Calibrating a temp sensor with a pot rheostat
WJG
Kelvin.
The voltage supply is 12V, through a resisotr of 4.5-9K, to the sensor with
a 10K rheostat on it, and then to ground/negative terminal.
My question is about the calibration. The expected output of the sensor
would be 2.93V give or take the uncalibrated error. What if I need to add
voltage to the sensor. I am looking at it - that the rheostat would only
be dropping voltage. But - should I be looking at it this way, the output
of the sensor would be considered a voltage divider? So essentially, I
have two voltage dividers in play, the obvious one is the rheostat, which
in turn effects the output of the sensor whose voltage is also effected by
an implicit voltage divider? (between the the drop across the resistor and
the sensor)
And how does one calculate the current through the rheostat in this
situation, to determine what is left for the sensor?
Thanks.
Dan Akers
Re;
The LM135 is essentially a device that operates in-circuit much like a
zener diode (a parallel voltage regulator); with the exception that the
zener voltage varies with temperature. The "breakdown" voltage is
proportional to the ABSOLUTE temperature of the device with a temp
constant of +10 mV/K. The uncalibrated "zener" voltage at 25C is 2.98V
+/-0.03V; that's +/-3C . The device operates with an "idle" current of
400uA to 5mA, but keep the operating current at a minimum for your
application to minimize self-heating. By adding a 10K pot across the
device with the wiper attached to the "adjustment" terminal, the above
accuracy can be improved by calibrating the device in-situ.
Thus, the 10K pot acting as a variable potential for the device adj.
input (practically zero current drain via the adj. terminal) is used
merely to more precisely calibrate the device and does not contribute
significantly to the overall "zener" divider scheme albeit it does
slightly reduce the available output current for a given series
resistor. The proportional output is still the voltage drop across "+"
and "-" terminals of the device (which also happen to be the legs of the
10k pot).
Example; at 25C with a 6.8K series resistor and 12V supply the circuit
current is (12-2.98)/6800=1.32mA. The pot would be carrying about
2.98/10k=0.298mA of that. The LM135 would be conducting the rest;
namely 1.022mA and dissipating 3.05mW (no-load conditions). The
limiting "load" you can put on the "output" of the circuit is then:
2.98/0.001022=~2900ohms. (Note that if you hadn't added the cal. pot.
the full 1.32mA would have been available for output; implying a
limiting output load of about 2300 ohms.)
I hope that helps...
-Dan Akers
Dan Akers
<snip>
"What if I need to add voltage to the sensor." <snip>
______________________________________
Re;
The LM135 series is essentially a parallel voltage regulator whose
regulated output is proportional to temperature. If you need greater
output voltage you'll need to amplify the output with an OP amp (or the
like); presumably set at some fixed gain. In other words I don't think
you can boost the "output" by very much with the 10k pot and the adjust
terminal; I'm guessing maybe 30mV or so...?
-Dan Akers
wg
THANK YOU.
My question about adding voltage is about the calibration. As I understand
it, the trim pots will cause voltage drops. But in the process of
calibration, I find that I need to ad an extra 30mV to the output. And it
seem that the trim pot would not help.
But I am a real amateur - so I think I am missing somethign because it
should be that difficult that if the factory produced LM135 might need an
op amp for correct calibration.
It seems to me, that the voltage from the supply is not being dropped
completely across the resistor because it still have to drop
proportionately arcoss the resistance contained in the LM135/rheostat, and
enough that gross changes could be made with a rheostat.
dig...@webtv.net (Dan Akers) wrote in news:24484-3E14964E-106@storefull-
2173.public.lawson.webtv.net:
--
Shop at http://www.handsonoptics.com
Peter Bennett
>I have not read your first post carefully yet, I have printed it out.
>THANK YOU.
>
>My question about adding voltage is about the calibration. As I understand
>it, the trim pots will cause voltage drops. But in the process of
>calibration, I find that I need to ad an extra 30mV to the output. And it
>seem that the trim pot would not help.
>
The adjustment pot on the LM335 messes with the internal workings of
the LM335, so can probably increase _or_ decrease the voltage across
the part.
--
Peter Bennett VE7CEI
GPS and NMEA info and programs: http://vancouver-webpages.com/peter/index.html
Newsgroup new user info: http://vancouver-webpages.com/nnq
Dan Akers
"My question about adding voltage is about the calibration. As I
understand it, the trim pots will cause voltage drops. But in the
process of calibration, I find that I need to ad an extra 30mV to the
output. And it seem that the trim pot would not help.
But I am a real amateur - so I think I am missing somethign because it
should be that difficult that if the factory produced LM135 might need
an op amp for correct calibration."
