|
Guys, Platinum RTD 100Ohm temperature measuring resistor is as common as thermocouples and thermistors. Its resistance is 100Ohm at 0C and about 140Ohnm at 100C. I am especially interested in 125-140 Ohm interval. Has anyone have practical experience with interfacing these with BX24? Any and all suggestions will be greatly appreciated. Thanks for your time. Cordially, Alex. |
|
|
Temperature measurments with RTD-100 Ohm.
Started by ●March 14, 2002
Reply by ●March 14, 20022002-03-14
|
The easiest way is to use an op- amp interface, as the available voltage swing from the pt100 is too small for accurate measurements when usd directly. . You need to supply current to the pt100 via a constant current source ( preferably) or say a 10K resistor from a stabilised power supply. 12v or so is better than +5 as the resistor value can be higher and the constant current effect better. The junction of the pt100 and the feed resistor is connected to the + input of the op amp. A voltage divider using a 10K ( or the same value as used to feed the pt100), a 100ohm ten turn pot and a 47ohm resistor is connected to the same +12 ( or whatever you used for the pt100) and to gnd. The slider of the pot goes to the - input of the opamp via a 1K. The pot is the "zero" adjustment. This arrangement of resistor values is such that the voltage at the - input is the same as the voltage from the pt100 when in an ice bath ( i.e at 100 ohms. You can easily calculate the values with ohms law. The gain resistor of the op amp ( from - input to output) is a fixed resistor of such a value ( try 100k initially) to give an op amp output of about 10mv per deg C. with a 0 output at 0 deg C. You then read this amplified output using the ADC on the bx24. In the interest of stability you need stabilised power supplies, good quality pots and padd out the pots using series/parallel resistors so that the effect of the pots is minimal. You don't want the zero pot for instance to cause wild swings in output. The pot should take 2-3 turns to move the op amp output thru 1 degree. Use a "good" op amp and metal oxide resistors for good temperature stability. neil; biovirus1 wrote: > Guys, > > Platinum RTD 100Ohm temperature measuring resistor is as common as > thermocouples and thermistors. Its resistance is 100Ohm at 0C and > about 140Ohnm at 100C. I am especially interested in 125-140 Ohm > interval. > Has anyone have practical experience with interfacing these with BX24? > > Any and all suggestions will be greatly appreciated. > > Thanks for your time. > > Cordially, > > Alex. > |
Reply by ●March 14, 20022002-03-14
|
Good assessment. Also don't use so much current that you heat the RTD. Mark Skala ----- Original Message ----- From: "Neil Jepsen" <> To: <> Sent: Thursday, March 14, 2002 3:37 PM Subject: Re: [BasicX] Temperature measurments with RTD-100 Ohm. > The easiest way is to use an op- amp interface, as the available voltage > swing from the pt100 is too small for accurate measurements when usd > directly. > . You need to supply current to the pt100 via a constant current source > ( preferably) or say a 10K resistor from a stabilised power supply. 12v > or so is better than +5 as the resistor value can be higher and the > constant current effect better. The junction of the pt100 and the feed > resistor is connected to the + input of the op amp. A voltage divider > using a 10K ( or the same value as used to feed the pt100), a 100ohm ten > turn pot and a 47ohm resistor is connected to the same +12 ( or whatever > you used for the pt100) and to gnd. The slider of the pot goes to the > - input of the opamp via a 1K. The pot is the "zero" adjustment. This > arrangement of resistor values is such that the voltage at the - input > is the same as the voltage from the pt100 when in an ice bath ( i.e at > 100 ohms. You can easily calculate the values with ohms law. > The gain resistor of the op amp ( from - input to output) is a fixed > resistor of such a value ( try 100k initially) to give an op amp output > of about 10mv per deg C. with a 0 output at 0 deg C. You then read this > amplified output using the ADC on the bx24. > > In the interest of stability you need stabilised power supplies, good > quality pots and padd out the pots using series/parallel resistors so > that the effect of the pots is minimal. You don't want the zero pot for > instance to cause wild swings in output. The pot should take 2-3 turns > to move the op amp output thru 1 degree. Use a "good" op amp and metal > oxide resistors for good temperature stability. > neil; > biovirus1 wrote: > > > Guys, > > > > Platinum RTD 100Ohm temperature measuring resistor is as common as > > thermocouples and thermistors. Its resistance is 100Ohm at 0C and > > about 140Ohnm at 100C. I am especially interested in 125-140 Ohm > > interval. > > Has anyone have practical experience with interfacing these with BX24? > > > > Any and all suggestions will be greatly appreciated. > > > > Thanks for your time. > > > > Cordially, > > > > Alex. > > > > > > > > > > > > ">http://docs.yahoo.com/info/terms/ |
Reply by ●March 14, 20022002-03-14
|
>The easiest way is to use an op- amp interface, as the available
