Message-Id: <5.1.0.14.0.20040513093940.00ba1718@u0019445.kuleuven.be>
Date: Thu, 13 May 2004 09:56:54 +0200
To: rsgb_lf_group@blacksheep.org
From: Rik Strobbe <rik.strobbe@fys.kuleuven.ac.be>
In-Reply-To: <40A2578C.2080600@usa.net>
Mime-Version: 1.0
Subject: Re: LF: VDC source with microvolt resolution
Content-Type: text/plain; charset=us-ascii; format=flowed
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Content-transfer-encoding: 8bit

Hi Alberto,

If you (or your company) can spend some money on it you could consider 
buying a ultra-stable voltage/current source such as
Analogic Model 8200 (http://www.analogic.com/Level4/TM4/8200.pdf).

An alternative route : in the early days, when anything above 12 bit DAC's 
were hard to find and very expensive, I built a pretty good DAC using a 
stable voltage source and PWM-ing that. After good filtering of the output 
signal I got 15 bit accuracy. Accuracy was mainly determined by the drift 
of the opamps, but todays state-of-the-art opamps should have much less 
drift than those of 20 years ago.

Or : try something with one of these 24 bit audio DAC's such as AD1955, 
they are cheap (< 10 Euro) and if you ask very nice they even send samples.

73, Rik


At 18:57 12/05/2004 +0200, you wrote:
>I wonder if someone on this list has already had a problem similar
>to this, and has found a simple solution.
>
>Problem:
>I have a super-filtered and stable DC voltage of about 8V.
>I need to make it variable (with negligible current requirements)
>from abt 0 to abt 8V, with a minimum step size of abt 10 uV.
>I can use part of the processing power of an AVR microcontroller
>for this task.  One solution that comes immediately to the mind is
>to use a 20-bit DAC, fed somehow with that stable voltage, but I am
>in search of something that doesn't need hard-to-find components.
>
>I played with the idea of a charge pump, where you control with
>a couple of output bits of a port of the micro the charge and discharge
>of a capacitor for variable lenghts of time, so to achieve the wanted
>result, but I am unsure of the implementation details. The response time
>requirements are very relaxed, there will be an adjustment of that
>voltage let's say once every two minutes or more.
>
>Has anybody suggestions or links ?  Many thanks
>
>73  Alberto  I2PHD
>
>
>