"Line Driver" for low volt D.C. 2

[a31ford]

Hi All !

Here's where my memory is getting foggy, I have a thermo-couple, and I have a Digital Panel Meter( NOT a DMM), I can get the 2 to work as a themometer when the wires between are short (< 3 ft.). here's where I'm at a loss: I need to do that measurement with 220 ft of cable between the T/C & the meter. The meter is now set for 0-20 mv dc. I can set it for 0-2v, 0-20v, etc. what would be a simple circuit that would take the mv. input of the T/C and convert it to a voltage that would run the distance. I'm thinking a 2n2222 as a simple amp. would do the trick, but at a loss as to config. (yes I can have a power source at the other end of that 220 ft of wire).

G.

 

[cbarn24050]

Hi, you don't appear to have any cold junction compensation! so it wont be very accurate. A &quot;simple&quot; dc amplifier will have more drift than your thermocouple produces, so any accuracy that you had left will be gone.A pt100 sensor would be a lot better and easier, assuming you are measuring a suitable temperature.

 

[a31ford]

Hi, Thanks for picking up the thread.

Ok, first off, what is a pt100 sensor ? and what is, or where can I find it's spec's ?

I have a hand full of KTY-10 semiconductor temp. sensors
BUT....

2 of the 5 sensors MUST be able to read around the 1500f mark +/- 500f (expanded scale). (anything below 500f for those 2 units is &quot;ignored&quot;. The other 3 units are in the 32f to 250f area, and should be good in the 100f to 250f area.

G.

 

[IRstuff]

The KTY-10 is a thermistor with RT value of 2K nominal, and is therefore similar to a PT100, which is a platinum thermistor and which has a RT value of 1K.

TTFN

 

[a31ford]

Wow ! way to fast !! lol, I just posted that ! TY IRstuff, see &quot;temp for computer&quot; (this forum) I think I found a simple 3wire sensor that will do for the low temp stuff, now just the high-temp probes....

Thanks,

G.

 

[xnuke]

I'd use a temperature transmitter that can read the thermocouple input and transmit a 4-20 mA signal. If you pass this through a precision 1 ohm resistor across the leads of the panel meter, you'll produce a 4-20 mV input signal, and can then set the zero on your meter accordingly to achieve a full scale reading. This will provide you an offset zero current signal to improve noise immunity, but sacrifices some of the resolution on your meter. If resolution is key, you should be able to also find transmitters that can send a 0-20 mA signal.

 

[automatic2]

Here's a possibility for you to get the job done and learn about some industry standards at the same time. Do an internet search on the I2C buss, specifically in regards to temperature sensors. Many instrument manufacturers utilize this communication method. Temperature sensation almost always leads to a reaction. Actuators of all types as well as fully integrated systems utilize this method of comm.

 

[a31ford]

Xnuke and Automatic, thanks for the replies,

Xnuke,
The current thing has crossed my mind, but the % of accuracy does not need to be 99-100%, consistancy is more the key here (and cost), this whole project is to simply &quot;keep track&quot; of temp's, in an &quot;outdoor wood gasification system&quot; (heater), &quot;district heating&quot; kinda thing, (house, shop & garage), we live out in the woods, and would like to burn wood instead of electricity, to heat it all.... (gets to minus 45 here, 4 mo. of the year). So the key is consistancy (if it's out, it's ok, as long as it's ALWAYS out the same amount).

Automatic,
I2C is a processor based system as far as I can see, and yes we have many computers around the house, but I'm not intrested in doing this via computer.

Actually the D.P.M's are going to be mounted in a wall box (with a &quot;decora&quot; cover plate), so one can glance from different areas of the house, (kitchen, rec room, office), and get an idea of how soon I have to go out and feed the fire again... (making a simple 1 of x auto switcher to cycle through the needed readings, with a led to indicate which is which. eg: wood hearth, reserve tank, outdoor, indoor, flue, etc.

G.

P.S. the D.P.M's are up and running, simply having indoor/outdoor temps are nice but.....

 

[a31ford]

Well, well, well, it appears that a good old 1N4148 diode does the job quite nicely (biased circuit) OK, so far, this is only 20 ft away from the source (so I have to figure out another method for the distance thing)... Go Figure, who would of thought that I could get a DIODE to be as acurate as a fluke Digital Therm. Meter that costs over 500.00, and my curcuit is under 35.00 (side note: working from only 0c to 500c at this point, BWTH thats the temp. area I need)...

p.s. will post findings as I do the work, and a web site with spec's for those that want (reply here. this thread, if intrested)....

G.

 

[gleaf]

Excuse the late entry:
Whay not go to an integrated circuit
1 mico amp per degree C sensor
and have distance immunity?
I believe AD still makes them.

 

[a31ford]

gleaf,

I had given that a thought, but having to sink 1A (not Ma) at 1000 c is a bit much. I ordered some samples from &quot;Microchip&quot; (a simple 3 wire sensor)(Voltage) so far, these appear to work well on the distance thing, but are limited to 200c. The solution I am working on for the two 1000c units appears to use a &quot;Honeywell&quot; thermocouple and a voltage follower circuit using a 741 op-amp (still in design).. Will post the results as time permits.

G

 

[47gandalf]

I've used the AD590 IC temp sensor for years with great success. I like its simplicity. Just push 1 uA per Kelvin through a high-accuracy 1 kohm resistance and get one millivolt per Kelvin. But it won't go near 1500 degF.

However, there is a good Analog Devices' Application Note for this part that shows how to obtain accurate remote temperature measurements at a distance of up to 1000 ft. That might be instructive about the error mechanisms that one can encounter. I'm using these suggestions in a current design.

I strongly suggest you look at the web site for Omega. They have some great information about various methods of measuring temperature. There's good discussion of the appropriate sensing technology for a given temperature range. And of the strengths and weaknesses of the various approaches. They discuss thermocouples, RTDs, thermistors, IC sensors, etc.

I previously used their products when measuring stress-strain curves of power metal alloys at temperatures up to 4000 degF with great results.

 

[a31ford]

47 Gandalf, thanks for the reply,
I will investigate your solutions further...

G