[sdiy] Tempco adjuster idea

[Ian Fritz]

Hi Rene --

>why not simply use the TCs in a divider and add a NTC over the upper leg. 
>I guess the NTC would be very large and parallel to the upper leg 
>resistor. So then the nonlinearity of the NTC shouldn't make a big 
>problem. Also the thermal coupling wouldn't have to be as tight for the 
>NTC as its influence is relatively small, so the errors that thermal 
>gradients could generate are even smaller.

Interesting idea.  I've always kept away from dividers because they distort 
the T dependence.  But maybe I should look at the actual response more 
closely.  Carbon film resistors have a negative tempco, so maybe common 
components on hand could be used.

OTOH, my new proposal has a "knob" to dial in either a positive or negative 
correction.

>Personally I wouldn't bother, but simply use the 3200ppm tempcos as they 
>are. I have oscillators compensated with 3000ppm which work just fine in 
>practice.

According to my calculations, 3000 ppm/K tempco coefficient corresponds to 
-247 ppm/K in scale-factor drift.  And the 3260 ppm/K units I just bought 
should only give -66 ppm/K scale drift.   So you are correct, these are 
pretty good.

OTOH Jim's active compensation results are incredibly good.  There is some 
scatter in his data, but if you take the end points you get 10 ppm/K scale 
drift.  If you use a least squares fit you get 40 ppm/K.

So I'm looking into how to get into that rarified atmosphere with passive 
compensation.

Best regards,

   Ian



>Cheers,
>  René
>
>Ian Fritz wrote:
>>Hi all --
>>Has anyone thought of making a circuit to compensate for tempco resistors 
>>that have incorrect coefficients? I've come up with an idea for this 
>>which is really quite simple, but I don't remember seeing it before.
>>The ideal tempco resistor (for compensating exponential current 
>>generators) has what's called a PTAT (proportional to absolute 
>>temperature) response. This means that the resistance is of the form R = 
>>AT, where A is a constant and T is absolute temperature.  This dependence 
>>cancels the 1/T factor in the exponential of the transistor I-V response 
>>function. The corresponding temperature coefficient (1/R)(dR/dT) is just 
>>1/T, or 3350ppm/K at room temperature.
>>So what if you buy some tempcos and their coefficients are different from 
>>this?  Since they are normally made of metals, their resistance will 
>>still have a linear T dependence, but they will not be PTAT.  In other 
>>words, their resistance can be well approximated as R = AT + B, where the 
>>constant B is the resistance extrapolated back to T = 0.  If B is 
>>positive then the coefficient is too low, and conversely if B is negative 
>>then it is too high.
>>So to correct for an incorrect coefficient one just needs to make a 
>>circuit that cancels the B term.  This looks easy to do, at least on 
>>paper, and the only drawback is that it takes several amplifiers instead 
>>of just one to condition the control voltage inputs.
>>The circuit I designed has four opamps.  The first one is a unity-gain 
>>inverting summer for the control signals.  The second and third op amps 
>>are driven in parallel from the first.  The second op amp is an inverter 
>>with with a gain of 0.1 and with the tempco resistor as the feedback 
>>resistor. (E.g., 10k input resistor, 1k tempco feedback resistor). The 
>>third opamp is a "switch-hitter" with gain adjustable over the range of 
>>-0.1 to 0.1.  This produces the signal needed to buck out the B 
>>coefficient.  The fourth op amp sums the outputs of the second and third, 
>>with the bucking signal attenuated to about 20% of the compensated 
>>signal.  A final 5.5 to 1 voltage divider feeds the compensated 18 mV/Oct 
>>signal to the base of the converter transistor.
>>The circuit should work with just about any tempco resistance value, as 
>>long as the second op amp's input resistor is chosen to be ten times as 
>>large.  The only potential problem I can think of with this idea is that 
>>noise and offsets could add up with four amplifiers in the circuit.
>>Keeping impedances low and using high-quality modern op amps should help 
>>alleviate these problems.
>>I hope to try this idea out in the near future so that I can use the
>>3200ppm/K units that KRL sold me.
>>Any comments most welcome.
>>   Ian
>
>--
>uzs159 at uni-bonn.de
>http://www.uni-bonn.de/~uzs159
>
>