voltage >swing from the pt100 is too small for accurate measurements when usd >directly. >. You need to supply current to the pt100 via a constant current source Maxim has a very good application note covering PRECISELY this problem. They interface a pt100 to a (I think) MAX197 a/d converter if I remember rightly (and the app. note is for that A/D device). However, there's one tiny flaw in the schematic. The + and - inputs to the section of the opamp processing the pt100 bridge output are reversed! This means that if you build the circuit exactly as shown, you get output DECREASING with increasing temp and being on the wrong side of ground. Reversing the + and - inputs to the opamp gives you a positive signal that increases with increasing temperature. Of course, if you're driving a BX then you don't need the Max chip either (since you'd be using the BX's inbuilt analog input). But the rest of the circuit stands. The reason I know so much detail about this is that I helped a guy (on another list?) with exactly this problem and exactly this circuit. And he managed to get it working. He'd even emailed Maxim tech support, and they said "nope, there's no known problems with the schematic". [That was before we figured out the + and - inputs were reversed]. Haven't mailed them since. Wonder what they'd say? Regards, David. _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp. |
|
|
Reply by ●March 15, 20022002-03-15
|
Thanks guys for all the help. I found the app note that David mentioned : http://dbserv.maxim-ic.com/appnotes.cfm?appnote_number'0 However, looking at it (fig 5, right?) I also see that MAX chip supplied the reference voltage to the circuit. being a really lousy electric engineer, I humbly submit my questions: in this circuit I see two parts: one - constant current source for the RTD and another one is doing something funny with the voltage across the RTD. I understand why I neeed constant current source - if I know that the current is constant, then measuring the voltage accross the RTD would give me the resistance of the RTD, case closed. But why do I need that other part? Does it conditions the voltage swing to more easily measurable for the ADC? Do I need to change any R values to make it better adjusted to BX24 ADC? And what about the current source? Are there simpler solutions than MAX 197 +opamp? Are there any ICs that can give me the same current source? Since I will not be using MAX197 for ADC, I think its a waste to have it just as a voltage source for op-amp current source. Thank you so much for your help. Alex. --- In basicx@y..., "David Emrich" <dav_ucc@h...> wrote: > >The easiest way is to use an op- amp interface, as the available voltage > >swing from the pt100 is too small for accurate measurements when usd > >directly. > >. You need to supply current to the pt100 via a constant current source > > Maxim has a very good application note covering PRECISELY this problem. They > interface a pt100 to a (I think) MAX197 a/d converter if I remember rightly > (and the app. note is for that A/D device). > > However, there's one tiny flaw in the schematic. The + and - inputs to the > section of the opamp processing the pt100 bridge output are reversed! This > means that if you build the circuit exactly as shown, you get output > DECREASING with increasing temp and being on the wrong side of ground. > Reversing the + and - inputs to the opamp gives you a positive signal that > increases with increasing temperature. > > Of course, if you're driving a BX then you don't need the Max chip either > (since you'd be using the BX's inbuilt analog input). But the rest of the > circuit stands. > > The reason I know so much detail about this is that I helped a guy (on > another list?) with exactly this problem and exactly this circuit. And he > managed to get it working. He'd even emailed Maxim tech support, and they > said "nope, there's no known problems with the schematic". [That was before > we figured out the + and - inputs were reversed]. Haven't mailed them since. > Wonder what they'd say? > > Regards, > David. > > _________________________________________________________________ > Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp. |
Reply by ●March 15, 20022002-03-15
|
>Thanks guys for all the help. >I found the app note that David mentioned : >http://dbserv.maxim-ic.com/appnotes.cfm?appnote_number'0 > >However, looking at it (fig 5, right?) I also see that MAX chip >supplied the reference voltage to the circuit. That's coz it has a hand reference output that's a nice "digital voltage" of 4.096V. However, anything constant voltage would do, since it's used to generate a constant loop current through the 3K3 resistor. (4.096V / 3K3 gives slightly over 1mA loop current). being a really lousy >electric engineer, I humbly submit my questions: in this circuit I >see two parts: one - constant current source for the RTD and another >one is doing something funny with the voltage across the RTD. Amplifying it! 10 times in fact. >I >understand why I neeed constant current source - if I know that the >current is constant, then measuring the voltage accross the RTD would >give me the