# favorite temp sensors

J

#### Jon Kirwan

Jan 1, 1970
0
Not in my experience. (but my experience is fiarly limited... a few
transitors tested.)

I always got a number that was a bit off ~0.3%, so about 1 degree at
room temp. I always assumed the error was due to the transistor
beta... Since the current is Ic and Ib. (I think I got a temperature
that was always a bit high, but I'd have to check my notebook.) You
could add some beta 'fudge factor'.... but then beta changes with
temperature too.

It also depended a bit on the collector current. (1 uA to 10uA were
'nice' currents)

Hi, George. I posted up a link to Linear's AN45 elsewhere
under this topic. See page 7 there. But I take your
not at the 'charged gas' theory level but at the higher (and
more usual for an EE) device modeling level.

The two-current pulse method, using say 1X and 10X currents,
depends upon the following:

dV = (k/q) * ln( 1+Ic/Is ) dT

Although k and q are known, the entire factor that includes
the ln( 1+Ic/Is ) part isn't knowable in advance. But if one
assumes that the +1 term is negligible then the two pulsed
currents results in:

dV1/dT = (k/q) * ln( Ic1 ) - (k/q) * ln( Is )
dV2/dT = (k/q) * ln( Ic2 ) - (k/q) * ln( Is )

Subtracting dV1/dT from dV2/dT yields:

dV1 - dV1 = (k/q) * ln( Ic2/Ic1 ) ) dT

And if the ratio of Ic2/Ic1 is known a priori then the entire
factor, (k/q) * ln( Ic2/Ic1 ) ), is also known. And as a
consequence could be used to measure temperature without
having to calibrate the system. Or so it seems at first
blush.

But it's also the case that the value of the saturation
current, Is, is itself a rather complex function of T:

Is(T) = Is(Tn) * (T/Tn)^3 * e^( -(q*Eg/k) * (1/T-1/Tn) )

Where Tn is some chosen T(nominal).

In fact, this particular component is what overwhelms the
first equation (which is positive vs temperature) and yields
the usually quoted -2mV/K figure (very approximately.) So, in
fact, Is(T) is the dominant factor in Vbe change over T and
in no possible way is it a simple function of T!

(Even the above Is(T) equation itself is a simplification.
The power ((T/Tn)^3) for example is an approximation and not
strictly true in practice. Same with Eg, which itself is also
taken as a single approximation value.)

Just as a guess, the idea of ln( Is ) being cancelled
entirely out of the equation by ratiometry, even assuming
that the die is at thermal equilibrium, would make me worry a
little. (I accept that pulsing the 1X/10X current change fast
enough or that using low enough currents, like the 1uA and
10uA you mention, would yield a near-equilibrium state.) I'm
just not sure that at this level of modeling, that _Is_
remains dead stable as a modeling parameter when facing a 10X
current change. There is a lot of linearity over orders of
magnitude change, as a broad statement. But exactly how
linear is it when provided a two point ratio a decade apart,
vs device variation?

I wonder that some temperature error is swept under this
Is(T) rug and hidden from the analysis, so to speak. Even
assuming thermal equilibrium. Because it may really be that
Is(V,T), not Is(T), as both the power (^3) and Eg are taken
as simple constants for simplification when they aren't, in
fact, invariant at this level of modeling.

You mention base currents as a possible error. I've ignored
that so far. The equation:

dV = (k/q) * ln( 1+Ic/Is ) dT

in the diode connected case refers to Ic. The currents
through it, on whole, are (beta+1)/beta times as much. If you
cobble up precision current sources at exactly 1X and 10X,
the ratio of Ic2/Ic2 would still be 10, even though you are
driving Ie, I think. However, beta itself changes vs Ic. So
there is that to account for, if you were only using a beta
level model. But the:

dV1 - dV1 = (k/q) * ln( Ic2/Ic1 ) ) dT

method doesn't use or rely upon beta. So I'm not imagining a
problem there because (1) the ratio is still 10X and (2) beta
isn't used in the analysis method.

Interesting problem getting past a certain level of accuracy,
though. There must be several papers that go beyond the AN45
app note I'd posted up earlier. I haven't seen one, yet.

Jon

J

#### John S

Jan 1, 1970
0
Cool, if you can't afford a FLIR.

Did you see my post that I bought the one you pointed out? It is due
here this afternoon sometime.

John S

B

#### Bill Sloman

Jan 1, 1970
0
Better way posted. Your way is asinine.

But accurate, and precise. Your idea of "better" needs work.

B

#### Bill Sloman

Jan 1, 1970
0
So midst the noise I've seen the following mentioned:

LM71
DS18B20
LM75

Of those, only a few seem to come in non-surface mount packages
and I'm not sure how we'd couple surface mount devices to the
copper cooling lines.

