# 8 Channel Thermocouple Based Temperature data logger

Status
Not open for further replies.

#### AtApi

##### New Member
Hi 2 all!
this is my first post:
like i said in the title i need to build a 8 Channel Thermocouple Based Temperature data logger for high temperature (max 2000F)
i study electronic when i was in school but now after 10 years i almost forgot everything :-(
theres any chance that somebody can help in this project? Unfortunately i dont have a lot of time to spend in this project...i just need the schematic and the list of component...
i know it sound lame but i need to build one in 10 days....
Thanx

#### blueroomelectronics

##### Well-Known Member
Good luck with that.

#### Diver300

##### Well-Known Member
I would use a Max6675. It converts a thermocouple input to SPI for connection to a microcontroller

http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3149

That will read to 1875 °F (1024 °C). I don't know what to do if you need higher than that.

You need to wire 8 Max6675 ics to a microcontroller. You can parallel all the clock inputs. The NCS lines of numbers 1 and 2, 3 and 4 etc can be connected together. The SO lines of all the odd numbered Max6675 can be connected together, as can the even ones. That way you can read 8 temperatures on 7 wires (1 clock, 2 data and 4 not chip select)

Then you can store the data on an EEPROM or send it to a serial line to use a PC to store the data.

You will need to be able to program a PIC.

If you need someone to design, build and program it, expect to pay for their time.

#### AtApi

##### New Member
Diver300
look at this
http://www.vincenzov.net/eng/design/max147.htm or
http://www.vincenzov.net/progetti/ntc/ntcschematic.gif

the first is a simple 8 channels / 12 bit ADC connected to the parallel port of a personal computer that is able to measure eight voltage and to transfer them to a PC.
The second is the schematic for a 16 channel Temperature data logger but how you can see he is using a MAX 187 that i dont think is good for my temperature goal...and both send the captured data to a pc with a serial port
i would like to build a data logger that is able to store on a eeprom or even better on a SD card and have a USB connection to transfer the data
Off course ill pay as long as is a reasonable price... PM me who is Interested...
I have big clients, if everything works we will make big money.

#### blueroomelectronics

##### Well-Known Member
Actually the MAX6675 that Diver300 suggested would appear to work like a charm, SPI is not a tough bus to work with and you can plunk an SPI EEPROM like the 25LC256 no problem.
Here's the rub, if you know what your doing it could be done in 10 days but that's if you've already got the parts and the design started. You still have to program and debug any working design.

Why only 10 days, sounds like a school project.

#### Diver300

##### Well-Known Member
It is 10 years since AtApi was at school, so he is either slow to graduate or it isn't a school project!

The Max6675 does work very well. There might be issues with availability. I only used a couple and they were samples from Maxim.

As with all thermocouple circuits, the voltage on the thermocouple is tiny and noise can be a problem. I had some problems getting rid of noise, (I can't remember the details) but it was on a board switching 10A at 230V ac.

It measures the temperature of the oven, for the temperature control. If it didn't work, I wouldn't eat and my wife would kill me.

#### blueroomelectronics

##### Well-Known Member
AtApi said:
Diver300
look at this
http://www.vincenzov.net/eng/design/max147.htm or
http://www.vincenzov.net/progetti/ntc/ntcschematic.gif

the first is a simple 8 channels / 12 bit ADC connected to the parallel port of a personal computer that is able to measure eight voltage and to transfer them to a PC.
The second is the schematic for a 16 channel Temperature data logger but how you can see he is using a MAX 187 that i dont think is good for my temperature goal...and both send the captured data to a pc with a serial port
i would like to build a data logger that is able to store on a eeprom or even better on a SD card and have a USB connection to transfer the data
Off course i`ll pay as long as is a reasonable price... PM me who is Interested...
I have big clients, if everything works we will make big money.
10 days is way to short for a finished product good enough to sell. It's short for a hand wired proof of concept even.

How many samples do you need to capture, over what time? Is it AC or battery? What probes are you going to use? Do you need a display in the design or a keypad, what enviroment will this device be used in? What size are you hoping for?

10 days no way.

Edit: I've looked at the two links, neither of those is a logger. Just simple A/D and mux. Your PC is the logger.

PS are you Vincenzo, AtApi?

Last edited:

#### Nigel Goodwin

##### Super Moderator
There are a number of different stages required here:

1) Convert the thermocouple reading to the required datalogger range (0-5V?), seperate opamp circuits required for each channel.

2) Convert to digital - a PIC 16F877 gives you 8 10 bit analogue inputs, which should be more than enough resolution?.

3) Storing the readings - obviously EEPROM is an easy option, and can provide reasonable amounts of storage.

4) Displaying the readings - either on LCD, or transfer via serial to a PC.

Something like a PIC16F877 should do the job nicely, and EPE did an 877 datalogger project years back, you just need to make the opamp thermocouple circuits.

