What effect will extreme temperatures have on characteristic impedance of a wire?
My scenario is this: radio equipment controlling an overhead crane inside a steel mill. Specifically it is used for transporting melting pots. With the current, unfortunate installation, the antenna and antenna coax cable are sitting exposed just above the melting pot where it can get many hundred degrees hot.
The coax is a regular RG58 50Ω with copper/tin wire, copper/tin shield inside polyurethane (PUR). These are rated for +80°C but the temperature is way out of specification.
My question is: what will happen with this coax wire in terms of characteristic impedance when the temperature just keeps rising? I'm assuming that the metal will expand, but will this have an affect on characteristic impedance and therefore also the RF signal requiring an ideal 50Ω?
(As a bonus, there's also a massive magnetic field from the melting process which I assume will get picked up by all metal parts, but lets keep it at one question per post.)
2 answers
RG58 cables come in different variations, even with Teflon dielectric suitable for 200°C operations. Your cable, being suitable for up to 80°C, seems to be one of those with a PVC jacket and Polyethylene inner insulator, like this.
It's bad news for several reasons, and Olin explained a few. I'll add that a problem you could have is not only the measured environmental temperature, but also radiated heat from molten steel.
In fact steel fusion temperature is at least 1300°C and can reach over 1500°C, depending on the alloy composition. During melting and processing it can go well over that temperature.
In that state the metal emits like a black body with an emissivity of about 0.4:
...
and its black-body radiation spectrum has a peak in near infrared around 1.6μm according to online calculators, for a temperature of around 1800K (~1500°C). For example from here:
Note that tiny number termed "intensity" (its precise name is radiant emittance): $5.6 \cdot 10^5 \; W/m^2 = 560 \; kW/m^2$ Multiply that by the surface emissivity of 0.4 gives about 230kW/m^2 of radiant power emitted by that surface!!!
It decreases as the square of the distance and you must multiply it by the "receiving" surface dimensions, but at (say) 10m it's still a hell of a lot of power. You can literally roast some meat with that!
So, if the cable is installed in line-of-sight of the molten steel surface, it can absorb a great deal of IR radiation that can heat it well beyond the measured ambient temperature (especially if its jacket is black, as it usually is).
Even before you get heat damage, you can have all kinds of deformations that can vary depending on the crane position and the way that the molten steel surface "illuminates" the cable.
A simple experiment to see if the radiated heat can be the cause of your problem is to wrap the cable with common aluminium [1] foil. That would reflect quite a lot of the incoming radiated power. If you measure any difference in the comm-link quality, then you will have an important clue.
-
no, I won't say aluminum! :-) ↩︎
0 comment threads
The expansion of the conductors is the least of your problems.
The real issue is the effect of excessive temperature on the insulators. The plastic in the cable has three electrical functions:
- To insulate the two separate conductors.
- To provide a specific constant dielectric between the conductors. This is necessary to make the transmission line impedance fixed and predictable.
- To hold the conductors at a fixed geometry relative to each other. This is also necessary to make the transmission line impedance fixed and predictable.
The cable is rated to a maximum temperature for a reason. The manufacturer isn't guaranteeing the cable's properties above 80°C, which probably means they know the plastic will start to loose necessary properties above that point. Most likely the impedance will suffer first as the dielectric constant changes. At a bit higher temperature, the mechanical properties of the plastic change to the point where the conductors might touch, resulting in a short.
Your question really comes down to "What happens when the maximum ratings are violated?". I think you already know the answer.
In this case it seems that unreliable operation of the crane not only has significant economic cost, but also serious safety problems. You really need to get this fixed now! A planned shutdown for a few hours is way better than a crane with a pot of molten steel suddenly having a mind of its own in the middle of a production run.
One advantage is that your transmitter and receiver can be fairly close to each other, so you should get good signal to noise ratio with relatively low power. You probably also have some distance before the signal gets off the premises where anyone else would care.
There must be a better place to put the antenna that doesn't require running 80°C coax right above molten steel. Where are the transceiver electronics now? Those must be in a cooler place. Why can't the antenna be near there, with maybe only a metal whip sticking out that can take much higher temperatures than polyurethane? With good placement of the fixed transceiver, there should be enough coupling even with less than optimal moving antenna placement.
As for the magnetic fields from the electric heaters, that should not be too much of an issue. The frequency should be much lower than whatever you are using for communication. Also, one of the advantages of coax cable is that all external magnetic and electric fields are common mode. Between the very different frequency and the fact that the interference is common mode, the external magnetic fields should not be that hard to deal with.
they poked a hole in the cabinet and placed the antenna outside, connected with an ordinary RG58
Why not have only a metal whip antenna poking out of the cabinet? Maybe the cabinet location is not optimal for an antenna, but maybe you can compensate by clever placement of the other antenna. Between that, the right frequency, the right power level, and the right protocol, this really should be doable. It sounds like your data rates are very low, which should also allow for better overall signal to noise ratio.
Looking back I'm confused now about which antenna is moving. I originally thought the problem was getting data to/from the moving crane, but you say the electronics are in a cabinet. Is the problem that the operator is moving around? If so, surely that could be dealt with by putting the fixed antenna and electronics in a cool part of the room.
0 comment threads