I connected a 50k stepped attenuator to my amplifier (lm3886) with a Minidsp 2x4HD as the source. When I have the volume knob turned all the way down I still have a considerable amount of volume coming out. Turning it all the way up increases the volume, but not by much. Should I have gotten a higher value pot, like 100k?
I think you have the attenuator connected wrong.
The input and output connections are probably reversed.
What's the brand/model, and do you have a data sheet?
Last edited:
I think you're right. I purchased it from ali express, "Dale 23 Step Attenuator Stero Volume potentiometer 2X 50K". I don't have the ground wire connected to it right now and I think that's the problem. There is a single tab for each channel and I connected the input to that. The output is connected directly to the last resistor on the pot which gives me a range of 0 ohms at full volume and 50k ohms at the lowest setting. I don't see any place to connect the ground wire too, though.
There must be a common (ground) connection on the attenuator for it to work.
You can find the common by measuring the resistance between the input
terminal and various other terminals on the control. The terminal that is always
50k, even when the knob is rotated, is the common.
Sorry I didn't post the diagram. I went ahead and wired it up the correct way, but the sound has degraded quite a lot. When I had it connected the wrong way with it acting like a 50k resistor in series with the inputs the sound was absolutely amazing. I wish I had taken some measurements to compare them with the sound I'm getting now, but I didn't. I really don't need volume control, it's just that my source produces too much noise going into the preamp on the board. The noise was reduced with the 50k in series to a low enough level that it won't bother me. Is there any reason that I can't remove the pot and stick a 50k resistor in series with the inputs? The distortion with the pot wired correctly is way too high and it makes the sound grainy and smeared.
That may work fine, if the shunt capacitance after the series 50k resistor is reasonably low.
If not, use a 10k resistor instead, unless the source would have a problem with that.
If you really want to try to get by without a volume control, I'd also add a mute switch that shorts
to ground after the series resistor.
The reason I don't need volume control is because I use the one on my PC. With the 50k series resistance of the pot it lowered the level of noise. If I turned the pot up (lowering there series resistance) it got louder but the audio quality suffered. At 50k I had more than enough volume, so I don't see any need to go lower unless there is a reason I should. I just want to be sure I wouldn't be hurting anything with the resistor in place of the voltage divider. It really does sound like s#!+ with the pot wired correctly. I found some older measurements that closely match the ones I just took and there is a considerable amount of added distortion with the pot.
Sure, give it a shot, and use a quality metal film resistor.
Which attenuator do you have?
The attenuator is a 50k stepped with Dale resistors (23 steps, I think). All of the resistors on thanks are Dale also, and there are no dc blocking caps in the signal chain on the amp. Except for the output caps on the minidsp 2x4hd, anyway. I went ahead and ordered some 50k Dale resistors to try instead of the pot. I know the output impedance of the minidsp is 560 ohms but I'm not sure of the input impedance of the amp. The input goes through a 523 ohm resistor while a paralleled 56.2k resistor and a 680pf cap run to ground.
A passive attenuator should not have audible distortion.
Is the minidsp overloading the input of your amplifier?
Are you sure there is no DC going to the amplifier input?
By adding a series 50k, you are cutting the input level by half,
either AC or DC.
There is no measurable DC current going into the op amp. I believe it is over loading the input of op amp. If I bypass the feedback loop and run the op amp as a buffer there is very little distortion and low noise but I can still hear it if I'm a few inches from the speaker. The problem is that the lm3886's are not putting out much wattage even at Max volume with the op amp run as a buffer. And if I use it with gain then the noise is quite audible within a couple of feet of the speakers.
If you go to page 2 of a previous thread you will see the schematic. The link is: LM3886 help, please!
That's really hard to decipher. I would start by patching in any coupling capacitor
that you happen to have, before the power amp. If that fixes the problem, there
must be DC at the amp input, and so you have to isolate it.
If that does not help, it must be a level problem instead.
You said there is a coupling capacitor at the source output, right?
There is an output coupling capacitor on the output of the minidsp. On the amp there was a 4.7uf coupling cap between the op amp and the DC servo, with an option to bypass the cap. I took measurements with an without the cap and the only difference is lower distortion without it. I also measured the DC voltage at the output of the minidsp and the output of the amp with an oscilloscope, both were super low.
If things work without the attenuator, but you want a level cut for noise reduction,
then the series resistor should be ok. That schematic is way too much work to decipher.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.