Thermally conductive adhesive? The AD7414 data sheet mentions this on part 14.

Given that we just want to measure the coolant temperature on
both sides of the liquid-cooled heatsink, and not reinvent
anything [1]; it looks like the DS18B20 may do the trick.

It's 2-4X the price of LM35's but still affordable.

If the copper inlet and outlet pipes run close together - or can be made torun close together, you might mount two AD7414 parts on a flexible printedcircuit, and glue the back of one to the inlet pipe and the back of teh other to the outlet pipe. You'd separate the two AD7414 parts on the flexy bymore than the worst-case separation of the two bits of pipe to which you were going to glue each one (with thermally conductive epoxy) so a to leave a bow in the flexy to accommodate any movement between the two points.

And you'd have to do a zero-point calibration when the heat-sink wasn't dissipating any power. The temperature error data on page 15 of the data sheetis a bit depressing. Interchangable thermistors can be quite a bit more accurate - put two (one on each pipe) in a bridge, and AC-drive with a bifilar wound centre-tapped transformer to set up a Blumlein bridge.

http://www.diybanter.com/metalworking/275783-precision-electronic-levels-blumlein-bridge.html

S

#### Spehro Pefhany

Jan 1, 1970
0
Cool, if you can't afford a FLIR.

Can your FLIR output an array of the temperatures for further
analysis?

Best regards,
Spehro Pefhany

B

#### Bill Sloman

Jan 1, 1970
0
Whenever is a thermistor "precise" ??

Whenever you pay for a "interchangeable glass-encapsulated thermistor", as I recommended earlier in this thread. I don't know when Yellow Springs Instruments first introduced them, but I got told about their parts in 1974, and other people jumped an that particular band-waggon early on. I've bought Betatherm parts from Farnell which are just as tightly specified.

You probably don't know about that kind of stuff - thermistors don't seem to be susceptible to integration onto silicon.

G

#### George Herold

Jan 1, 1970
0
Hi, George. I posted up a link to Linear's AN45 elsewhere
under this topic. See page 7 there. But I take your
not at the 'charged gas' theory level but at the higher (and
more usual for an EE) device modeling level.

The two-current pulse method, using say 1X and 10X currents,
depends upon the following:

dV = (k/q) * ln( 1+Ic/Is ) dT

Although k and q are known, the entire factor that includes
the ln( 1+Ic/Is ) part isn't knowable in advance. But if one
assumes that the +1 term is negligible then the two pulsed
currents results in:

dV1/dT = (k/q) * ln( Ic1 ) - (k/q) * ln( Is )
dV2/dT = (k/q) * ln( Ic2 ) - (k/q) * ln( Is )

Subtracting dV1/dT from dV2/dT yields:

dV1 - dV1 = (k/q) * ln( Ic2/Ic1 ) ) dT

And if the ratio of Ic2/Ic1 is known a priori then the entire
factor, (k/q) * ln( Ic2/Ic1 ) ), is also known. And as a
consequence could be used to measure temperature without
having to calibrate the system. Or so it seems at first
blush.

But it's also the case that the value of the saturation
current, Is, is itself a rather complex function of T:

Is(T) = Is(Tn) * (T/Tn)^3 * e^( -(q*Eg/k) * (1/T-1/Tn) )

Where Tn is some chosen T(nominal).

In fact, this particular component is what overwhelms the
first equation (which is positive vs temperature) and yields
the usually quoted -2mV/K figure (very approximately.) So, in
fact, Is(T) is the dominant factor in Vbe change over T and
in no possible way is it a simple function of T!

(Even the above Is(T) equation itself is a simplification.
The power ((T/Tn)^3) for example is an approximation and not
strictly true in practice. Same with Eg, which itself is also
taken as a single approximation value.)

Just as a guess, the idea of ln( Is ) being cancelled
entirely out of the equation by ratiometry, even assuming
that the die is at thermal equilibrium, would make me worry a
little. (I accept that pulsing the 1X/10X current change fast
enough or that using low enough currents, like the 1uA and
10uA you mention, would yield a near-equilibrium state.) I'm
just not sure that at this level of modeling, that _Is_
remains dead stable as a modeling parameter when facing a 10X
current change. There is a lot of linearity over orders of
magnitude change, as a broad statement. But exactly how
linear is it when provided a two point ratio a decade apart,
vs device variation?

I wonder that some temperature error is swept under this
Is(T) rug and hidden from the analysis, so to speak. Even
assuming thermal equilibrium. Because it may really be that
Is(V,T), not Is(T), as both the power (^3) and Eg are taken
as simple constants for simplification when they aren't, in
fact, invariant at this level of modeling.

