PT100 conditioning circuit

Go To Last Post
37 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello friends,
During my search for pt100 conditioning circuit. I came across the circuit.
Pls find it in the attachment.
I have build this circuit using lm358 on a general purpose board with following changes.

I have shroted the ref & vcc pin to 5v.
I am not using the MAX197 IC.
Rest of the circuit remains the same.

----------- Testing ----------

Instead of sensor I have connected 50 ohm,100 ohm,180ohm resistor (all mfrs). The output of diff amp. is 3.749v,3.74v,3.74v in the output.

1. I'm not getting good diff. in output.What is the problem ?
2. I need to measure -40 deg to 400 deg cel. Will this circuit provide me the range ?
3. What necessary changes the circuit requires ?
4. I'm using 3 wire pt100? How it should be connected in the circuit?

pls help me :(
Thanks in advance

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Suggest you repost as a graphic (GIF/JPEG) as I doubt that anyone here would be mad enough to open a possibly virused .doc file from an unknown source.

Cliff

PS as I said to someone else asking about PT100 this very morning (kind of coincidental!) a search for "PT100" here will hit about 38 threads - surely one of those has something to tell you about how to use it?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

sorry
now I've attached a zip file.

Instead of sensor I have connected 50 ohm,100 ohm,180ohm resistor (all mfrs). The output of diff amp. is 3.749v,3.74v,3.74v in the output.

1. I'm not getting good diff. in output.What is the problem ?
2. I need to measure -40 deg to 400 deg cel. Will this circuit provide me the range ?
3. What necessary changes the circuit requires ?
4. I'm using 3 wire pt100? How it should be connected in the circuit?

pls help me Sad
Thanks in advance

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Posted to the wrong thread - that was previously locked...

Attachment(s): 

You can avoid reality, for a while.  But you can't avoid the consequences of reality! - C.W. Livingston

Last Edited: Mon. May 19, 2008 - 01:01 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Perhaps you could ask the moderators to append this to your existing thread?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Threads now merged.

vijay.naikwade, please do not start multiple threads about the same topic (AKA "cross posting")

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello friends,
Still I am not able to build working circuit for pt100 conditioning.
:(
I tried two options.

--- Wheatstone bridge circuit ------

1. using wheatstone bridge balanced for -40 deg cel

---- problems --
1. Not able to derive formula for T < 0 deg cel.

----- constant current source method ----

I have built a circuit using lm358 which provides 1 ma constant current.
pls find the circuit in the attachment.

----- problems -----------
1. how to connect pt100 in the circuit?

2. when tested for 50,100,180 ohm res. I haven't got a distinct voltage drop.

Let me thank :D all the friends for their help.

if possible, pls give a circuit which I can evaluate.

If it is not possible, pls tell me from where I can buy a pt100 conditioning board with following specifications.

-----specifications----

sensor - pt100 ( 3 wire )
no. of channels -- 3
temperature range -40 deg cel to 400 deg cel
temp. accuracy - 2% of full scale.
supply volt. - 3v3

I will be connecting this board to my meshnetics development board.
The same conditioning circuit should be used for all 3 channels.

1. what will be the price?

Thanks in advance !!!

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

vijay.naikwade wrote:

----- problems -----------
1. how to connect pt100 in the circuit?
??? Please be more specific.
Quote:

2. when tested for 50,100,180 ohm res. I haven't got a distinct voltage drop.
I don't understand the "sense"-names of your schematic.
To start very simple: If you take the Op-Amp providing the 1.2mA current, your pt100, the 3.3k resistor and ignore the second op-amp and all 47x-resistor you should be able to measure 1.2mA at any temperatur. pt100 means 100Ohm at 0 deg. Celsius. Taking a digital voltage meter (DVM) you should measure 1.2mA * 100R = 120mV.
The second 0p-amp is only an differtial input amplifier with an amplification factor of 10. So 120mV should become 1.2V.
Your business is now to think of your requirements, because 3.3V supply voltage can now generate 4,096V Ref-voltage. And also you should be looking to the input voltage range of your ADC. That's a little bit of playing and calculating of the Ref-voltage and the ration of the 3.3k resistor, the variation of the pt100 resistance and the input range of the ADC. The differential amplifier can be set to a factor of 1 by making all 4 resistors identical (e.g. 47k) and finish.

