[Possum]

Recently, after I added a digital pot (the DS1802) to my amp in place of a Panasonic that used conductive plastice, I noticed 2 situations where I would get noise, or more noise than before:

1) When I move through certain areas of my room with my headphones on, I would here very soft clicking. If I moved back and forth through a specific spot, I could repeat the clicking. I tried moving through the same spot with the headphones unplugged, and did not hear any clicking, so I'm guess there may be RF or EMI interference being transfered from the headphone cable back to the amp, then back to my headphones as this clicking noise. Could this be the cause, and could I remedy this with a few ferrite beads around the wires soldered to my headphone jack?

2) Certain sounds, if they were somewhat noisy with my analog pot, are now very very noisy with my digital pot. So noisy that you'd think something was clipping. Examples: the starting sound "Welcome to Gamespy" when you load the Gamespy online-game server browsing program. Before I could here a little bit of clipping-like noise with and without my amp. Now the noise is very pronounced. Another example, downloading the poorly ripped 320-kbps (48000 Hz sampling rate) MP3's of Blink 182's "Rock Show" song from certain sharing services. With the old pot it just sounded like the song was ripped very loudly, however, with the digital pot it just sounds like it's clipping vert severly. Could this have ANYTHING to do with my previous careless assumption that the digital pot did not depend on a specific current requirement (see thread about voltage dividers)?

EDIT: Case 2 is definitely clipping. I was able to reproduce severe clipping to the point where the digital pot resets itself. It seems to happen if there's A LOT of bass information in the song. I simulated this with any song with decent bass, then used my portable CDP's bass enhancer to add bass, and got the clipping. So the new questions is: Why is this happening with the digital pot?

 

[ppl]

Check the grounds on the Analog and Digital parts of the Chip. ya make sure the Chip is supplied with enought Current and voltage. Clipping on large input signals using this type of attuator has always been a problem that analog pots do not have. I was hoping that this DS-1802 chip just used a switched resistor array as denoted in the Block Diagram in whitch case Input overload should not be a problem. Ferite beeds on all input and output Hot lines of any amp is a good idea also i always use them. I am about to do the same thing using that chip. wen i get some time. let me know how this works out.

 

[thomas]

i'm also planning on using these pots....

Can you please tell me if they're ttl/cmos/other, and what voltage they require? thanks...

 

[ppl]

The low curent drain of the DS-1802 would tend to indicate CMOS but go to maxium semiconductor site and check out the data sheet. The supply voltage is specified at about 2.5-5.0 volts.

 

[eric343]

Well, I know that the cable that I use to connect my Cmoy to the source works as an AM antenna; when I hold the plug in the air and I'm in the right spot I can get talk radio

 

[ppl]

if you are using an fet input opamp i would investigate the possibility of oscilation. Jfets most often do not Demodulate AM signals like Bipolars due In most all cases that anfet input amp is playing AM radio the Circuit was unstable. What happens is the oscilation forms a supperhetrodine receiver and since the input is not tuned any atray am input can be demodulated and heared also if you tutch the input of most any amp with the headphones on your head then this will drive evean a stable amp into unstability and thus pick up AM Radio. In any case ferite Beeds should be on the Input leeds

 

[Possum]

Well, I was able to try a few combinations yesterday before my power went out for the rest of the afternoon and night.

Using a voltage divider of lower resistors allowed more music to come through with the clipping, and higher resistors allowed almost no music with the clipping (so you'd just hear thumpy-click thumpy-click thumpy-click etc.). So I tried as low a combination as I could get (3.3 ohms in between Vout-and ground, and 10 ohms between Vin and Vout to provide the required Vcc voltage, Vin = ~17V) - well, I got to see a resistor burn and flame up for the first time

. The 10 ohm was toast. Then I tried doubling up the resistors: 2 3.3 ohms in the back, and 2 10 ohms in the front in parallel. Yay! TWO resistors flamed and burnt up (hint hint for you pyros looking for good electronically-startable fuses). I was going to try paralleling more resistors of slightly higher value so the end value is low resistance but high wattage. I hope I'm on the right track. If not, warn me so I don't get too fire happy.

 

[artmusic247]

I have two questions about ferites. How do ferite attachments work and would it help a "normal" signal, one that's has no problem with radio waves? Thanks.

 

[ppl]

Ferrite Beeds are a small graphite Beed with a Hole in the Center. Thay slip over the Wire. Place one upon each Hot Wire in both Input and output Leeds, Close to the Jack.. Thay act as an inductor to roll off the Radio frequency waves with out affecting the Audio band at all. In this way all the desireable qualities of the wire are intact yet still afford RFI attinuation. It is also a good idea to use these on the Non ground side of the DC powersupply input. For AC operated units radio Shack selles Large Freetie Beeds that go upon the AC mains Line Cord.

 

[Tomo]

Umm ...

No resistors should burn because of human error. Resistors do have rights too. >

Your use of voltage divider is probably incorrect. Voltage divider is suited to set a voltage, but it is not suited for high current applications. If you need to set some other voltage than the voltage from the PSU. You can use simple regulator assembly with 78XX family chips. (or you can take your PSU apart.)

I assume you have to do this to power your digital pot. In this case, you should build another regulator assembly.

Your input should not be larger than supply rail. You should set up the input source to satisfy that condition.

Tomo

P.S. My stepped attenuator is not picking up RF and stays very quiet.

 

[Possum]

Yah, in the other thread I ended up using an LM317 to provide 4.96 volts for the IC's Vcc. I also had to lower the input signal so the sound wouldn't clip, as aos and you have pointed out as a cause.