| autoexec |
hello
i had a search around, i couldn't seem to find anything for it, i cant seem to see any specific place for noobs, so as its chipamp based, here i am
i constructed a lm3886 based amp with point-to-point wiring
when the speaker is connected, it sits fully foreware (then backward if i swap the wires to the speaker around)
from what i can gather, this would mean that there is a large dc signal being outputed
is this a common problem for someone with little experience (only made a lm3875 b4, but that one worked)
is there a simple thing i did wrong, or is this a complex problem that is going to take me forever to fix?
cheers
matt |
|
|
| Nuuk |
Yes, it would appear to be a large DC offset. I hope that you have been using an old speaker for testing!
Without a circuit diagram of what you have done, we cannot really help you though! ;) |
|
|
| autoexec |
i followed the data sheet, exept i didn't use a cap on the mute, would that be the cause?
i used a 1.8k on Rm
im using 56v CT trans, with 4700uf instead of the 1000uf on the datasheet
what causes a dc offset? |
|
|
| Nuuk |
| quote: | | im using 56v CT trans, with 4700uf instead of the 1000uf on the datasheet |
I have not yet built any 3886 GC's so I hope somebody else will help you but if your transformer has 56 VAC secondaries, your rail voltages will be around 80 volts DC which is far to high for that chip! :att'n: |
|
|
| Sheldon |
Yes, that voltage would be a problem, unless you are talking about 28-0-28.
If voltage is not the problem, one mistake that will cause the amp to output the full voltage of either rail, is to have an open circuit in the feedback loop. The amp is a differential amp, so it tries to zero the difference between the inverting and non-inverting inputs by changing the output. Feed back is required for this to work. No feedback, the amp pushes all the way to the rail.
Sheldon |
|
|
| mateo88 |
Try putting a 20k resistor from input to ground. Without a resistor there and with no source connected, the output will be the straight DC that's only a few volts less than what your rails are at.
You really don't need a capacitor for the mute pin. I think the capacitor just allows for a "pop and click free" startup and shutdown.
Measure the voltage with a multimeter, also, before connecting to a speaker.
Good luck!:smash: :) |
|
|
| thomas997 |
| quote: | Originally posted by mateo88
Try putting a 20k resistor from input to ground. Without a resistor there and with no source connected, the output will be the straight DC that's only a few volts less than what your rails are at. |
Yeah, also try connecting some sort of input, something cheap incase there is a problem with the circuit. |
|
|
| pinkmouse |
| Or just short the input for the time being... |
|
|
| autoexec |
it doesn't change when i put an input on it
its a 28-0-28 trans so i end up with around 40v rails to give me 80v
| quote: | Originally posted by Sheldon
If voltage is not the problem, one mistake that will cause the amp to output the full voltage of either rail, is to have an open circuit in the feedback loop. The amp is a differential amp, so it tries to zero the difference between the inverting and non-inverting inputs by changing the output. Feed back is required for this to work. No feedback, the amp pushes all the way to the rail. |
i dont really understand how an amp works at a transistor level, so i guess i better read up on that, i just know how to solder and follow a diagram
which pins are called the inverting and non-inverting inputs?
is it the Vin+ and Vin- ? or are you talking about that connection between the output and the Vin- ? |
|
|
| Sheldon |
| quote: | Originally posted by autoexec
it doesn't change when i put an input on it
its a 28-0-28 trans so i end up with around 40v rails to give me 80v
i dont really understand how an amp works at a transistor level, so i guess i better read up on that, i just know how to solder and follow a diagram
which pins are called the inverting and non-inverting inputs?
is it the Vin+ and Vin- ? or are you talking about that connection between the output and the Vin- ? |
Vin+ is the non-inverting input. Vin- is the inverting input. If the input goes to Vin+, the amp is non-inverting (does not invert the signal). If the input goes to Vin- the amp will invert the signal at the output. Feedback from the output always goes to Vin-. You don't have to understand all the inner workings. For a simplified version of a non-inverting differential amp: If you apply a signal of +10mv to the non-inverting input, the amp will raise the output voltage until it sees a signal of +10mv at the inverting input. If your feedback ratio is 20:1 (the ratio of the feedback resistors) the output will need to raise to 200mv before Vin- is at +10mv. For an inverting amp: If you apply +10mv to the inverting input (the Vin+ is grounded), the amp will go to -200mv to get the Vin- back to 0 volts. If you've got an open connection from the output to the inverting input, the amp will go positive (negative for the inverted amp) until it reaches its maximum, which is just short of the rail voltage.
try mateo88 and pinkmouses suggestions too. Easy things first.
Sheldon |
|
|
| autoexec |
i put a 22k across the input, still get the 40v
since its point to point, im going to undo the connections between the components and start that again, hoping not to make the same mistake which i still cant find |
|
|
| autoexec |
| before i do that, can someone tell me if i need to connect both V+ pins, or can i just choose one of them |
|
|
| Nuuk |
| quote: | | before i do that, can someone tell me if i need to connect both V+ pins, or can i just choose one of them |
Both of them. |
|
|
| autoexec |
its all working now
thanks everyone |
|
|
| pinkmouse |
| What was the problem? |
|
|
| Nuuk |
| quote: | | What was the problem? |
I'd guess posts 12 and 13 may give us a clue! ;) |
|
|
| autoexec |
actually, i dont know what the problem was
i just rebuilt it and it worked this time. |
|
|
|