Instead of a 4 line connection as shown by your wird file the low 2 connection can be replaced by one (connect the 47k resistor from the "-" input of the op-amp to the 3.3k resistor directly).

Please avoid word files - .gif / .png are smaller and easier to handle for none-PC users!
Knut

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

And don't open a new topic every time you post.
https://www.avrfreaks.net/index.p...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

vijay.naikwade

That's twice now that I've merged threads - like CountZero says would you PLEASE stop this habit of starting new threads on the same subject. It's called cross posting. What happens is that you start one thread, abandon it, start a new one, others reply to the first thread then find that someone else has given the same answer on the new thread. This makes them very CROSS!

(and the moderators who have to keep merging things back together again)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry, From here onwards I will never do that mistake again.

1. how to connect pt100 in the circuit?

Since the word " sense " is shown 4 times in the schematic. I thought that the connection is for 4 wire pt100 sensor & there might be some change for the three wire pt100. That's why I asked the question.

2. when tested for 50,100,180 ohm res. I haven't got a distinct voltage drop.

Since I had pt100 connection doubt, instead of pt100 I have connected 3 resistance of above value.

thanks !!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have not taken a look at your circuit, if you are unable to post it as a png or gif why not post it as a pdf?

At least one of the references mirocarl gave you deals with connecting 3 wire sensors.....

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

All the sense wires do is compensate for the resistance of the wires that connect the RTD to the amplifier. There is nothing magical about the sense wires. In fact, generally, as the resistance of the sense wiring is "Self-Canceling " you really need not take them into account in any of the analysis of the design.

You can avoid reality, for a while.  But you can't avoid the consequences of reality! - C.W. Livingston

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am referring the same doc which Microcarl has suggested.
pls find it in the attachment.

I'm talking about the figure 9: digital approach.

As per your suggestion I have tested the circuit by removing the diff. amplifier.

(op-amp lm358) vcc=vref=3v3 and 3k3 is kept as it is.

When 100 ohm resistor connected in place of pt100.
I got 1.936v at the output pin.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

vijay.naikwade wrote:
Sorry, From here onwards I will never do that mistake again.

1. how to connect pt100 in the circuit?

Since the word " sense " is shown 4 times in the schematic. I thought that the connection is for 4 wire pt100 sensor & there might be some change for the three wire pt100. That's why I asked the question.
thanks !!


DId you hear about wikipedia? If not forget your project and start with a book like How to use a PC :twisted:
Search at wikipedia for resistance termometer. There you will find a detailed description of 2 / 3 / 4 wire connection...
If it is easier for you to do something than to read take a ohmmeter and measure the cables of your pt100. Two of them show zero ohm resistance and one will show something greater than 100ohm. Together with the description of wikipedia your measurement result and a self-made sketch you should be able to find a solution by yourself.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

kschwiRG wrote:
Instead of a 4 line connection as shown by your wird file the low 2 connection can be replaced by one (connect the 47k resistor from the "-" input of the op-amp to the 3.3k resistor directly).

Nope. You need a 4 wire connection especially if these wires are long. This is to cancel the resistance fluctuations on the sense wires. I think microcarl also explained this.

This is a really common measurement technique whenever accurate results are a must. Nothing wrong with this one.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hi to all
this website is very good.
thanks alot

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,
while listing through this thread I came across post of Microcarl and document of Microchip "Precision Temperature-Sensing With RTD Circuits".
There seems to be pretty simple circuit for 3-wire Pt100, which I would like to use for my application - temperature measurement from -50°C to 100°C (so I think no linearization will be needed). I want to use Pt100, because I have few of them in my drawer.
I suppose that I will use refference of 2.5V, which is on the picture. I would prefer to use ATmega onchip 10bit ADC.

At -50°C the resistance of Pt100 is 80,31Ohm, therefore A3 opamp ouput voltage will be 0,08031V
At 100°C the resistance of Pt100 is 138,5Ohm, therefore A3 opamp ouput voltage will be 0,1385V
The voltage after A3 is so low, that I need to amplify it at least 10x with A4. Difference between -50 and 100°C will be 0,5819V. When I suppose that I will use 2,5V refference and 10bit ADC, I will have 0,00244 step on ADC and between -50 and 100°C will be step of 0,63°C, which is acceptable, if it will be accurate!

