LM311 V-F

[AJ]

HI,

I am currently using a LM311 as a V-F. I have it set for 150mV/Hz with the
following components.

Rt = 200K
Ct = 100nF
Rs = 13.8K
Cl = 1uF
Rl = 100K

For testing I have used some components that I have had laying around and I
am finding that the output is greatly effected by temperature. The
resistors are surface mount metal film but I have used an XR7 for Ct and a
Tantalum for Cl. The datasheet recommends NP0 or C0G series for Ct and I
have found one with 30ppm/C in an SMT package ok but it also says I should
use a Mylar Cap for Cl. I have had a bit of a look around and the only 1uF
Mylar caps I can find are huge and not available in SMT. Can anyone
recommend a good choice for Cl? I was also wondering if Cl had to be 1uF, I
am guessing it does but the datasheet isn't very clear, maybe I could use a
smaller value?

Best regards.


AJ

 

[Eeyore]

HI,

I am currently using a LM311 as a V-F. I have it set for 150mV/Hz with the
following components.

Rt = 200K
Ct = 100nF
Rs = 13.8K
Cl = 1uF
Rl = 100K

For testing I have used some components that I have had laying around and I
am finding that the output is greatly effected by temperature. The
resistors are surface mount metal film but I have used an XR7 for Ct and a
Tantalum for Cl. The datasheet recommends NP0 or C0G series for Ct and I
have found one with 30ppm/C in an SMT package ok but it also says I should
use a Mylar Cap for Cl. I have had a bit of a look around and the only 1uF
Mylar caps I can find are huge and not available in SMT. Can anyone
recommend a good choice for Cl? I was also wondering if Cl had to be 1uF, I
am guessing it does but the datasheet isn't very clear, maybe I could use a
smaller value?

1uF plastic film caps are quite large but mylar typically aren't the smallest.
Mylar is just a DuPont trade name. Look for generic polyester film instead. A
leaded cap is much more likely to be available at this value btw.

Graham

 

[Ancient_Hacker]

AJ wrote:


The rate is determined by the ratio of the two capacitors, so they
GOTTA be very stable.


The actual capacitor values are not critical, it's only their ratio
that is important.

Just dont go much below say 500pf as then stray capacitances may become
significant.


So you could scale down the values by a factor of 100 easily. Maybe
even 1000 if there isnt much stray capacitance.

 

[Eeyore]

AJ wrote:

The rate is determined by the ratio of the two capacitors, so they
GOTTA be very stable.

In which case make them *both* polyester film as the tempco will at least track.

Graham

 

[Eeyore]

Mylar SMT capacitors are not available, because polyester cannot
survive the high temperatures required for the SMD process.

Is that what it is ? Id never realyl thought much about it.

Here's one series that goes up to 1uF in SMT, with somewhat different
characteristics because of the different dielectric:

http://www.panasonic.com/industrial/components/pdf/ef038_ecpu_dne.pdf

There a very high dissipation factor with that type.

Graham

 

[Spehro Pefhany]

HI,

I am currently using a LM311 as a V-F. I have it set for 150mV/Hz with the
following components.

Rt = 200K
Ct = 100nF
Rs = 13.8K
Cl = 1uF
Rl = 100K

For testing I have used some components that I have had laying around and I
am finding that the output is greatly effected by temperature. The
resistors are surface mount metal film but I have used an XR7 for Ct and a
Tantalum for Cl. The datasheet recommends NP0 or C0G series for Ct and I
have found one with 30ppm/C in an SMT package ok but it also says I should
use a Mylar Cap for Cl. I have had a bit of a look around and the only 1uF
Mylar caps I can find are huge and not available in SMT. Can anyone
recommend a good choice for Cl? I was also wondering if Cl had to be 1uF, I
am guessing it does but the datasheet isn't very clear, maybe I could use a
smaller value?

Best regards.


AJ

Mylar SMT capacitors are not available, because polyester cannot
survive the high temperatures required for the SMD process.

Here's one series that goes up to 1uF in SMT, with somewhat different
characteristics because of the different dielectric:

http://www.panasonic.com/industrial/components/pdf/ef038_ecpu_dne.pdf

And some nice 5mm lead pitch through-hole stacked-film caps (molded
case and dipped, respectively):

http://www.epcos.com/inf/20/20/db/fc_05/MKT_B32520_29.pdf
http://www.panasonic.com/industrial/components/pdf/abd0000ce8.pdf


Best regards,
Spehro Pefhany

 

[Eeyore]

1uF plastic film caps are quite large but mylar typically aren't the smallest.
Mylar is just a DuPont trade name. Look for generic polyester film instead. A
leaded cap is much more likely to be available at this value btw.

See....
http://uk.farnell.com/jsp/search2/b...&Ntt=1uf+polyester+63v&Ntx=&_requestid=355656

Graham

 

[Fred Bartoli]

Eeyore a écrit :

In which case make them *both* polyester film as the tempco will at least track.

It'll be better, but saying they'll track is overstating.
Caps' tempco depends on dielectric but also on the cap construction.
For ex. try a stacked vs wound cap, with the same dielectric.

 

[Eeyore]

Eeyore a écrit :

It'll be better, but saying they'll track is overstating.
Caps' tempco depends on dielectric but also on the cap construction.
For ex. try a stacked vs wound cap, with the same dielectric.

I had a couple of 'box polys' in mind.

Graham

 

[Rich Grise]

AJ wrote:

The rate is determined by the ratio of the two capacitors, so they
GOTTA be very stable.

The actual capacitor values are not critical, it's only their ratio
that is important.

Just dont go much below say 500pf as then stray capacitances may become
significant.

So you could scale down the values by a factor of 100 easily. Maybe
even 1000 if there isnt much stray capacitance.

If you're going that small, wouldn't that put it into a range where you
could use NP0s or C0Gs for both? Would they track better?

Thanks,
Rich

 

[AJ]

Is that what it is ? Id never realyl thought much about it.



There a very high dissipation factor with that type.

Graham

Thanks to all for the replies. I used some Polyester Caps and it fixed the
temperature stability problem but im still finding that the output is not
linear to the input for some reason. Rather than spending a lot of time
with it I have decided to use a PIC12F510 instead. Its cheaper, comes in a
surface mount package, requires few external components and can do
everything the LM331 was meant to do for me with enough accuracy for my
application.

Best regards


AJ