Re;
Like I said, the LM135 is essentially a shunt (parallel to the output)
voltage regulator whose voltage drop is proportional to it's absolute
temperature. Designing a circuit that uses it is essentially the same
as designing a simple circuit that utilizes a zener diode as a shunt
regulator with the exception that the power dissipation in the LM135
needs to be kept at a minimum to minimize self-heating. The voltage
drop across the LM135 (and the pot) is determined by the temp of the
device and to some extent by the pot wiper voltage at the adjust
terminal of the LM135. My guess is that you may be able to alter the
regulated voltage by about +/- 30mV at 25C. The literature I have
doesn't specify the cal. slope change capabilities of the adj. terminal.
I infer the +/-30mV @25C from the cal/not-cal specs. In any case, the
calibration correction that you make at any given temperature to the
voltage/temp curve, is extrapolated to the range of the device; ie the
device temp/voltage characteristic (10mV/K) is re-scaled from origin 0
deg K.
You don't need the OP amp for correct calibration. I was merely
suggesting that if you want greater voltage output, proportional to
temperature, then you'll need some sort of linear amplifier. Take a
look at the LM3911. This an 8 pin DIP; essentially a LM135 and OP amp
in one package.
______________________________________
WG wrote;
"It seems to me, that the voltage from the supply is not being dropped
completely across the resistor because it still have to drop
proportionately arcoss the resistance contained in the LM135/rheostat,
and enough that gross changes could be made with a rheostat."
______________________________________
Re;
The supply voltage is dropped proportionately across the LM135 with
respect to the device temperature; never completely across the series
resistor. It is a shunt regulator of sorts, in that respect.
Regardless of the series resistor that you use, within limits of shunt
regulator design concepts of course, the voltage drop across the device
is ultimately determined by the device temperature along with the
characteristic sloping factor that can be slightly altered at the adj.
terminal of the device.
What are your exact requirements for a temp/voltage characteristic?
Current? Temp. range? What exactly are you trying to do???
-Dan Akers
wg
hobby of mine. If I can create the circuit, I save about $150 from the
cost a similar commercial product for about $20 in parts. But - I just
don't want to build it - I want to also understand the circuit and what is
happening.
dig...@webtv.net (Dan Akers) wrote in news:12797-3E15D85B-679@storefull-
2171.public.lawson.webtv.net:
--
Shop at http://www.handsonoptics.com
Dan Akers
a hobby of mine. If I can create the circuit, I save about $150 from the
cost a similar commercial product for about $20 in parts. But - I just
don't want to build it - I want to also understand the circuit and what
is happening."
Re;
Well I can appreciate that, but your still being a bit cryptic about
what you want to build; a thermometer? A thermostat? A differential
temperature indicator? Required temperature indicating range? What are
you going to drive with it? A meter? etc, etc... Have you taken a look
at National's website? I think you can get application notes there that
might help. I've looked at a 1995 and 1982 copy of their application
notes and sad to say, not much detail on the capabilities of the adj.
terminal and I don't have one handy to experiment with right at this
moment :-(. How about a Google search on "LM335"?
Simply said, the adj. terminal senses the ratio of the voltage at the
adj. terminal to the device voltage drop to determine/alter the
calibration slope constant. Without the external pot, this value
defaults to 10mV/K via internal integrated resistors. I don't know how
I can explain it more plainly... You do realize that in order to
accomplish the linear voltage/temp. characteristic from it's internal
RTD (resistance temp. detector) in a two terminal, "zener" (shunt)
fashion, this IC has some 16 internal transistors contained within the
little TO-92, SO-8, or TO-46 package? I mean; what is that you seek to
understand; the device, or how to use it? Sometimes we have to accept
integrated circuits for the black-boxes that they are... :-)
-Dan Akers
wg
degrees above the current ambient temperature (specifically, a telescope
objective lens, to prevent dewing, and to not keep the heater on to long
and create tube air currents that would degrade viewing through the
telescope).
So before I build it - I want to understand how the components will impact
the circuit. For instance - the LM135 - if I needed to increave the output
for calibration, its not obvious that a trim pot would do that. My
background is chemical engineering, so I do have an interest, and the
ability to learn how basic circuits are working - although I am global
learner.
This was the last thing that puzzle me about the circuit. The only other
item was someone's suggection of putting a capacitor in series with the
feedback resistor on the comparator (I have a schmitt trigger/hysteresis)
around it.
dig...@webtv.net (Dan Akers) wrote in news:23461-3E1730D2-889@storefull-
2172.public.lawson.webtv.net:
--
Shop at http://www.handsonoptics.com
Dan Akers
"I want to build a circuit that controls a heater, to heat an object to
2-3 degrees above the current ambient temperature (specifically, a
telescope objective lens, to prevent dewing, and to not keep the heater
on to long and create tube air currents that would degrade viewing
through the telescope)."
Re;
Wondering; are you trying to prevent condensation on the inside of the
lens or on the outside? Either situation is possible and if inside
condensation is the problem, then 2-3 C will probably not do the trick;
especially if the telescope is being hauled from indoors (where the
internal telescope dew point may have come to an equilibrium 10C) to
outdoors were the prevailing ambient temp. may be well below 0C.
