RG174 vs RG316: Loss, Temperature and Crimping Guide

RG174 and RG316 look almost the same: two thin 50-ohm coax cables, both about as thick as a shoelace, both used for antenna pigtails and short jumpers. Put them side by side and the only obvious difference is colour — RG174 is usually black PVC, RG316 is usually amber or tan and slightly translucent.

The difference that matters is inside. The plastics and the plating are not the same, and that changes how much heat the cable survives, how tightly it can bend, how it behaves above 2 GHz, how it must be crimped, and how much it costs.

Short version:

• Short indoor pigtail, low power, cost matters → RG174. It is 4–6 times cheaper per metre and, below 1 GHz, often the lower-loss cable of the two.
• Heat, engine bay, UV, oil, repeated bending, bench/test use, or work above ~2 GHz → RG316. PTFE and FEP give it a -55 °C to +200 °C rating and much better soldering tolerance.
• Neither one can meet your loss budget → stop forcing a mini cable. Move up to RG58, LMR-100 or LMR-200.

What Is Actually Inside a 50-Ohm Mini Coax

Every coax cable in this class has four layers, from the middle outward. A plain-language walkthrough makes the RG174/RG316 differences easy to follow.

1. Center conductor — the wire that carries the signal

Both cables use a 7-strand, roughly 26 AWG bundle, so they feel similar when you strip them.

• RG174: 7 × 0.16 mm strands, bare copper or copper-clad steel (CCS) depending on the maker.
• RG316: 7 × 0.175 mm strands, silver-plated copper-clad steel. Silver is the best electrical conductor available, and at high frequency the current only travels in the outermost skin of the wire — so a thin plating on the surface is where most of the current actually flows.

Why steel at all? A pure copper strand this thin snaps easily. A steel core makes the cable survive pulling and repeated flexing, at the cost of higher resistance at low frequency. This trade-off is behind the counter-intuitive loss behaviour explained further down.

2. Dielectric — the white or clear plastic around the conductor

This is not just packing material. It sets the impedance, the signal speed, and how much of the signal is turned into heat inside the plastic.

• RG174 uses solid PE (polyethylene), the same family of plastic as a milk bottle. Electrically it is good; thermally it is the weak point, because PE melts at roughly 105–135 °C.
• RG316 uses solid PTFE (Teflon). PTFE melts at about 327 °C, and its RF losses stay extremely low and stable across temperature and frequency — NIST measurements put its dissipation factor below 2 × 10⁻⁴ over a very wide temperature range.

Two terms in everyday language:

• Dielectric constant (εr): how much the plastic slows the signal down. Walking through water is slower than walking through air. PE is about 2.3, PTFE about 2.1 — both verifiable on a neutral materials database such as MatWeb. Lower εr means a faster signal, which is why RG316 has a higher velocity of propagation.
• Loss tangent (tan δ): how much signal the plastic itself eats and turns into heat. Think of clean glass versus frosted glass in front of a lamp. PTFE is the clean glass, with tan δ around 0.0002.

3. Braided shield — the metal mesh

The braid does two jobs: it is the return path for the signal, and it keeps outside noise out. Coverage is typically 90–96 % on both cables.

• RG174: tinned or bare copper braid.
• RG316: silver-plated copper braid, which resists oxidation at high temperature and keeps resistance low at microwave frequencies.

Standard RG316/U has one braid layer. That matters in noisy environments — see the double-shield section below.

4. Jacket — the outer skin you touch

• RG174: black PVC. Soft, slightly rubbery, easy to strip, cheap. PVC hardens in the cold and softens near heat, and ordinary PVC is not built for long-term UV exposure.
• RG316: amber/tan FEP, semi-translucent, slick to the touch and noticeably stiffer. FEP belongs to the same fluoropolymer family as PTFE, so it resists heat, oil, most chemicals and UV far better than PVC.

A practical field test: if the jacket is glossy amber and slides between your fingers, it is a fluoropolymer cable. If it is matte black and grips your skin slightly, it is PVC.

