RTD linearization
in [Design]
|
From: John Larkin on 13 Dec 2009 14:33 On Sun, 13 Dec 2009 12:46:09 -0500, Spehro Pefhany <speffSNIP(a)interlogDOTyou.knowwhat> wrote: >On Sun, 13 Dec 2009 09:50:22 -0600, the renowned "RogerN" ><regor(a)midwest.net> wrote: > >> >>"John Larkin" <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in message >>news:p1u7i5thbjmtjvqcj63b291l19rf7ktllp(a)4ax.com... >>> >>> >>> Does anybody remember the value of negative resistance that linearizes >>> a 100 ohm platinum RTD? >>> >>> John >> >>I thought RTD's were supposed to be linear. The 100 ohm resistance being at >>0 Degrees C and a change of .385 ohms (for a 100 Ohm RTD) per Degree C. > >DIN-standard Pt100 sensors change 38.5 ohms between 0�C and 100�C, but >it's not linear (although the error would be rather modest over a >typical 'comfort heating' temperature range*). Using a taste of >positive feedback you can make it very close to linear (the residual >error is an S-curve). Typically we would use controlled current >sources because it's hard to compensate for leadwire resistance in >industrial setups otherwise. > >The positive feedback could be set to make the reading dead on at a >third point (eg. mid-scale) with zero and span getting the first two >points, or one could minimize the error according to some cost >function. > >BTW, there is/was also a US standard of 39.2 ohms change from zero to >100 C, but it's about dead now- pretty much everyone uses the European >standard. > >Back the last time I worked with Pt RTD instrumentation design (a few >weeks ago now) I simply implemented a binary search to invert the >Callendar-Van Dusen equations to give T(Rx) (I needed to convert in >both directions-- T(Rx) and Rx(T) for that application). We just do RTD table lookup and interpolation in our thermocouple instruments where we use an RTD as the reference junction sensor, which is just one direction. We have to do that for the thermocouples anyhow, so it's just another, rather small, lookup table. We mux the excitation and sensing at 1/8 duty cycle, and use a 24 bit ADC with low excitation current, so self-heating is nil. ftp://jjlarkin.lmi.net/1300.JPG John
From: Jan Panteltje on 13 Dec 2009 14:46 On a sunny day (Sun, 13 Dec 2009 11:19:49 -0800) it happened John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in <steai5hk9pm4gu0lim9lmv35mgig93m6t6(a)4ax.com>: >On Sun, 13 Dec 2009 17:54:56 GMT, Jan Panteltje ><pNaonStpealmtje(a)yahoo.com> wrote: > >>On a sunny day (Sun, 13 Dec 2009 09:28:53 -0800) it happened John Larkin >><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in >><8k7ai51i2m2us5a8nl49cb0llls0lq4jdr(a)4ax.com>: >> >>>On Sun, 13 Dec 2009 12:20:05 GMT, nico(a)puntnl.niks (Nico Coesel) >>>wrote: >>> >>>>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >>>> >>>>>On Sat, 12 Dec 2009 12:46:37 -0800, Joerg <invalid(a)invalid.invalid> >>>>>wrote: >>>>> >>>>>>John Larkin wrote: >>>>>>> >>>>>>> Does anybody remember the value of negative resistance that linearizes >>>>>>> a 100 ohm platinum RTD? >>>>>>> >>>>>> >>>>>>No uC at hand for this job? Maybe this helps: >>>>>> >>>>>>http://pdfserv.maxim-ic.com/en/an/AN3450.pdf >>>>>> >>>>>>But you don't have to use a Maxim opamp :-) >>>>> >>>>>I'm thinking I'll use 1K RTDs for the automation project, and lay out >>>>>an interface board... easier than hand wiring. The little RS232 widget >>>> >>>>1k RTDs are easier to interface. I used one to control my floor >>>>heating. 