In either case, how do you intend to apply the heat? Are you using two
LM135's to accomplish the 2-3C differential regulation via the
comparator? One for ambient sense and one for "lens" sense?
-Dan Akers
Dan Akers
"So before I build it - I want to understand how the components will
impact the circuit. For instance - the LM135 - if I needed to increave
the output for calibration, its not obvious that a trim pot would do
that."
Re;
If you don't find out before hand, I'll let you know just how that adj.
terminal functions in a day or so, as I am expecting some LM335's in the
TO-92 package to arrive at any time.
-Dan Akers
wg
telescope moving it from inside to outside. The big problem would be
internal air currents that would degrade the telescope's performance.
The circuit I have in mind is for a telescope already outside and in
equilibrium where the telescope is usually radiating heat and having a
temperature below that of ambient. And keeping the objective lens above
ambient would prevent the dew/frost from forming.
dig...@webtv.net (Dan Akers) wrote in news:23937-3E19DB9A-1162@storefull-
2174.public.lawson.webtv.net:
--
Shop at http://www.handsonoptics.com
wg
anything about the terminal adjustment at all. I plan on writing a letter
or calling National to see what is up. I am sure there is something in the
LM135 that allows this, or the LM135 could simply all pass inspection at
National providing if they are out of calibration, they giving too high of
a voltage and the use must reduce the voltage no matter what.
dig...@webtv.net (Dan Akers) wrote in news:23938-3E19DF0B-550@storefull-
2174.public.lawson.webtv.net:
--
Shop at http://www.handsonoptics.com
John Popelish
>
> I want to build a circuit that controls a heater, to heat an object to 2-3
> degrees above the current ambient temperature (specifically, a telescope
> objective lens, to prevent dewing, and to not keep the heater on to long
> and create tube air currents that would degrade viewing through the
> telescope).
If the two temperatures are accurately sensed, you don't have to worry
about how long the heat is on, because the control circuit will pulse
it on and off as required to maintain the temperature differential.
> So before I build it - I want to understand how the components will impact
> the circuit. For instance - the LM135 - if I needed to increave the output
> for calibration, its not obvious that a trim pot would do that. My
> background is chemical engineering, so I do have an interest, and the
> ability to learn how basic circuits are working - although I am global
> learner.
The pot steals a bit of current from the supply resistor based on
ohm's law (I=E/R) where E is the output voltage of the LM335 at what
ever temperature in is immersed in, and R is the pot total
resistance. For example, if you use the recommended 10k pot and the
output voltage is about 2.4 volts, then the pot will steal
2.4/10,000=.24 milliamp from the supply resistor, so you have to lower
that resistor enough to supply both the 1 milliamp or so needed to
make the 335 operate, and also .24 milliamp for the pot. The tap on
the pot (sliding contact) produces a voltage that is some fraction of
the output voltage. Varying this fraction trims the millivolts per
degree over a narrow range. I think you will have no trouble
producing a 2 or 3 degree miscalibration, so that the two temperatures
have to differ by 2 or 3 degrees to make the the output voltages match
and put the control circuit in the middle of its on off range.
> This was the last thing that puzzle me about the circuit. The only other
> item was someone's suggection of putting a capacitor in series with the
> feedback resistor on the comparator (I have a schmitt trigger/hysteresis)
> around it.
The capacitor in series with the feedback limits how long a power
pulse can be if the temperatures differ by exactly the desired
amount. Without that cap, the actual temperature will swing more and
less than the desired offset (because the on-off state of the heater
alters the switching temperature offset), with much longer power
pulses and much longer off time between them.
The series capacitor value allows you to choose the maximum on and off
time so that a rapid stream of narrow pulses holds the offset
essentially constant if the maximum on and off times are much shorter
than the thermal time constant of the heater and scope. A few seconds
max. on and off time should be fast enough to produce no measurable
temperature ripple.
--
John Popelish
Dan Akers
"There usually isn't a problem of condensation on the inside of the
telescope moving it from inside to outside. The big problem would be
internal air currents that would degrade the telescope's performance.
The circuit I have in mind is for a telescope already outside and in
equilibrium where the telescope is usually radiating heat and having a
temperature below that of ambient. And keeping the objective lens above
ambient would prevent the dew/frost from forming."
Re;
That makes sense; now I fully understand. Anyway, I got hold of some
LM335's. I powered one up with a 12V supply using a 10k series
resistor. Without anything connected to the adj. term the output
(voltage drop across the LM335 was 2.940V) was right on the money for
the room ambient; namely 21C. I then hooked up a 10-turn, 10k pot
across the device with the wiper attached to the adj. term of the LM335
and was able to vary the device voltage drop from 2.498V to 3.610 at the
extremes of the pot. So it looks like you can adjust the calibration
slope constant from about 8.5mV/K to about 12.3mV/K. I hope that helps
you out...
-Dan Akers