Coaxial Cable Capacitance Change Due to Temperature

[KEggemeyer]

I am looking for information on the amount of change in capacitance a coaxial cable exhibits based on a change in temperature. The dimension of the dielectric will change with temperature, won't the capacitance also change? If it is more than 1% over a 10 C to 80 C temperature change, I'll have to account for it. Any thoughts or direction on where to find information? Thanks!

 

[CDRIVE]

I would think that this data would be found in the manufacturer's data sheet. Are you using the coax unterminated as a capacitive element? I believe a coax line is represented as lumped C, L and R - (including dielectric loss) per foot @ 25 deg C.

Chris

 

[KEggemeyer]

I did find data on capacitance per foot, but no temperature was given. There was also no data on how a temperature change would effect the capacitance. What I'm thinking I can do is take the expansion coefficient of the dielectric, determined the diameter change based on the temperature change, and calculate the capacitance values at both ends of the temperature range. When I do this, the change is around 0.7% for the entire temperature change of 70 °C on a short cable (5-foot). This is acceptable, but I'm unsure that my approach is accurate.

I'm using the coax to hook a capacitive probe to the measurement electronics. The probe only changes 80 pF over the span of measurements, so any change in cable capacitance will make a difference. It's okay for me to add the cable's capacitance to the probes capacitance, it just gets calibrated out---as long as the cable's capacitance doesn't change based on temperture.

 

[davenn]

so what brand of coax is it that you are dealing with ?

Dave

 

[KEggemeyer]

Dave,

I haven't selected a brand, or even a specific cable type. Just need to know if the capacitance changes.

Kevin

 

[CDRIVE]

Is this a long period temp test of the capacitive device: as in a lab (temperature stress) chamber?

Chris

 

[davenn]

what are you using it for that makes it so critical ?

1% seems to be a tolerance that is unlikely to be obtainable in any form of transmission line

Dave

 

[KEggemeyer]

The application is for a capacitive continuous liquid level sensor in a process vessel (a brewery boil kettle) where temperatures range from 10 C to 100 C. The probe will heat up due to the process, and that will heat the cable. The electronics are away from the probe due to the temperature of the environment, but only by a few feet, so the cable is not long (< 5 feet). The cable will have as much capacitance as the probe, but it's not the base capacitance of the cable that is the problem because the liquid level will be calibrated with this capacitance in place (at say 20 C). However, if the capacitance changes by much due to the changing temperature of the cable, it would have to be compensated for or the reading will be off. I had heard that cable capacitance changes with temperature, but I wasn't sure if it was by a little or a lot.

Thank you for your questions and replies to this!

Kevin

 

[davenn]

OK interesting use

I use coax cables in extreme outdoor enviroments, in the RF field, but the max temp would be ~ 45-50C

for us RF techs, a few % change in capacitance isnt going to affect the impedance overly much ( and impedance and dB loss per metre of the cable are our primary concerns)
if the cable changed from 50 to 55 Ohms impedance its no big deal even changing to say 60 Ohms would not have any dire effect on transmission quality/losses

I suspect you are going to have to go to a couple of mil spec cable manufacturers and ask the same questions and give them the temp range you expect the cable to work in

Dave

 

[(*steve*)]

Why not just test it?

 

[CDRIVE]

First comment on the list is all coax are not created equal. There is a vast selection available with varying types of dielectric, conductors, shielding and sheathing. The price per foot is just as great. So, it's doubtful that they all would have the same C vs Temp curve.

In lieu of not having any manufacturer data here's a thought. Seal the end of a length of the coax you're using and measure it under your temperature conditions. Then you can just factor it in.

Personally, I think delta C is going to be almost insignificant unless the coax is el cheapo. I would think that a coax model using a Teflon dielectric would be superior, while foam type dielectric would be inferior. What's your coax type (PN) name. like RG58U etc.?
Mil Spec stuff is usually superior.