Baseline specifications side by side

Both military designations can be looked up by part number on DLA QuickSearch, and background on impedance, Vp and coax construction in general is covered by the neutral reference Coaxial cable — Wikipedia. For a specific build, always confirm figures against the exact part datasheet, because an “RG” number describes a construction family, not one guaranteed set of values.

Loss by Frequency, and the Result That Surprises Buyers

Attenuation is signal loss along the cable — water pressure dropping along a long garden hose. It is quoted in dB per 100 m and it always rises with frequency.

A quick way to read dB:

• 3 dB = half the power is gone.
• 6 dB ≈ only 25 % left.
• 10 dB = only 10 % left.

Read the table twice. Below about 1 GHz the cheaper cable is also the lower-loss cable. Only at 2.4 GHz and above does RG316 catch up and pull ahead.

Why RG174 can beat RG316 below 1 GHz

This contradicts the common assumption that “RG316 is the better cable, so it must have less loss”. Two mechanisms explain it.

Below ~1 GHz, conductor resistance dominates. RF current does not fill the whole wire; it crowds into a thin surface layer, like traffic piling into the outside lane. That depth is called skin depth — see Skin effect — Wikipedia. In copper it is roughly:

• 6.5 µm at 100 MHz
• 2.1 µm at 1 GHz
• 1.3 µm at 2.4 GHz
• 0.9 µm at 5.8 GHz

Silver plating on a copper-clad steel strand is only a couple of micrometres thick. At 100–400 MHz the current runs several micrometres deep, so it passes through the plating and into the copper cladding and the steel-influenced region underneath. Steel is far more resistive than copper and magnetic as well, which concentrates the current even further. The plating brings almost no benefit here, while RG316’s smaller conductor and dielectric geometry still cost some loss. Result: at 100 MHz the numbers are 27.56 dB/100 m for RG174 against 36.09 dB/100 m for RG316.

Above ~2 GHz, the surface and the plastic take over. Skin depth shrinks to about 1 µm, so essentially all the current rides in the silver layer, where resistance is at its lowest. At the same time dielectric loss — the plastic quietly heating up — grows in proportion to frequency, and PTFE’s tan δ of roughly 0.0002 is far better behaved than PE. RG316’s curve therefore rises more slowly and overtakes RG174 around 2.4 GHz, with a clearer advantage above it.

Two honesty notes when comparing datasheets:

• Many RG316 sheets publish maximum attenuation while many RG174 sheets publish nominal values, which exaggerates the low-frequency gap.
• RG174 is normally specified only to 3 GHz. Using it at 5.8 GHz means running outside its specified band, where sample-to-sample spread widens.

Calculate your own loss budget

Use this for any RF path:

Total loss (dB) = (cable length in m × attenuation in dB/m) + connector losses + adapter losses

Working values for the extra parts:

• Well-made SMA, BNC or MMCX termination: 0.1–0.2 dB each
• U.FL/IPEX micro connector: 0.2–0.3 dB each
• Each added adapter (SMA-to-RP-SMA, SMA-to-BNC): 0.2–0.3 dB each

Example 1 — 2 m Wi-Fi antenna extension for a PC, 2.4 GHz, RG174:

2 m × 1.50 dB/m = 3.0 dB, plus two SMA ends ≈ 0.3 dB → ≈ 3.3 dB. More than half the power is gone before the antenna. Acceptable for a desktop link, poor practice for a long-range one.

Example 2 — 30 cm indoor GPS pigtail, 1.575 GHz, RG174:

0.3 m × 1.18 dB/m = 0.35 dB, plus two ends ≈ 0.3 dB → ≈ 0.65 dB. Comfortable for an active antenna with a built-in LNA.

Example 3 — 20 cm FPV video lead, 5.8 GHz, RG316:

0.2 m × 2.15 dB/m = 0.43 dB, plus an SMA and a U.FL ≈ 0.4 dB → ≈ 0.85 dB. The two small connectors cost about as much as the cable itself — at these frequencies, termination quality matters as much as cable choice.

Rule of thumb: if the connectors contribute more loss than the cable, stop optimising the cable and shorten the path or remove adapters instead.

Inside a crowded enclosure, bend radius and kink resistance often decide the design before attenuation does.

Mechanical Behaviour and Routing in Tight Enclosures

Soft is not the same as bend-friendly. RG174 feels more flexible because the PVC jacket and PE core are soft, and that same softness is why a sharp kink squashes the dielectric and leaves a permanent impedance dent. RG316 pushes back when you bend it, so it resists knotting and returns to shape — and its published static bend radius is roughly half that of RG174.

Practical routing guidance:

• Long, loose runs behind a panel or inside a plastic box: RG174 is easier to handle and cheaper.
• Tight loops, corners, or a cable that gets bent every time a lid closes: RG316 keeps impedance more stable and survives repeated flexing better.
• Any rotating or hinged joint (robot arm, folding antenna mount, drone canopy): assume thousands of bend cycles, choose RG316, add strain relief, and secure the cable so the bend never happens right at the connector.
• Confirm the exact minimum bend radius on the datasheet of the part you buy. Manufacturers publish separate static, repeated and dynamic values, and they differ substantially.

Temperature, Sunlight, Oil and Chemicals

One limit worth stating plainly: a temperature rating is not a weatherproofing rating. Neither cable is sealed against water ingress at the connector. For outdoor installations, specify the exact jacket’s UV rating and seal the connector interface with self-amalgamating tape or a boot.

The 0.29 mm Crimp Trap: Never Share Ferrules Between the Two

The outer diameters differ by only 0.29 mm — 2.79 mm for RG174 against 2.50 mm for RG316. That gap is invisible to the eye and a common cause of failed assemblies, because a crimp ferrule and its hex die are sized for one cable, not both.

Crimping RG316 with an RG174 die (3.25 mm on a 2.50 mm cable): the ferrule never closes properly. The braid is only lightly pinched, so the ground contact is intermittent and pull-out strength drops well below the 40 N target. The assembly may pass a bench test and then fail after a few weeks of vibration.

Crimping RG174 with an RG316 ferrule (2.67 mm on a 2.79 mm cable): the opposite problem, and the worse one electrically. The tool crushes the soft PE dielectric — dielectric pinching. Squeezing the insulation reduces the spacing between the inner conductor and the shield, and in a coax that spacing ratio sets the impedance, so the local value drops well below 50 ohm.

What that costs in RF terms: a section sitting at 40 ohm reflects about 11 % of the arriving voltage, i.e. a local VSWR of roughly 1.25 on its own. VSWR simply describes how much signal bounces back instead of going forward — like an echo in a tunnel, where 1.0 means no echo at all. One dent is survivable; a crushed dent plus a mediocre connector plus an adapter add up, and because the reflections combine differently at every frequency, a 2.4 GHz or 5.8 GHz assembly can easily land above 1.5 and lose repeatability across a production batch.

Practical rules:

• Order the connector, ferrule and die set for the exact cable family, and label the tooling.
• Keep RG174 and RG316 connector kits in separate boxes; the parts look nearly identical.
• Check the first article: measure retention force (≥ 40 N), then check return loss on a VNA if the design works above 1 GHz.
• If a batch shows random VSWR outliers, cut one sample open and inspect the dielectric under the ferrule before blaming the antenna.

Soldering Heat: Why Prototypes Fail on RG174

This one bites hardest during hardware development, when engineers hand-solder their own test leads.

• PE (RG174) melts at 105–135 °C. A soldering iron tip sits at 320–380 °C. Hold the SMA center pin for more than about two seconds and the PE dielectric retracts from the heat — shrinkback. The dielectric pulls away, the braid strands relax forward, and one stray strand touching the center pin shorts the RF path. Sometimes it is a dead short you find immediately; sometimes it is an intermittent contact that only appears under vibration.
• PTFE (RG316) melts near 327 °C and will not shrink back during a normal soldering cycle. It tolerates a slow hand, a second attempt, and the extra heat needed on a board-edge joint.

Recommendations:

• For hand-built test jigs, VNA calibration leads, board-to-board direct soldering, and anything reworked more than once, use RG316.
• If RG174 must be soldered, pre-tin the pin, keep contact under two seconds, use a heatsink clip behind the joint, then inspect: any visible gap between the dielectric and the connector body means the assembly should be scrapped.
• For volume production on RG174, prefer crimped assemblies over hand-soldered ones.

Single Shield vs Double Shield: A Naming Trap

Many buyers assume RG316 is double shielded. Standard RG316/U built to MIL-DTL-17/113 has a single braid. Typical shielding effectiveness is around 45 dB — enough for ordinary environments, and broadly comparable to what RG174’s single braid delivers.

If the cable runs next to brushless motors, ESCs, switching power supplies or inverters, a single braid may let noise couple in.

• What to specify instead: RG316-DS or RD316, the double-braid construction covered by M17/152-00001. Two braid layers typically push shielding effectiveness past 80 dB.
• What changes: outer diameter grows to roughly 2.9–3.0 mm, the cable becomes stiffer, and the ferrule and hex die change again. Do not reuse RG316 tooling on RD316.
• How to order without ambiguity: write “double-shielded RG316-DS / RD316 per M17/152-00001” on the purchase order. “RG316, double shielded” alone gets interpreted differently by different suppliers.

Military specification sheets can be looked up by part number on DLA QuickSearch, the neutral source for what each designation actually requires.

Typical Applications and Real Selection Guidance

Indoor, office and consumer builds — RG174 is the cost-effective pick

Good fits:

• 1–2 m Wi-Fi antenna extension for a desktop PC or router, accepting about 3 dB at 2.4 GHz over 2 m
• Indoor GPS receiver pigtails, especially with an active antenna
• 433/868/915 MHz LoRa modules and utility meters inside a plastic enclosure
• Internal wiring for RF modules, set-top boxes and test fixtures used at room temperature
• Cost-sensitive assemblies in volume, where 0.35–0.50 USD/m against 1.80–3.00 USD/m is a real line item

Why it works: the environment is controlled, power is low, the run is short, and below 1 GHz RG174 actually has the lower attenuation. Paying 4–6 times more for RG316 buys nothing in this scenario.

Harsh and high-performance environments — RG316 is the safer choice

Good fits:

• FPV drones and 5.8 GHz video links: lower loss at 5.8 GHz than typical low-grade RG174, plus kink resistance inside a cramped frame. Pair with RD316 if the ESC noise floor is a problem.
• Outdoor base-station and rooftop antenna jumpers: UV and weather exposure rule out standard PVC.
• Industrial robot joints and cable carriers: repeated flexing needs the springy FEP jacket and the 15 mm bend radius.
• Automotive electronics: engine-bay heat, oil mist and cleaning chemicals are all outside PVC’s comfort zone.
• Lab, bench and calibration leads: frequently handled, frequently re-soldered, sometimes routed near hot equipment.

When mini coax is no longer the right answer

Both cables are miniature, so both are lossy by design. Stop stretching them and move to a thicker family when any of these is true:

• Length above about 3 m at 2.4 GHz or higher. At 1.5 dB/m, a 5 m RG174 run costs about 7.5 dB — roughly 80 % of the power. LMR-200-class cable is around three times lower in loss at the same frequency.
• Continuous transmit power above a few tens of watts, or any run sitting in an enclosed hot space at high duty cycle.
• A loss budget tighter than about 1 dB for a run longer than half a metre.
• Outdoor runs longer than 1–2 m, where both a larger conductor and a jacket built for sunlight or burial help.

Upgrade path, in rough order of size and performance: RG174/RG316 → LMR-100 (similar diameter, lower loss, but a foam dielectric that is easier to crush) → RG58 → LMR-200. Low-loss cables trade away flexibility and need their own connector families, so check the bend radius before redesigning the enclosure.

Five-Step Selection Checklist

1. Check the frequency. Below 1 GHz, RG174 is usually both cheaper and lower in loss. At 2.4 GHz and above, RG316 pulls ahead.
2. Calculate the length budget. Length × dB/m + connectors + adapters. If the total exceeds your link margin, shorten the run or move to a thicker cable rather than hoping for a better cable of the same size.
3. Check the temperature and environment. Anything above 70 °C, below -20 °C, in sunlight, or near oil and chemicals points to RG316. A sealed room-temperature box is fine for RG174.
4. Measure the space. Compare your tightest bend against 28 mm (RG174) and 15 mm (RG316), and count how many times per year that bend will move.
5. Compare cost against risk. RG316 costs 4–6 times more per metre. In prototypes and harsh environments that is cheap insurance; in a high-volume indoor product it is money spent for no measurable gain.

One extra step for assemblies: confirm the connector family, the ferrule and hex die size, and whether you need single or double shielding — before the first sample is built.

What To Send a Supplier

The more precisely the RF and mechanical conditions are stated, the less likely a supplier is to quote a cable that fits the housing but misses the link budget.

Common Questions

Is RG174 the same as RG316?

No. Both are 50-ohm miniature coax families with a 7-strand, ~26 AWG conductor, but RG174 uses a PE dielectric with a PVC jacket, while RG316 uses PTFE with an FEP jacket plus a silver-plated conductor and braid. The temperature ratings differ by a wide margin: -20 °C to +70 °C against -55 °C to +200 °C.

Does RG316 always have lower loss than RG174?

No, and this is the most common misconception. At 100 MHz typical figures are 27.56 dB/100 m for RG174 against 36.09 dB/100 m for RG316. RG316 only overtakes around 2.4 GHz, once skin depth shrinks enough for the silver plating to pay off and PTFE’s very low dielectric loss starts to dominate.

Can I use the same crimp tool for both?

No. RG174 normally uses a .128″ (3.25 mm) hex die and RG316 uses .105″ (2.67 mm) or .120″. Mixing them either leaves the braid loosely held, with retention force below 40 N, or crushes the dielectric and drops local impedance, which shows up as poor VSWR at 2.4 GHz and above.

Is RG316 double shielded?

Standard RG316/U per MIL-DTL-17/113 has a single braid, with shielding effectiveness around 45 dB. For strong EMI environments, specify RG316-DS or RD316 per M17/152-00001, which uses two braids and typically exceeds 80 dB. Expect a larger diameter, a stiffer cable and different crimp tooling.

Which cable should be used for a 5.8 GHz FPV video link?

RG316, in short lengths. At 5.8 GHz it runs around 215 dB/100 m, so a 20 cm lead loses roughly 0.4 dB in the cable plus 0.3–0.4 dB in the connectors. It also handles the heat and vibration inside a drone frame better than PVC.

Is RG174 good for GPS antenna cables?

Yes, for short indoor pigtails. At 1.575 GHz a 30 cm run plus two connectors costs well under 1 dB, which an active antenna with a built-in LNA absorbs easily. Keep it short, avoid sharp kinks, and check the total loss if the receiver has a tight link budget.

How long can an RG174 cable be?

There is no universal limit; it depends on frequency and link margin. A workable guide: up to 2–3 m at 433/915 MHz, about 1–2 m at 2.4 GHz, and well under 1 m at 5.8 GHz. Beyond that, calculate the budget and consider LMR-100, RG58 or LMR-200.

Why does RG174 melt so easily when soldering?

Its PE dielectric melts at 105–135 °C, far below a soldering iron tip. More than about two seconds of heat on the center pin causes shrinkback, and the braid can then touch the center conductor and short the cable. The PTFE in RG316 melts near 327 °C, which is why it is the practical choice for hand-built leads.

Conclusion

RG174 and RG316 are not a good/better pair. RG174 is the sensible default for short, low-power, room-temperature pigtails below 1 GHz, where it is both cheaper and slightly lower in loss. RG316 earns its 4–6× price premium through its -55 °C to +200 °C rating, chemical and UV resistance, tighter bend radius, forgiving soldering behaviour, and better performance above 2.4 GHz. When neither can meet the loss budget, the right move is a thicker cable, not a longer mini one.

For a custom RF cable assembly quote, send the frequency band, cable length, connector pair, impedance, target insertion loss and VSWR, operating environment, shielding requirement and quantity. Browse our RG174 coaxial cable and RG316 coaxial cable. For termination choices, see the antenna connector types guide; for a thicker 50-ohm comparison, see RG58 vs RG400.