Step attenuator problem

[sci525]

Hey everyone and thanks for reading my post. First I put a VALAB 20K 23 step attenuator at the output of my lm3886 chipamp board. Very slight increase in volume over the first 21 steps then a huge jump at 22 and nothing at 23. So I decided to see what would happen if I put the attenuator at the input of the amp board. Very slight increase in volume over the first 11 steps then a huge jump at 12 and small progression from there. Also there was a hum.

Any ideas as to what may be wrong? Link to the attenuator below. Thanks!

VALAB 23 Step Attenuator Potentiometer 20K Stereo Log | eBay

[lanchile]

From the RCA terminals go to the input of potentiometer and from potentiometer to input of amplifier board. Make sure to connect the ground. and put it as far as possible from Transformer.

PS: are you sure the potentiometer is connected right?

[AndrewT]

Sci,
the attenuator has 6 wire connections.
3 wires belong to the Left channel.
3 wires belong to the Right channel.

Does your retailer show you how to connect these 6 wires?

If not then ask them for details.

[sci525]

Yeah, that is what I did the second time around. The symptom changed, but there is still a problem. The hum is likely from the transformer, and I have some ideas on how to fix it, but one problem at a time. As for having the attenuator connected properly, I am confident that I have hooked it up as indicated by the picture on the ebay listing. However, I have no other reference. I hooked it up exactly the same way when put before the amp and after the amp and it responded differently as detailed above?

[sesebe]

Can you put a drawing in the these two cases?
You connect the atenuator on output of LM3886? And after atenuator you connect the speaker?

Plese post a schematic. If I'm right you do not respect the "input impedance" and the "output impedance" ranges of the atenuator (to not say more) .

[GloBug]

Bust out the DMM and measure those SMD, you should be able to sort it out.
Check the values with the numbers on the resistors, look for ones with no contact.

If you are unsure of the "wiper", find two wires that dont change the impedance on your ohm meter when you turn the dial, the third wire will be the "wiper".

You should be able to figure out if it is defective or not and which way to hook it up with your meter. Use a "regular" pot for reference if needed.

[sesebe]

Now I see the photos. Pay attention to SMD. Is the worst possible way to solder SMD.
Due to variations in temperature and mechanical stress, appear cracks between resistive layer and the soldered zone.
These cracks are invisible to the naked eye and I find very often when SMD are soldered in this way. You need a microscopes to see these cracks.
And please clarify the connection schematic.

[sci525]

Sesebe, At first the attenuator was connected between the amp and the speakers. I'm a newb, what can I say. But, now I have it connected between in the RCA input and the amp.

The right channel works pretty well as is, but doesn't get to full volume, and the left channel suffers from the issue mentioned earlier where there is little volume change till the 12th step when it makes a huge jump.

I took out the multi meter but didn't spend much time on it because it only took me a couple seconds to notice that the impedance's where not the same at the same step for each channel.

Later today I will resolder it and see if that helps, if not I will drop in a cheap reference pot and make sure the issue is in the attenuator.
Is it likely that I blew out the attenuator having it originally hooked up after the amp, or was it more likely defective?

[sesebe]

Hi sci525,

Seems that in the left channel you already have a cracked SMD.
My advice is to change all resistor from both channels with true-hole resistors to not have problems in the future. I know is not easy but you will not regret. Small variation from original values are not a problem. You will not see the difference from original values and new ones changed.

For the right channel, at maximum the signal must be directed connected so the only problem can be from a too high impedance of the source (if without attenuator you have full power).

[sci525]

Thanks for your input everyone! I resoldered the whole thing very methodically and it is working great. Sesebe, I'm going to let it run right now, but if there is another crack I'll take your suggestion and replace the resistors with through hole.

The VALAB 23 step attenuator sounds great, good value, BUT be prepared to spend some time getting it working.