Chris

 

[CDRIVE]

I suspect you are going to have to go to a couple of mil spec cable manufacturers and ask the same questions and give them the temp range you expect the cable to work in

Dave

Why not just test it?

Geeeeez, I hit submit and there you both were. Now I feel like an echo!

 

[KEggemeyer]

I'm sure the approach of just testing it is where we'll end up. The issue is that the cable will not heat up with the process vessel in a linear fashion and I wasn't sure how to deal with this lag or the measurement of the cable temperature.

Forums are full of people with great and varied experience! : ) I will take the question to a cable manufacturer to see what they say, but I'm probably overthinking it and will take Steve and CDrive's advice.

Thanks for the help!

Kevin

 

[(*steve*)]

First comment on the list is all coax are not created equal.

Sure. Find a quality cable, and test it. If the results indicate the problem is insignificant, then just specify the use of that cable.

There is no need to worry about something until you have evidence it actually exists.

 

[CDRIVE]

I'm sure the approach of just testing it is where we'll end up. The issue is that the cable will not heat up with the process vessel in a linear fashion and I wasn't sure how to deal with this lag or the measurement of the cable temperature.

Thanks for the help!

Kevin

Does this Cap Meter have a computer interface?

Chris

 

[KEggemeyer]

Chris,

This is something I'm building from scratch for fun. It's a one off, not something I'm looking to produce in quantity. The idea is a relaxation oscillator (comparator with the probe as capacitor plus a few resistors, output to an optocoupler for level shifting to the PLC--very simple) that sends a signal (frequency) to the PLC that controls the brewery (Siemens S7-1214 capable of measuring frequencies up to 100 kHz). The entire idea is a revision of the Nuts and Volts magazine article, "Measure Water Level Without Getting Wet". The probe is entirely different, and so is the oscillator (not a 4060). I've already built the probe out of 0.5-inch diameter stainless covered in 0.125-inch PTFE. It produces a base capacitance of about 100 pF in air and around 280 pF when immersed in water to 24 inches (~80 pF per foot, 1 pF is significant). It actually works better than what I thought it would for being homemade!

I felt the electronics (oscillator) needed to be separated from the probe due to the temperture of the environment over the boil kettle. So I was looking into a cable. However, cables have capacitance and that capacitance, according to what I've read in articles and books on industrial control, changes with temperature. However, none of them ever stated if the change was large or small. Also, I don't know why coaxial cable was recommended, but I'm trying to learn.

Now I'm beginning to think it might be worth investigating two versions of the circuit: 1) temperature capable components that sit immediately on top of the probe (goes to 100 C), and 2) components that are separated by a cable--but not worrying about any capacitance changes in the cable.

I appreciate your thoughts and help on this!

Kevin

 

[sirch]

I may be mising something with this suggestion but assuming you are fixing the cable to the wall of the boiler why not just run two widely separated wires? That way they would have negligable capacitance. Or create something like ladder feeder - again with wide separation - to minimize the cable capacitance.

 

[CDRIVE]

Guys, I searched Kevin's Nuts & Volts article and found the PDF. The preface alone is an enjoyable read. We have quite a few requests for water level circuits, so I found the list of air/water properties (alone) to be worth the read and dendrite stimulating to boot..

http://www.parallax.com/Portals/0/Downloads/docs/cols/nv/vol1/col/nv27.pdf

Chris

 

[KEggemeyer]

Chris,

This one is also very good, though a lot more technical for someone at my level of understanding of these things:

http://www.foxanddole.com/Practical Guide to RF Level Controls.pdf

It's long and contains a lot of information not useful to my application, but page 50 of the PDF, in section 7.4 describes the idea for the circuit. How a capacitive sensor works is described elsewhere, and the construction of the probe and calculations for capacitance are given.

Sirch -- I was concerned about EMI and interference from the motors and large electric heating elements, though I don't know if my concerns are justified.

Kevin

 

[KEggemeyer]

Just a follow up... I contacted Belden and they said that the change in capacitance over the specified temperature range would be "little".

Kevin