Maker Pro
Maker Pro

LM317 Current Loop (LM34DZ)

B

Bill Stock

Jan 1, 1970
0
I built a temperature sensor based on the LM317 current loop in the LM34
docs. I went this route, as the sensor is about 100 feet from the input
device. But the temp is jumping up and down by a couple of degrees all the
time (noise?). I went with the current loop due to the length of the wire to
the sensor/317. The one note I did read in the LM317 docs suggested tieing
VOut to the case. I haven't tried this yet.

Has anyone else used this circuit? any suggestions? I'd also appreciate any
help with the math (resistance values) if possible.
 
F

Fred Bloggs

Jan 1, 1970
0
Bill said:
I built a temperature sensor based on the LM317 current loop in the LM34
docs. I went this route, as the sensor is about 100 feet from the input
device. But the temp is jumping up and down by a couple of degrees all the
time (noise?). I went with the current loop due to the length of the wire to
the sensor/317. The one note I did read in the LM317 docs suggested tieing
VOut to the case. I haven't tried this yet.

Has anyone else used this circuit? any suggestions? I'd also appreciate any
help with the math (resistance values) if possible.

The diagram does not show it, but you will have to bypass that 317
supply with something like 0.1uF minimum. That application circuit is
out of spec for the 317 anyway- not enough minimum current worst case
for regulation. There are other diagrams on the LM34 datasheet
-immediately after the specification pages, that show how to set it up
for remote sensing through twisted pair cable- these are also
temperature dependent current sources that overcome line length voltage
drops and they are simpler than the 4-20mA loop circuit.
 
B

Brian

Jan 1, 1970
0
Bill Stock said:
I built a temperature sensor based on the LM317 current loop in the LM34
docs. I went this route, as the sensor is about 100 feet from the input
device. But the temp is jumping up and down by a couple of degrees all the
time (noise?). I went with the current loop due to the length of the wire to
the sensor/317. The one note I did read in the LM317 docs suggested tieing
VOut to the case. I haven't tried this yet.

Has anyone else used this circuit? any suggestions? I'd also appreciate any
help with the math (resistance values) if possible.

In the data sheet suggestions, it has a de-coupling circuit. Have you
tried this? If you are not sure how to add this to your circuit, I made a
drawing on how to do it. You can see it at
http://www.fncwired.com/Temperature/
Hope this helps,
Brian
 
B

Bill Stock

Jan 1, 1970
0
Brian said:
appreciate

In the data sheet suggestions, it has a de-coupling circuit. Have you
tried this? If you are not sure how to add this to your circuit, I made a
drawing on how to do it. You can see it at
http://www.fncwired.com/Temperature/
Hope this helps,
Brian

Thanks Brian, talk about service. :)

My understanding of the (4-20) current loops was that they provided a two
wire interface immune to noise. But if I still have to decouple the sensor,
then I might as well run 3 wire over the 100 feet?
 
B

Brian

Jan 1, 1970
0
Bill Stock said:
Thanks Brian, talk about service. :)

My understanding of the (4-20) current loops was that they provided a two
wire interface immune to noise. But if I still have to decouple the sensor,
then I might as well run 3 wire over the 100 feet?

What has to be done, depends on how noisy the environmet is, it has to
work in. Sometimes, this can only be found by trial and error.
 
T

Terry Given

Jan 1, 1970
0
Bill said:
I built a temperature sensor based on the LM317 current loop in the LM34
docs. I went this route, as the sensor is about 100 feet from the input
device. But the temp is jumping up and down by a couple of degrees all the
time (noise?). I went with the current loop due to the length of the wire to
the sensor/317. The one note I did read in the LM317 docs suggested tieing
VOut to the case. I haven't tried this yet.

Has anyone else used this circuit? any suggestions? I'd also appreciate any
help with the math (resistance values) if possible.

Hi Bill,

Have you used twisted pair for your wires? CAT5 cable is very handy
stuff, and cheap too.

Cheers
Terry
 
B

Bill Stock

Jan 1, 1970
0
Terry Given said:
Hi Bill,

Have you used twisted pair for your wires? CAT5 cable is very handy
stuff, and cheap too.

Cheers
Terry

I ran 4-wire for a serial device, but it got RMA'd. So I'm using that for my
LM317/LM34, while the serial device get's fixed up. I realize that this is
not the ideal wire, but I didn't plan on this problem. I guess I should
ground the two unused wires in the mean time.
 
F

Fred Bloggs

Jan 1, 1970
0
Brian said:
In the data sheet suggestions, it has a de-coupling circuit. Have you
tried this? If you are not sure how to add this to your circuit, I made a
drawing on how to do it. You can see it at
http://www.fncwired.com/Temperature/
Hope this helps,
Brian

The circuit output is not the terminal labeled OUT on the datasheet
circuit, that OUT refers to the LM34 OUT terminal. The circuit output is
the total current drawn from the 5-30V power supply and not a voltage.
 
T

Terry Given

Jan 1, 1970
0
Bill said:
wire to



I ran 4-wire for a serial device, but it got RMA'd. So I'm using that for my
LM317/LM34, while the serial device get's fixed up. I realize that this is
not the ideal wire, but I didn't plan on this problem. I guess I should
ground the two unused wires in the mean time.

I have no idea what "4-wire" is. Use twisted pair, and yes do ground
unused wires

Cheers
Terry
 
B

Brian

Jan 1, 1970
0
Bill Stock said:
I built a temperature sensor based on the LM317 current loop in the LM34
docs. I went this route, as the sensor is about 100 feet from the input
device. But the temp is jumping up and down by a couple of degrees all the
time (noise?). I went with the current loop due to the length of the wire to
the sensor/317. The one note I did read in the LM317 docs suggested tieing
VOut to the case. I haven't tried this yet.

Has anyone else used this circuit? any suggestions? I'd also appreciate any
help with the math (resistance values) if possible.

Since this is a relatively slow dc signal you are looking at, probably all
you need is a rc decoupling network at the receiving end.
Something like a 2K resistor and a 4.7 uf capacitor.
Brian
 
Top