Noise with a digital pot (X9511) - diyAudio
Go Back   Home > Forums > Design & Build > Parts

Parts Where to get, and how to make the best bits. PCB's, caps, transformers, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 15th June 2005, 11:21 PM   #1
Ayah is offline Ayah  Sweden
Account disabled at member's request
 
Join Date: May 2005
Location: Sweden
Default Noise with a digital pot (X9511)

Hi,

I have a problem with a circuit that I'm working with. There are some inputs on a 4051 switch (+/- 5V), output to a non-inverting buffer and an inverting op that work as an attenuator with a digital pot (controlled by a PIC). If I use a sinus wave (0.4V, 2kHz) on as input I get a very noisy and "unstable" (on an oscilloscope) output. The picture I attached is not from my output but it looks similar.

I'm sure the problem is the X9511 (10k) because if I disconnect it and use a normal resistor instead (for example 5k) the output is half the input and it's the same "shape" and frequency on the output.

Is it my design that isn't correct or could it be something else? I'm also interested in other ways how to improve the circuit except for the noise.


Fredrik
Attached Images
File Type: jpg circuit.jpg (32.8 KB, 194 views)
  Reply With Quote
Old 16th June 2005, 08:01 AM   #2
Ayah is offline Ayah  Sweden
Account disabled at member's request
 
Join Date: May 2005
Location: Sweden
the sinus wave with nosie..
Attached Images
File Type: jpg sinus-noise.jpg (7.2 KB, 145 views)
  Reply With Quote
Old 16th June 2005, 12:31 PM   #3
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Looks like you have "clock noise" from the PIC coupling onto the NE5532's etc. -- decouple the supply pins of the OP-AMPs, multiplexer and digital pot with 100nF ceramic and 1 uF tantalum capacitors. Also -- the MUX address and Digital Pot control pins -- you can feed these with a 10K resistor in series -- this will decouple a lot of noise from the PIC. Be careful with layout --
  Reply With Quote
Old 17th June 2005, 05:18 AM   #4
Ayah is offline Ayah  Sweden
Account disabled at member's request
 
Join Date: May 2005
Location: Sweden
Thanks for your answer. I added resistors on the signals from the PIC and I it helped some. I also noticed that my DC supply wasn't very "clean" so I put capacitors (all I had was polyester) to the ground from +/- 5V which helped a lot.

In my circuit I use the same +5V for all chips and a -5V for the OP's and the MUX (VEE). Do I have to decouple all the supply pins or isn't it enough to do it once at the +/- 5V supply since they're all linked together? Why use tantalum capacitors?

How is the "clock noise" transmitted to the other circuits? Through ground?

Thanks again,
Fredrik
  Reply With Quote
Old 17th June 2005, 04:09 PM   #5
diyAudio Member
 
jackinnj's Avatar
 
Join Date: Apr 2002
Location: Llanddewi Brefi, NJ
Quote:
Originally posted by Ayah
Thanks for your answer. I added resistors on the signals from the PIC and I it helped some. I also noticed that my DC supply wasn't very "clean" so I put capacitors (all I had was polyester) to the ground from +/- 5V which helped a lot.

In my circuit I use the same +5V for all chips and a -5V for the OP's and the MUX (VEE). Do I have to decouple all the supply pins or isn't it enough to do it once at the +/- 5V supply since they're all linked together? Why use tantalum capacitors?

How is the "clock noise" transmitted to the other circuits? Through ground?

Thanks again,
Fredrik
Each of the +/-V pins of the analog AND digital components should be bypassed -- it isn't sufficient to bypass the supply alone. the traces themselves will pick up radiated energy from the micro-processor. while you can't hear it, the radiated energy is sometimes sufficient to cause opamps to behave in a less than ideal way.

The switching speed of the PIC may not be high enough for it to matter (are you using 4MHz or 20MHz?) but a ceramic seems to be preferred to polyester film in this application and they are cheap. using a tantalum or aluminum electrolytic is a bit of "belt and suspenders".

One last thing -- I believe that the NE5532 will operate with lower noise and THD if the supply voltage is higher. There are families of devices that are designed to work with 3.3V or 5V, but the NE5532 isn't of that generation. If you are committed to the NE5532 that's OK, it is a much venerated opamp

another thing -- see the attached JPEG -- you might find this method of coupling the PIC to your MUX and POT works even better.

btw, there is a good thread with some knowledgeable (SP?) guys on multilayer ground planes. since you have mixed analog-digital you should hop on board.
http://www.diyaudio.com/forums/showt...245#post665245
Attached Images
File Type: jpg resistors.jpg (41.4 KB, 132 views)
  Reply With Quote
Old 23rd June 2005, 08:02 AM   #6
mobyd is offline mobyd  New Zealand
diyAudio Member
 
mobyd's Avatar
 
Join Date: Sep 2004
Location: Auckland
Those Xicor 95 series pots use an on-chip substrate bias generator (charge pump) and are as noisy as all hell.
You are better off using a pot designed for audio (quiet) like a Dallas DS1267 or subsequent versions thereof. There are many others by other makers as well, but the Xicor is easily the worst.
M
  Reply With Quote
Old 23rd June 2005, 04:04 PM   #7
macboy is offline macboy  Canada
diyAudio Member
 
Join Date: Oct 2003
Location: Ottawa, Canada
I can see one likely source of trouble with your circuit: the inverting opamp can be operated with a gain of less than 1 (depending on the pot value of course). This is may cause it to be unstable, and oscillate, which can produce an output much like what you are seeing. Is there any particular reason why you chose to use the pot to vary the gain of the amp, rather than simply to attenuate the signal?

I'd recommend using the digital pot as a variable attenuator between the non-inverting buffer, and the inverting amp (connect VH as input, VL grounded, and W as output). Select the gain of the inverting amp to give you the desired maximum output level with the pot at maximum output.

If the pot has a linear taper rather than log, you can approximate a log taper by putting a resistor (equal to 1/5 to 1/10 of the pot resistance) between the VH and VL terminals.

I would recommend putting a 10 to 22 ohm resistor in series with the power supply pins of each component. Make sure to use bypass capacitors after these resistors, as close as possible to each device's power supply pins. This forms an RC filter which will help keep digital noise out of the power supplies. You'd be surprised how well this can work.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
noise problem with digital line and s/i ss stage then_dude Digital Source 2 18th November 2007 09:04 PM
RF noise in digital equipment... (detection with AM radio?) ash_dac Digital Source 2 22nd April 2006 01:46 PM
Digital Pink Noise generator...? Lyra Digital Source 12 22nd March 2006 02:06 AM
discman car conexion digital noise mbahi Car Audio 1 1st June 2004 10:32 PM
optocouplers to reduce digital noise? dorkus Digital Source 9 23rd April 2002 05:41 AM


New To Site? Need Help?

All times are GMT. The time now is 12:49 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2