And that's where my questions start:
1. What values of rezistors and capacitors should I use around A4? I tried to simulate it with Microchip SW "FilterLab" - that's the second picture (I can only hope, that it is for MCP604 also - I cannot choose type of opamp there).
2. Do the simulated values seem to be real? Will it be stable and accurate?
3. How will this circuit be susceptible to noise? Will I get meaningful results?
4. How would You improve this schematic?

Thank You very Much for Your help!

Note:"Current Generator Circuit" rezistors are on previous picture and they are 25kOhm.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Vijay - not sure if you have sorted your problem but from your original cct, the differential amplifier inputs are the wrong way round - neg input should go to the feedback network and pos input should go to the potential divider network.

Cheers

Ron

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Vijay's original question was 1.5 years ago!

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, it's more than year old thread :shock: , but I didn't want to make new one, it is same topic :idea:

Please could You focus on my post with MCP604? :?:

Thank You Very Much :!:

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hmmm, for this temperature range I would advise an NTC, much simplier (few pasives) and plenty of resolution and precission if adjusted.

As an example: https://www.avrfreaks.net/index.p....

If you still want to lear how to use RTD's, then be prepared for more complex circuitry and/or programming. I had success to interface up to 5 Pt100 with one ATmega128/1281 using an external 16bit ADC. Although a fixed current source is advised, and there are even some clever tricks for a single Pt100 that can make your life easier, I had done it as voltage divider, a 2V5 precision VRef, and the same linearization technique explained in the example linked above, with really good results.

In some thread in these forums there is an schematic where a 16 bit ADC with differential inputs and external reference works quite nice, with a current supply that flows through the RTD and a fixed high precision 100 ohms as reference.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So Yo think, that it isn't possible to use internal ADC and proposed connections?

Here https://www.avrfreaks.net/index.p... You wrote:

Guillem Planisi wrote:
Yes, Daqq, I already know that, and I had been using Pt1000 for a while with much better results than with Pt100 respect to noise and tolerances

Why are they better? They can be fed by 0,1mA (recommended) current versus 1mA of Pt100. Therefore they have same voltage on the output (on the input of first opamp).
What current do You use for Pt100 or Pt1000 :?:
They are better because We don't need 3-wire (for short distances), but that's all I think - Am I right?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, the Pt1000 are better since for short distances you don't need 3-wire or 4-wire, but also because you have better 'resolution'. If you have Pt1000, then it could be interesting to use the same setup for them as the one I had explained for NTC's: a simple voltage divider.

A current source is not needed if you use a voltage divider, but then you need to linearize the measurement. An OpAmp acting as amplifier will also be recommended if you want to use an expanded scale with a lower resolution ADC, like the one that cames with ATmegas (Xmegas have a 12 bit one). With some tricks, one can even try to use the internal amplifier that cames with some ATmegas when using differential imputs (that means to use a Wheatstone bridge) and override the external OpAmp, but that means a certain study, math, linearization calculations, etc. A nice and entertaining engineering task that can be good to learn some basics and not so basic engineering tools.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
you have better 'resolution'.

I do not think You have better resolution - voltage on the output of RTD will be the same in case of Pt100 and Pt1000, because of lower current in case of Pt1000.

As I wrote:

Quote:
Why are they better? They can be fed by 0,1mA (recommended) current versus 1mA of Pt100. Therefore they have same voltage on the output (on the input of first opamp).
What current do You use for Pt100 or Pt1000?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

One of the big differences is exactly due to the lower current.

In particular, you get additional voltage drops in the connecting wires. With PT100, this is typically solved with a 3-wire or (better) 4 wire arrangement (also known as Kelvin connection). With PT1000, the connecting wires are a smaller fraction of the total resistance, hence contribute fewer errors.

There are other factors, however. One is noise power (higher with higher resistance). There may be others, because PT100 is, by far, the more common sensor.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

True, but I cannot see the difference in "resolution" as Guillem Planisi wrote.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

With a 10 bit ADC, you probably will not see any resolution difference.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Simple: 0.34 ohms per ºC with Pt100, 3.4 Ohms for the same delta with Pt1000. If you use the same current supply, then you have 10 times more variation in voltage with a Pt1000 than with Pt100.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That's not true, because You CANNOT use same current source - Pt1000 can be fed by 0,1mA (recommended by most manufacturers, have a look in datasheets. You are "overheating" Pt1000 if You do use 1mA) current versus 1mA of Pt100.
Which means You will have same voltage on the Opamp input independently of whethet You will use Pt100 or Pt1000.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

That's not true, because You CANNOT use same current source - Pt1000 can be fed by 0,1mA (recommended by most manufacturers, have a look in datasheets. You are "overheating" Pt1000 if You do use 1mA) current versus 1mA of Pt100.

