Is there any ic with on chip cold junction compensation and amplification ,which can be used for all types of thermocouples?
I don't understand your list of PGA gain values because the AD590 output is a current. What you need is programmable output voltages of 50uV/°C, 60uV/°C, 40uV/°C, etc. to match the thermocouple Seebeck Coefficient at ambient temperature, right? The AD590 output is 1uA/K, so this really only requires a small resistor of 50 ohms or so and a level shift current to subtract 273uA (to convert K to °C). They show a circuit that does this for a J-Type thermocouple in figure 20 of the AD590 data sheet.
If you do need gains of 0.005, 0.006, etc., I don't think you will find an integrated dc-precise PGA with just the right gains for you. Anyway, the amplifier would add noise and offsets along with the Cold-Junction Compensation voltage, which could be trouble when you are trying to reliably measure single microvolts. The exception would be if you could add the CJC voltage after gain, as it was done in the first article that I linked above. That way the compensation voltage is larger and noise and offsets contribute less uncertainty.
If you do use the AD590 and you can't add the CJC voltage after gain, I would suggest switching in a resistor based on the type of thermocouple, instead of using a PGA. This can be done with a modification of the circuit above. Just for clarity, see the sketch below of what I was thinking based on the numbers you listed above.
ADI thermocouple amplifiers such as AD8494 and AD8495 are designed for J and K-type thermocouples, which were chosen for their sensitivity, availability, and relatively wide linear temperature range. Some legacy parts such as the AD595 can be calibrated for different types of thermocouples with external resistors, but the process is time consuming and not practical for mass production. There is no ADI thermocouple amplifier with selectable gain and cold-junction compensation for many thermocouple types.
Are you designing a universal thermocouple interface, or will each input only interface to a single thermocouple type?
Here are a couple articles about the topic that are worth reading:
I hope this helps.
I am designing a universal analog input module with four channels.Each channel has to accept input from all thermocouple types,two ,three and four wire RTD,4 to 20mA ,and voltage signal(one at a time).In order to interface the thermocouples,I thought of using a temperature sensor(AD590) to measure the cold junction temperature,and then use a summing amplifier to add the ouput of AD590 with the thermocouple voltage so as to get the cold junction compensated output.But the AD590 output has to be made proportional to any thermocouple type used.Is it effective to use a PGA at the output of AD590 for this purpose?
That is one way to do it and it works well. It is similar to the first article that I shared, where they attenuated the CJC voltage based on the type of thermocouple using a multiplexer, and then they used the AD8227 instrumentation amplifier to sum it with the gained thermocouple voltage. The difference being that AD590 has a current output, so you would switch in different resistors or trans-impedance gains rather than attenuating the way they did in the article.
Another nice way to do it is to use a temperature-to-digital sensor to measure the reference junction temperature and then do CJC in the digital domain. This tends to reduce board space, increases flexibility, and allows linearization over a wider reference junction temperature range. If you need more info, the process is outlined in this video. They did a nice job of implementing this method using a flex circuitboard and an ADT7320 in CN0172.
If I use a PGA at the ouput of AD590,I would require different gain factors so as to make its output proportional to the thermocouple type used.So, is there any PGA available whose gain can be software programmable according to my requirements(like gain of .005,.006,.004,.0025,.0005)?
Retrieving data ...