to share, that to my ears the sound got way warmer without loosing any brilliance.
As soon as the AD8656 arrives, I will give it a try in my other D03K.
I know, perception of audio is very personal. Nevertheless, I want to share, that to my ears the sound got way warmer without loosing any brilliance.
As soon as the AD8656 arrives, I will give it a try in my other D03K.
It looks as good as a factory fit and close up shots are very very revealing indeed.
Well, not that good of a job this time. 😀 I used too little flux. But I learned that the marked cap is "adding resistance" to the op amp. My hot air blew it away, and I was lucky to find it again. I didn't resolder it properly, and the right channel was softer than the left one. Now I fixed it.
To be honest: I hear no, absolute no difference between these two op amps (MUSES8832 and AD8656). I asked my wife, to switch between channels. I could easily hear the difference between the SGM358 and the MUSES, though. So in case anybody has to decide which part to throw in, I would advise using the AD8656 as it is cheaper.
To be honest: I hear no, absolute no difference between these two op amps (MUSES8832 and AD8656). I asked my wife, to switch between channels. I could easily hear the difference between the SGM358 and the MUSES, though. So in case anybody has to decide which part to throw in, I would advise using the AD8656 as it is cheaper.
Last edited: