temperature compensation.

[Rene Schmitz]

At 10:29 27.07.99 GMT, Paul Maddox wrote:
>Dear all,
>
>  Ive been toying with tempature compensation of VCO's.. for most circuits 
>we need a NTC 3300 PPM resistor..
>
>  If you look at the ASM one we have this
>
>  OPAMP> -----VVVV----+------ base of transistor
>                      >
>                      < NTC resistor
>                      >
>                      |
>                      0v
>
>  Why couldnt you simply swap the two around and use a PTC, these seem to be 
>more common, ie....
>
>  OPAMP> -----VVVV-----+---- Base of transistor
>             PTC res   |
>                       >
>                       <  Fixed value res
>                       >
>                       |
>                       0v
>
>  Am I missing something obvious as to why this couldnt be done?
>
>  Paul

Hi Paul and List

Above you swapped NTC for PTC. The tempco resistor in the ASM is a PTC. 
Now, I've sucessfully done what you propose here. I found that all the
cheap NTC resistors have an exponential temperature coefficient of -4%/K.
This value depends 
on the band gap of the used semiconductor material, which seems to be the
same for all the cheap NTCs. Si (?). When you replace the 54k/1k(tempco)
divider with a 51k+4.7k(NTC)/1k divider one gets pretty good temperature
cancellation. 
(Theres a rule of thumb for the series resistor: 11 times the value of the
NTC because 4%/12=0.33%)
Of course the compensation is not exact at all temperatures, because of the
exponential temperature dependance. But I needn't. You can even find an
example, on my website. I've compensated one of my VCO 2s with  this
divider scheme. Works nice. I did also simulations which showed that good
stability can be archeived arround room temperature.


Bye
 René (aka Tempco Geek)

             ,   : (uzs159 at uni-bonn.de)
|") [" |\ | ["   : (http://www.uni-bonn.de/~uzs159)
|"\ [_ | \| [_   : (http://members.xoom.com/Rene_Schmitz)