You mention base currents as a possible error. I've ignored
that so far. The equation:

dV = (k/q) * ln( 1+Ic/Is ) dT

in the diode connected case refers to Ic. The currents
through it, on whole, are (beta+1)/beta times as much. If you
cobble up precision current sources at exactly 1X and 10X,
the ratio of Ic2/Ic2 would still be 10, even though you are
driving Ie, I think. However, beta itself changes vs Ic. So
there is that to account for, if you were only using a beta
level model. But the:

dV1 - dV1 = (k/q) * ln( Ic2/Ic1 ) ) dT

method doesn't use or rely upon beta. So I'm not imagining a
problem there because (1) the ratio is still 10X and (2) beta
isn't used in the analysis method.

Interesting problem getting past a certain level of accuracy,
though. There must be several papers that go beyond the AN45
app note I'd posted up earlier. I haven't seen one, yet.

Jon- Hide quoted text -

- Show quoted text -

Jon, I'm going to have to reread some of that.
But I've got some data for a TIP31C. (Shipped out a month ago.)
The transistor is at the end of a ~12" cable and 18" SS probe.
The to220 tab is stuck to a hunk of copper at the bottom of the
probe.
Here's the forward voltage vs current.
I list the temperature as 297.75K
There's a FET opamp with a DC offset of -0.30mV

I(uA) V(mV) delta
0.01 243.365 63.42
0.1 306.785 60.82
1.0 367.606 59.91
10 427.519 59.50
100 487.020 59.57
1k 546.590 61.40
10k 607.986

The delta's are that current minus the one below.
The measurement were just taken in room air.. no control,
and the temperature drifted up (voltage down about 20uV) during the
measurment.

I get a T of 299.6K at the lowest voltage difference.

Oh.. it's a 1/2 kelvin connection... there's just one ground lead and
two leads for current and voltage to the emitter.

I think I did better with a 2N3904.. except it had more resistance.

George H.

G

#### George Herold

Jan 1, 1970
0
Cool, if you can't afford a FLIR.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot comhttp://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation- Hide quoted text -

- Show quoted text -

How much is an FLIR?

George H.

G

#### George Herold

Jan 1, 1970
0
What do you have to pay for 0.1% thermistor interchangeability?

I think we buy 0.2 degree C thermistors from YSI for a few bucks
each....
(no more than $5) George H. G #### George Herold Jan 1, 1970 0 IC temperature sensors are all crap. Every single one, AFAICT--none that I know of claims accuracy better than 1 degree C, even the trimmed ones. Why would that be, if it's so fundamental? Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot nethttp://electrooptical.net- Hide quoted text - - Show quoted text - Still, A sensor that's got 'built in' calibration to ~1% is kinda nice. (and it costs a dime or two.) I've got data for 77K.. I think I had a ~+2 degree error there too. (I guess it can't be a beta error, cause that would be in the other direction.(?) It's the dreaded 'non-ideality factor', which last time we checked had something to do with recombination in the base.. IIRC) George H. G #### George Herold Jan 1, 1970 0 I think we buy 0.2 degree C thermistors from YSI for a few bucks each.... (no more than$5)

0.2 C accurate at one temperature, or over the whole span?

George H.

[snip]

...Jim Thompson
--
| James E.Thompson                                 |    mens     |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon athttp://www.analog-innovations.com|    1962     |

I love to cook with wine.     Sometimes I even put it in the food.- Hide quoted text -

- Show quoted text -

Over some span around room temp. Maybe 0 to 60C.. you can pay more
and get better specs.

It's totally worth it for us. I use them for diode laser thermal
control, diodes are measured in my setup, but will then 'work' in
everyone else's.

(I measure wavelength at one temperature.)

George H.

G

#### George Herold

Jan 1, 1970
0
I don't know of a single commercial unit that uses the ratiometric
current method, they're all PTAT's, some trimmed, some not.

Jim, I'm confused (or just ignorant again).. I thought PTAT's used the
current ratio trick.

George H.

G

#### George Herold

Jan 1, 1970
0
0.2 C accurate at one temperature, or over the whole span?
George H.
BTW... my chopper-stabilized thermistor apparatus made with Germanium
devices... 51 years ago...
http://www.analog-innovations.com/BS_Thesis_JE_Thompson_1962.pdf
...Jim Thompson

...Jim Thompson
--
| James E.Thompson                                 |    mens     |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus   |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon athttp://www.analog-innovations.com|   1962     |
I love to cook with wine.     Sometimes I even put it in the food.-Hide quoted text -
- Show quoted text -

Over some span around room temp.  Maybe 0 to 60C.. you can pay more
and get better specs.

It's totally worth it for us.  I use them for diode laser thermal
control, diodes are measured in my setup, but will then 'work' in
everyone else's.