2k2 (IIRC) in series from 3.3V and then fed directly into an >>>>ADC. In a limited temperature range, the output is quite linear so >>>>there is not really a need for fancy math. >>> >>>I'm thinking along these lines... >>> >>>ftp://jjlarkin.lmi.net/RTD.jpg >>> >>>All the 1Ks will be 0.1%. >>> >>>The opamp will have a gain of about 8, input and output centered on >>>+2.5. This should be accurate to a fraction of a degree C, so we'll >>>know how to dress. The Z-wave home automation systems have a combined >>>motion detector and temperature sensor (no temp-only nodes I know of), >>>but I don't trust their accuracy and they only go down to 40F, not >>>good for outdoors. >>> >>>I can run RG174 to the RTDs. No EMI/ESD hazards as with semiconductor >>>sensors. >>> >>>John >> >>For a reasonable long cable to the sensor, a 4 wire system with current source would be better. > >Too complex. My engineering budget (design, pcb layout) is one day. I >can do the software up in the cabin during the holidays. > >>Else you will have to include the variance of the resistance of the copper, >>plus the voltage drop in the cable. > >RG174 is about 40 milliOhms per foot, and I doubt I'll need to run >over 25 feet to the outdoor sensor (I get to use my Makita hammer >drill again!) That's about a 0.25C error, which I can mostly correct >for anyhow. > >>You will also have to take into account cut or shorted cables, and the sort of signals >>that wil lgenerate, and how to handle that. > >Why would my cables get cut or shorted? And it's not as if that would >go undetected. > > >>I find this a bit primitive. > >Simple is not primitive, it's good engineering. > >Go to a store and look at electronic or mechanical thermometers on >display. You may see a 5 degree C spread. I'll be happy with 1 degree >true accuracy. > >John Then why not use a simple chip sensor? i2c or any other protocol exists. My inside thermostat is the temp_pic: http://panteltje.com/panteltje/pic/temp_pic/index.html You can see the system working in real time here now: http://panteltje.com/panteltje/view_sensors.php I have written the soft so it is in 1 C steps... The outside sensor is a LM135 or something like that, almost connected like your cicuit, but only 3 meters wire. It will be freezing here tonight perhaps. Nice and warm inside, all temps ara logged too, and can be controlled from anywhere in the world. >
From: Nico Coesel on 13 Dec 2009 15:01 John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >On Sun, 13 Dec 2009 12:20:05 GMT, nico(a)puntnl.niks (Nico Coesel) >wrote: > >>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Sat, 12 Dec 2009 12:46:37 -0800, Joerg <invalid(a)invalid.invalid> >>>wrote: >>> >>>>John Larkin wrote: >>>>> >>>>> Does anybody remember the value of negative resistance that linearizes >>>>> a 100 ohm platinum RTD? >>>>> >>>> >>>>No uC at hand for this job? Maybe this helps: >>>> >>>>http://pdfserv.maxim-ic.com/en/an/AN3450.pdf >>>> >>>>But you don't have to use a Maxim opamp :-) >>> >>>I'm thinking I'll use 1K RTDs for the automation project, and lay out >>>an interface board... easier than hand wiring. The little RS232 widget >> >>1k RTDs are easier to interface. I used one to control my floor >>heating. 2k2 (IIRC) in series from 3.3V and then fed directly into an >>ADC. In a limited temperature range, the output is quite linear so >>there is not really a need for fancy math. > >I'm thinking along these lines... > >ftp://jjlarkin.lmi.net/RTD.jpg > >All the 1Ks will be 0.1%. > >The opamp will have a gain of about 8, input and output centered on >+2.5. This should be accurate to a fraction of a degree C, so we'll >know how to dress. The Z-wave home automation systems have a combined How do you cancel the offset error from the opamp? I'd go for an 12bit A/D converter with internal mux. In my circuit the voltage across the RTD varies 2.8mV per degree Celcius. With a 12 bit A/D referenced at 3.3V (same 3.3V that supplies the RTD so errors cancel) you'll get 806uV per bit so thats 3.5bits per degree. Since temperature changes are slow you can take the average value of a lot of samples so accuracy may be better than 12 bit. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... "If it doesn't fit, use a bigger hammer!" --------------------------------------------------------------