Like Will Rogers, "All I know is what I read in the papers."

US Sensors doesn't have a dire warning: "Maximum applied current: 1 mA"
http://www.ussensor.com/pdfs/pro...

but Vishay (kinda) does: "0.1 mA to 0.25 mA"
http://www.vishay.com/docs/28762...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I suppose sensor will burn above "Maximum applied current: 1 mA", but current for best measurement is 0.1mA.
We might talk about same thing...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I used a 10K 1% resistor as 'current source'. Even if you can apply 0.1mA or 1mA to your PtXXXX, that doesn't mean that you have to. Self heating can be an issue, so this current must be kept to a low value. The products I had developed worked with Pt100, Pt1000, TC J, TC K, NTC, PTC, 4-20mA loops, 'digital' inputs, all within the same two or three terminal blocks.

After some exhaustive tests, I've got better results from Pt1000 than from Pt100 when applying EMI fields, electrical noise, long wires (Pt1000 spec'ed <2m long wires in any case, though), etc. Accuracy, when correctly (hard process, by the way) adjusted and calibrated, was by far better with NTC than with Pt100.

Pt100 had been a kind of standard, and general market is highly 'static'. A thing that had been used for a long time doesn't need to be the best.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry Guillem, but Your posts are a bit vague. Could You recommend some particular schematic? Or would You improve something on proposed schematic?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

In many latin languages, vague can be translated as Lazy. I have to admit, I'm both things ;)

I had added a reduced schematic about my implementation. I can't post more detailed information, since it is a commercial product, but at least you can get the picture with that.

At the left you can find a trhee wire Pt100, a 'load' resistor (10K) and a two low pass filter. They go to a differential/two single ended inputs of an SPI 16bit ADC (you can use ADC7718 instead, with 24 bits resolution) with an internal PGA.

There is a reference voltage not depicted of 2V5. Be warned that there is a 'virtual ground' done with a schottky diode, that is necessary since the negative input of the ADC must be minimum 100 mV above real Ground. The 2v5 reference is tied to this virtual ground instead of real ground.

Software is similar to the code that I had published for the NTC interfacing in the tutorials forum. If you place an NTC within the upper terminal and virtual ground, you can use the same circuitry, and you can attach a thermocouple at the same time between the virtual ground and the terminal at the center (where the compensation wire is attached).

Computing the compensation is more elaborate, but still doable. The main idea is that you have four resistors: the 10K load, the RTD, and two equal wires. You sense the voltages at three points and guess that the return wire has the same resistance than the 'excitation' wire (the one that goes from the RTD to the load resistor).

Another detail is that to properly create the virtual ground, some minimum current must flow trhough the schottky diode, in order to obtain a minimum of 100mV required by the ADC.

Attachment(s): 

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Guillem Planisi wrote:
In many latin languages, vague can be translated as Lazy. I have to admit, I'm both things ;)

I didn't mean it, really! :D

Is the value of rezistor, which is connected to 2,5V reference, really 10k? Too much isn't it? There would be 0,25mA going through Pt100 only.
Why didn't You use more "traditional" way - for example "Opamp" way?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I succesfully tested Pt100 conditioning circuit made according to above mentioned AN687 Understanding RTDs from Microchip.
https://www.avrfreaks.net/modules...

It is pretty stable now (even on breadboard, with long unshielded wires), works fine. I connected it to internal 10-bit ADC of ATmega128 and used external TL431 referrence (dunno if it is necessary, but I will need it anyway for 4-20mA sensors refference).
I will post complete schematics here in 14 days:
https://www.avrfreaks.net/index.p...

I have had problems with LM2575, which I used as power source for my board. I wasn't able to filter it. It made so much noise (it caused mistake of 10°C approx:( ) I had to replace it with 7805 linear stabilizer, which works fine, without noise.