Just googled, EPE August/September 1999 was the 8 channel datalogger project. http://www.epemag.wimborne.co.uk/0899.htm

#### mcs51mc

##### Member
blueroomelectronics said:
How many samples do you need to capture, over what time? Is it AC or battery? What probes are you going to use? Do you need a display in the design or a keypad, what enviroment will this device be used in? What size are you hoping for?
What accuracy do you need?
What resolution do you expect?
What type of thermocouple will be used or do you expect accurate temperature measurment over the whole range 0 to 2000F with only one type?
What about thermocouple correction? Every thermocouple has it's own error that needs to be entered in the system to have correct readings.
What is the maximum distance between thermocouples and logger?

maxim-ic datasheet of MAX6675 said:
This converter resolves temperatures to 0.25°C, allows readings as high as +1024°C, and exhibits thermocouple accuracy of 8LSBs for temperatures ranging from 0°C to +700°C.

#### mcs51mc

##### Member
If you can't afford the equipment I suggested before
you can try this one: http://www.picotech.com/thermocouple.html

So what's the point of reinventing the wheel
Can you offer MORE for a better price? I don't think so

#### AtApi

##### New Member
Ok first of all Thanx Thanx Thanx you guys make this forum one of the best!!!

This is the specs that this device should have:
sample data rate as quickly as 1000 times per second (Analog)
Uploading recorded data to PC is done via a SD memory card or USB or EEPROM
Battery Voltage
it would be nice to have a LCD display but not necessary
Is gonna be used in racing enviroment
Real-time clock stores data and time
12 bit resolution i think should be fine
i m gonna use the probe showed in this web site http://ifamilysoftware.com/8307.html and actually the data logger in this site is pretty much what i want to build but 8 channels.
i can even buy some kit if i find where...
i found this http://www.magenta2000.co.uk/kits/877.htm do you guys think that i can use it for my project?
or i can pay someone to draw the schematic for me...
i can push to 30 days but i need to know if someone can help me in this project.

P.S: for Nigel Goodwin

how i can get the schematic for the project that EPE did on August/September 1999 ? (http://www.epemag.wimborne.co.uk/0899.htm)

Anyways thanx for your help.... you guys are great!!!!

#### Nigel Goodwin

##### Super Moderator
AtApi said:
i found this http://www.magenta2000.co.uk/kits/877.htm do you guys think that i can use it for my project?

P.S: for Nigel Goodwin

how i can get the schematic for the project that EPE did on August/September 1999 ? (http://www.epemag.wimborne.co.uk/0899.htm)

Anyways thanx for your help.... you guys are great!!!!
Magenta do a kit for it, you mentioned it above!.

Like I said before, it's 8 channel but 10 bit - however, 10 bit is probably more than enough?, with 12 bit the extra resolution is likely to be lost in noise anyway, and requires very careful PCB layout.

#### mcs51mc

##### Member
Nice, unrealistic specs you have there... ...
1) Sample rate 1000S/s for temperature? Come on, what temperature slope do you expect 100° in 1sec? Even then you have enough with 200S/s

2) You wrote: "12 bit resolution i think should be fine" What kind of spec is that. Please answer my questions: "What accuracy do you need?
What resolution do you expect?" then we can do the math and see how many bits you need. I can already tell you that 10 bit isn't enough
It's very easy math you know, I wonder why you didn't already figure it out by yourself.
Did you read the Pico specs: 20 bit resolution

You can use the Magenta kit if you have a temperature to voltage converter. Be aware that you need at least 1.22mV accuracy. So like Nigel already said, it will require very careful PCB layout.

About the probe, they say 0.1 resolution in a -328 to 2498 range, to achieve that you need at least 15 bit resolution. I think they use standard 16 bit ADC's

BTW it's for racing environment, then try one of these: http://www.picotech.com/auto/index.html

#### AtApi

##### New Member
My specs are correct,
1000/s is correct... have you ever log the temperature that coming out of a dragster muffler? or a NASCAR race car? this cars use nitro-methane where there s detonations and temperature shock (extremely rapid changes in temperature).I have already the thermocouple probes that are inserted into the header pipe to measure the temperature of the exhaust gas as it exits the cylinder head.They are Type-K thermocouple assemblies (Nickel-Chromium/Nickel-Aluminum). Each require an available analog channel and must be used with the appropriate thermocouple amplifier module. All probes are 12” in length and are terminated with a two pin male miniature-connector.

You may have right when you said 15 bit resolution...the higher the better,this guys want precision......
look at this multi logger:
http://www.racepak.com/V300SD.php

I dont think that http://www.picotech.com/auto/index.htm is what i need...
the logger has to stay on-board..

and has to be strong enough so G forces will not tear this unit apart.

Ok now heres my thing:
I move to USA for work, i used to do electronics more then 10 years ago, then i study informatic and i became an IT guy, and unfortunately i forgot a lot of things so im sorry.. be patient.
I have friends that are in the high level racing enviroment and they all use dataloggers so they can study how the engine react so they can improve performance.
Now if theres someone who wanna work with me in this project let me know...