<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<title></title>
</head>
<body text="#000000" bgcolor="#ffffff">
Are you sure that's true for the CEM3374, Paul? I know it's true for
the 3340. The 3374 datasheet says "...the 3374 includes an on-chip
temperature sensor which generates an output voltage, nominally +2.5v,
proportional to chip temperature (i.e. with a TC of +3300 ppm); by
ensuring that the exponential control voltage is derived from this
sensor output (for instance, by applying it to the reference input of
the system DAC), unsurpassed oscillator stability can be acheived).<br>
The sensor voltage mentioned most certainly will be a band gap
generated voltage. This can be scaled internally to +3300ppm by a
simple linear amplifier and does not require a multiplier. The
multiplication is done by applying it to the reference voltage of a
multiplying DAC whose digital inputs are the control voltage. Once
upon a time I looked into using AD633's to act as the external
multipliers, but ended up using a couple of tempco resistors instead
(<a class="moz-txt-link-freetext" href="http://home.comcast.net/~sbernardi/elec/og2/og2_cem3374.html">http://home.comcast.net/~sbernardi/elec/og2/og2_cem3374.html</a>). I
think Jim's idea of using the LM13700 multiplier like in his temp
compensated VCO would work fine. You'd need two of them.<br>
The 3340 on the other hand, has an onboard analog multiplier which
multiplies the internal PTAT voltage times the input control voltage. <br>
<br>
<br>
Paul Schreiber wrote:<br>
<blockquote type="cite"
cite="mid000d01c3945a$82460420$0201a8c0@speakeasy.net">
<pre wrap="">Correct, it uses an on-chip multiplier.
Paul S.
</pre>
</blockquote>
<blockquote>Karl Dalen wrote:<br>
<br>
It also say that a ±90mv covers a 10 octave range
<br>
wouldent that put quite some precision requirements
<br>
on the OP amps in the DAC reference feedback path
<br>
in terms of V offset drift with TC?!
<br>
</blockquote>
The DAC should run at the full scale voltages (1v/octave). Scale it
down after the DAC with an inverter, 100K input R and a say a 1.5K +
500 trimmer (for scale trim) in the feedback path. . This is no
different than what other VCO designs have.<br>
<br>
<br>
<blockquote>
<pre wrap="">Its no doubt that it works splendid but how sure
could one be that the temp drift are equal on both
VCO's?
</pre>
</blockquote>
There should be pretty good matching between the two VCO's with their
exponential transistors and the bandgap generator transistors close to
each other on the same chip. <br>
<pre wrap="">
</pre>
<pre class="moz-signature" cols="72">--
Scott Bernardi
<a class="moz-txt-link-abbreviated" href="mailto:sbernardi@comcast.net">sbernardi@comcast.net</a></pre>
</body>
</html>