From: Jim Thompson on 13 Dec 2009 15:24 On Sun, 13 Dec 2009 20:01:06 GMT, nico(a)puntnl.niks (Nico Coesel) wrote: >John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > >>On Sun, 13 Dec 2009 12:20:05 GMT, nico(a)puntnl.niks (Nico Coesel) >>wrote: >> [auto-snip] >>>1k RTDs are easier to interface. I used one to control my floor >>>heating. 2k2 (IIRC) in series from 3.3V and then fed directly into an >>>ADC. In a limited temperature range, the output is quite linear so >>>there is not really a need for fancy math. >> >>I'm thinking along these lines... >> >>ftp://jjlarkin.lmi.net/RTD.jpg >> >>All the 1Ks will be 0.1%. >> >>The opamp will have a gain of about 8, input and output centered on >>+2.5. This should be accurate to a fraction of a degree C, so we'll >>know how to dress. The Z-wave home automation systems have a combined > >How do you cancel the offset error from the opamp? I'd go for an 12bit >A/D converter with internal mux. In my circuit the voltage across the >RTD varies 2.8mV per degree Celcius. With a 12 bit A/D referenced at >3.3V (same 3.3V that supplies the RTD so errors cancel) you'll get >806uV per bit so thats 3.5bits per degree. Since temperature changes >are slow you can take the average value of a lot of samples so >accuracy may be better than 12 bit. Why not a variation on Jim Williams' idea... http://analog-innovations.com/SED/TemperatureSensor.pdf I've recently implemented such a variation on a custom chip. Advantages... AC Gain avoids DC offset accumulation It's trivial to implement DC restoration (aka pick your own baseline) ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | Help save the environment! Please dispose of socialism properly!
From: Spehro Pefhany on 13 Dec 2009 16:03
On Sun, 13 Dec 2009 13:24:49 -0700, the renowned Jim Thompson <To-Email-Use-The-Envelope-Icon(a)My-Web-Site.com/Snicker> wrote: >On Sun, 13 Dec 2009 20:01:06 GMT, nico(a)puntnl.niks (Nico Coesel) >wrote: > >>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Sun, 13 Dec 2009 12:20:05 GMT, nico(a)puntnl.niks (Nico Coesel) >>>wrote: >>> >[auto-snip] >>>>1k RTDs are easier to interface. I used one to control my floor >>>>heating. 2k2 (IIRC) in series from 3.3V and then fed directly into an >>>>ADC. In a limited temperature range, the output is quite linear so >>>>there is not really a need for fancy math. >>> >>>I'm thinking along these lines... >>> >>>ftp://jjlarkin.lmi.net/RTD.jpg >>> >>>All the 1Ks will be 0.1%. >>> >>>The opamp will have a gain of about 8, input and output centered on >>>+2.5. This should be accurate to a fraction of a degree C, so we'll >>>know how to dress. The Z-wave home automation systems have a combined >> >>How do you cancel the offset error from the opamp? I'd go for an 12bit >>A/D converter with internal mux. In my circuit the voltage across the >>RTD varies 2.8mV per degree Celcius. With a 12 bit A/D referenced at >>3.3V (same 3.3V that supplies the RTD so errors cancel) you'll get >>806uV per bit so thats 3.5bits per degree. Since temperature changes >>are slow you can take the average value of a lot of samples so >>accuracy may be better than 12 bit. > >Why not a variation on Jim Williams' idea... > >http://analog-innovations.com/SED/TemperatureSensor.pdf > >I've recently implemented such a variation on a custom chip. > >Advantages... > >AC Gain avoids DC offset accumulation > >It's trivial to implement DC restoration (aka pick your own baseline) > > ...Jim Thompson There's also a clever way to use a third current to compensate for the base resistance of Q2 and lead wire resistance. Hard to filter the snot out of it if you're switching currents around though, and diode-connected transistors are reputed to rectify. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff(a)interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com |