(I measure wavelength at one temperature.)

George H.- Hide quoted text -

- Show quoted text -

PS
Sorry, just thinking about how important voltage references are to the
whole signal chain.

Geo

S

#### Spehro Pefhany

Jan 1, 1970
0
Whenever you pay for a "interchangeable glass-encapsulated thermistor", as I recommended earlier in this thread. I don't know when Yellow Springs Instruments first introduced them, but I got told about their parts in 1974, and other people jumped an that particular band-waggon early on. I've bought Betatherm parts from Farnell which are just as tightly specified.

You probably don't know about that kind of stuff - thermistors don't seem to be susceptible to integration onto silicon.

Both YSI and Betatherm were acquired by Measurement Specialties about
seven years ago.

Best regards,
Spehro Pefhany

B

#### Bill Sloman

Jan 1, 1970
0
What do you have to pay for 0.1% thermistor interchangeability?

Newark sells a +/0.2C part for $9.57, 1-off Newark Part Number: 10M5351, Fenwall (Honeywel) Part No: 192-303QET-A01 The Betatherm part, also +/0.2C costs$4.79 1-off Newark Part Number: 93K1990
Betatherm part no.:1K2A1B

Thermometrics do a +/-0.1C part for $8.18 Newark Part Number: 65J5900 but it's on special today for$5.43 (1-9)
BTW... my chopper-stabilized thermistor apparatus made with Germanium
devices... 51 years ago...

http://www.analog-innovations.com/BS_Thesis_JE_Thompson_1962.pdf

Terrible circuit diagram, not that American schematics from that period were ever much better. I'd have to redraw if before I could make any sense of it.

S

#### Spehro Pefhany

Jan 1, 1970
0
IC temperature sensors are all crap. Every single one, AFAICT--none
that I know of claims accuracy better than 1 degree C, even the trimmed
ones. Why would that be, if it's so fundamental?

Cheers

Phil Hobbs

That was true a few years ago, but the ADT7320 has quite impressive
specs- +/-0.2°C from -10 to 85°C with a 3V supply and +/-0.25°C from
-20 to 105°C with supply +/- 10%.

Nasty package for field use, of course, and it's going to measure it's
own temperature rise, and not particularly cheap (>\$5 in q100), but it
does include its own ADC. I'd worry about the solder fatiguing if it
saw a lot of cycling.

I've recently designed in a similar TI part with specs about twice as
bad, but still quite good (+-/ 0.5°C, IIRC)and I didn't care very much
about the absolute temperature- it was for compensation, so trimming
was built into the process).

Best regards,
Spehro Pefhany

G

#### George Herold

Jan 1, 1970
0
Different ratio.  PTAT's generate a voltage that indeed is ratio'd by
using two devices of differing areas, thus different current
_densities_.

The Jim Williams' technique, using resistors and adapted by me to use
current sources, uses the same device, "measured" at two carefully
ratio'd currents.

The PTAT effect depends on lots of other variables as well.

I'll dredge thru my files and find one I can show.

...Jim Thompson
--
| James E.Thompson                                 |    mens     |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon athttp://www.analog-innovations.com|    1962     |

I love to cook with wine.     Sometimes I even put it in the food.- Hide quoted text -

- Show quoted text -

Ahh, got it thanks. I must have read the Jim Williams app note.. I
remember the baby bottles.

George H.

D

#### David Lesher

Jan 1, 1970
0
Jim Thompson said:
Here it is done at the device-level on one of my chips...

Of course you can do it with off-the-shelf OpAmps and analog switches,
which is what the devices are all about.

I'm sorry if I gave anyone the wrong idea. We don't want to
invent new temp monitoring schemes; just choose a chip to use.

Trying to hang surface mount chips onto coolant lines has
little appeal to us...

D

#### David Lesher

Jan 1, 1970
0
If you want the difference to be accurately measured, that's
actually an ideal place to put in a thermocouple, or several
thermocouples in series.

No, I dealt with TC's in my youth, but I've been in recovery

D

#### David Lesher

Jan 1, 1970
0
Both LM35 (analog) and LM71 (digital) types could have problems at the
ends of long cables. LM35s love to oscillate if they see much load
capacitance, and any induced noise can make them do weird things. SPI
would of cource have to be handled carefully at the end of a cable.
RTDs and thermocouples are better for field temperature measurements.

I used LM35's on far longer cables than this in years past. It's
only a foot or less from the motor controller CPU to the JFET
heatsink. There's not that much room under a Miata hood.

Replies
1
Views
886
Replies
0
Views
2K
M
Replies
26
Views
2K
M
R
Replies
21
Views
2K
rickman
R
K
Replies
4
Views
1K
Klaus Kragelund
K