Yet another post lost to the 'reply to:' feature -------- Original Message -------- Subject: Re: [PICLIST] [EE]: Searching for Digital Potentiometer with highprecision/tolerance Date: Sat, 17 Feb 2001 19:10:03 -0700 From: Robert Rolf Organization: U of Alberta To: Neil Gandler Most digital pots have lousy tempcos (100's of PPM) and not very good linearity (at least not as good as you seem to want). I would suggest individual calibration as the only viable way to make this work (along with a temperature sensor for the chips). I also would suggest that perhaps you should choose a low value XDCP, and make up the bulk of your resistance needs from a precision resistor. That way the contribution of the digital pot's PPM drift is less. In other words a 12.3K precision resistor plus a 200 ohm digital pot. Another technique you might consider is PWM with a FET switch. One can get a very high resolution, highly linear, relatively drift free variable resistance if your circuit allows for the required configuration and bandwidth constraints. This method was commonly used back in the 70's where low noise variable gain cells were built from scratch. http://www.xicor.com/products_xdcps.html 100K, 1024 tap part. SPI control. +/- 300PPM/C, 1.5% linearity http://www.xicor.com/product_summaries/x9110_summary.html http://64.70.222.78/pdf_files/x9110.pdf page 11 for detailed specs. Robert Neil Gandler wrote: > > I would like to use digital potentiometers for digital tuning of an analog filter, as an alternative to a noisy Switched Capacitor filter, or more expensive and complex DSP solution. The problem with most Digital Pots, is that the end to end resistance tolerance is 20% of greater. I am searching for a digital potentiometer with precision resistance settings <1%. So a setting of 12.4K would be 12.4K +/-1%. I would appreciate any leads. Thanks! -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics