Both carry high-frequency signals over a copper cable with a connector on each end, and they are not the same technology. Coaxial cable carries a single signal and a return on a shield; twinax carries a differential pair. One is 75 or 50 ohm and unbalanced, the other is 100 ohm and balanced.
Confusing them matters because they serve different jobs and neither substitutes for the other.
Quick answer
| Coax | Twinax | |
|---|---|---|
| Signal | Single-ended, referenced to the shield | Differential, on a balanced pair |
| Impedance | 75 ohm or 50 ohm | 100 ohm |
| Typical use | Video, CCTV, satellite, antenna, RF test | Short high-speed data links in a rack |
| Connector | F, BNC, RCA, N | Fixed assemblies with the transceiver built in |
| Reach | Governed by loss at the frequency in use | Short, by design |
| Where it ends | Fiber for very long runs | Optics and fiber beyond its reach |
What a DAC assembly is
A direct attach cable takes the transceivers that would normally be separate pluggable modules and builds them into the cable ends. The result is a fixed assembly: connector, cable, connector, with the electronics sealed in.
Three consequences follow.
It is cheaper than two pluggable modules and a length of fiber, for the distances it covers.
It uses less power, because the transceiver circuitry in a short passive assembly is simpler than an optical module’s.
It is less flexible. The assembly is one unit at a fixed length; if it fails, the whole thing is replaced, and different switch vendors have their own compatibility requirements.
That combination makes it the standard choice for connecting equipment within a rack or between adjacent racks — the distances where fiber would be over-specified and where a pluggable module at each end would cost more than the link is worth.
Why data links use twinax rather than coax
A high-speed serial data link carries a differential signal: two conductors, with the information represented by the voltage difference between them. Interference that arrives on both conductors equally largely cancels at the receiver, which is what makes differential signaling reliable at high rates and low signal levels.
Coaxial cable is built for the opposite arrangement. The center conductor carries the signal and the shield is the return, and the signal is referenced to that shield. It is the right structure for video and RF, and the wrong one for a differential data link.
The impedance values reflect the difference. Coax is standardized at 75 or 50 ohm depending on the application; twinax is 100 ohm to match the differential interface it serves.
Where each one belongs
The dividing line is the signal, not the speed.
| Application | Cable |
|---|---|
| Television, satellite, cable service | 75 ohm coax |
| CCTV camera, analog or HD analog | 75 ohm coax |
| Coaxial digital audio | 75 ohm coax |
| Antenna feed, two-way radio, test equipment | 50 ohm coax |
| Switch to switch in a rack | Twinax DAC, or fiber with modules |
| Server to top-of-rack switch | Twinax DAC |
| Anything past the DAC’s reach | Fiber with transceivers |
The two worlds do meet in a data center, where cameras or video distribution might be installed alongside network equipment. They do not share cable.
Passive, active and optical
Three variants of the same idea, and the difference is how far the link goes.
Passive DAC has no active electronics in the cable. It relies on the host’s transceivers and the cable’s quality. It is the cheapest and shortest-reaching option.
Active DAC includes signal conditioning in the assembly, which extends the reach over copper.
Active optical cable replaces the copper with fiber inside the assembly, keeping the same electrical connectors at each end. It reaches much further and is what you use when a copper assembly will not span the distance.
That progression is the same shape as the coax world’s answer to distance: coax up to a limit, then fiber with converters. The AOC side of that is covered in the site’s optical cable material, and the principle is identical — the copper has a reach, and past it the signal changes medium.
The practical selection points
Four things decide a DAC purchase, and none of them is the cable’s construction quality in the abstract.
The reach. Passive assemblies have a shorter working range than active, and the vendor’s compatibility list will state it.
Vendor compatibility. Equipment makers qualify specific assemblies. An assembly that works between one vendor’s switches may not work between another’s, and the programming of the connectors is part of why.
Data rate. The assembly has to match the interface speed.
Locking and strain relief. These cables are handled during moves and changes. A connector that does not latch properly is a link that drops when a rack is serviced.
Scenarios
Two switches in the same rack. A passive DAC, short, on the vendor’s compatibility list.
Adjacent racks, several meters apart. An active DAC, or fiber if the length is at the edge of what the vendor supports.
Twenty meters to a switch at the end of the row. Past the copper assembly’s reach. Optical assembly or fiber with modules.
A CCTV system and a network core in the same room. Coax to the cameras, twinax between the switches. Two cable stocks, clearly labeled and stored separately.
FAQ
Is twinax the same as coax? No. Twinax is a balanced pair at 100 ohm; coax is a single-ended cable at 75 or 50 ohm.
Can I use coax for a 10G data link? Not in the normal way. The interface expects a differential pair, and coax is a single-ended structure.
What is a DAC cable used for? Short, high-speed links between network equipment, typically within or between adjacent racks.
How far can a DAC run? It depends on whether the assembly is passive or active, and on the equipment vendor’s qualification. Copper assemblies are for short reaches by design.
What is the difference between a DAC and an AOC? Both are fixed assemblies with the electronics built in. A DAC uses copper and reaches a shorter distance; an AOC uses fiber and reaches further.
Why do DAC cables need to be vendor-compatible? Because the equipment vendors qualify specific assemblies, and the connector programming and signal characteristics are part of that qualification.