resistance of the RTD, case closed. Yes. Except the change in resistance for a given temperature change is very small. Therefore the change in voltage is also correspondingly small. To increase the "swing" of voltage for the same temperature change, the second half of the op-amp (the one with the 47K and 470K resistors) multiplies the voltage change by 10. (470K / 47K). >But why do I need that other part? Does it conditions the voltage >swing to more easily measurable for the ADC? You could say it that way, yes. :-) OH, btw, VERY IMPORTANT POINT.... When I was talking about swapping the + and - inputs of the opamp, I was actually lying my a**e off. You need to swap the "sense" pins of that circuit (leaving the resistor arrangement around the opamp exactly as it is!). That is, where they show two straight lines going left from the little circles called "sense -" and "sense +", they should actually cross over before they get to the RTD. The "bottom" (closest to ground) of the RTD should go to the 47K resistor connected to the "-" of the opamp, and the "top" of the RTD should go to the 47K resistor connected to the "+" of the opamp. >Do I need to change any >R values to make it better adjusted to BX24 ADC? No. The input range of the MAX197 in this circuit is 0-5V. This matches the BX24. >And what about the current source? Are there simpler solutions than >MAX 197 +opamp? Are there any ICs that can give me the same current >source? Probably the easiest way to get a current source is to replace the MAX197 with a 78L05 (or similar). Note, it's NOT a pin-for-pin replacement ;-) Since you can get dual opamp chips very easily (that is, two separate op-amps inside one chip package), using the same circuit as Figure 5, but creating your own "reference" voltage is pretty easy. Make sure your reference voltage isn't higher than 5V though, and I wouldn't go any lower than 4V. Where the second (half of the) opamp connects to "CH.x" of the Max chip, replace that with whatever ADC input you want on the BX. What isn't mentioned in figure 5 is the power supply for the op-amp. The circuit that I know worked (using the MAX197 as it happens, but that's not relevant) used a "dual supply",.. .that is a positive AND negative supply and ground. Even though the output of the opamp should never go negative (since the current flow is always "down" the RTD) having the split rails makes the biassing nicer, and keeps the circuit more stable. If I remember rightly, we used plus and minus 5V rails to the opamp (the 5V rail was the same as the one that supplied the MAX197 and the microprocessor that the guy was using). The minus 5V rail doesn't have to be very strong at all (hardly any current required), and could be derived from the plus 5V rail using one of the Maxim / Dallas / Intersil style of low-current DC inverters (I think the ICL7660 is one). >Since I will not be using MAX197 for ADC, I think its a waste >to have it just as a voltage source for op-amp current source. Most definitely! It was the rest of the circuit I was suggesting, since it makes the job of reading an RTC circuit quite a bit simpler. >Thank you so much for your help. You're welcome. Good luck with the project! Regards, David. _________________________________________________________________ Chat with friends online, try MSN Messenger: http://messenger.msn.com |
|
|
Reply by ●March 15, 20022002-03-15
|
> Hi all Another way to do this is by utilizing a Wheatstone bridge, using 3 fixed resisors and the PT100 as the fourth. This gives a differential voltage dependant of the temperature. By adding a instrumentation amplifier of suitable gain, a voltage span of wanted range can be adjusted to be uses by the expected temperature range. The beauty of this solution is that there are no need for a constant current source, just a voltage reference. On my WEB-pages, there is an exaple using a PT1000 (to reduce current consumption for a battery powered system). The URL is :http://home.online.no/~ofremmi/F3F/WeatherDataSystem/Measuring%20PT1000%20element%20with%20ADC.pdf Best luck with whichever solution you might want to use. Ola Fremming Norway |
Reply by ●March 15, 20022002-03-15
|
On Fri, 15 Mar 2002 10:01:00 +0800, you wrote: > >The reason I know so much detail about this is that I helped a guy (on >another list?) with exactly this problem and exactly this circuit. And he >managed to get it working. He'd even emailed Maxim tech support, and they >said "nope, there's no known problems with the schematic". [That was before >we figured out the + and - inputs were reversed]. Haven't mailed them since. >Wonder what they'd say? > That was me! <G> I was tearing my hair out. I'm thinking - "Hey these Maxim guys have _got_ to know what they're talking about, they make this stuff". I must have torn that circuit down and built it back up from scratch a dozen times thinking that I _had_ to be doing something wrong. Then Dave took a look at it and said "hey Grant, they've got the connections backward." Went down to my bench, switched the wires around and it worked. I think if anyone at Maxim had actually taken a look at that schematic they would have seen the mistake. grant |
