Differential signaling vs Single-ended data rates

What is the main reason of higher data rates in differential signaling as compare to single ended signaling?

Tline appears resistive plus delay while single ended appears capacitive to the driver (and this ruins risetime @ RX).

dick_freebird said:

Tline appears resistive plus delay while single ended appears capacitive to the driver (and this ruins risetime @ RX).

Click to expand...

I don't think that's true. There's capacitance in the line regardless. Differential signaling is faster because, generally, the voltage levels are lower, e.g. LVDS, and because the required bandwidth for differential signaling is less than comparable single-ended signals.

Hi

barry said:

I don't think that's true. There's capacitance in the line regardless.

Click to expand...

For low frequency it looks capacitive.
But at the signal frequency of interest (signal frequency) it should be close to resistive.
--> A 110 Ohms signal pair ... looks like a 110 Ohms resistive load to the driver. (thus the name)

barry said:

Differential signaling is faster because, generally, the voltage levels are lower,

Click to expand...

to add some details:
* the signal speed on the line does not depend on signal level.
* but the rise time may be smaller. Resulting in getting more "information per time" on the line.
(With identical rise rate a 0 to 400mV step would be 5 times faster than a 0 to 2000mV step.)

***
Differential pair in best case cause no GND currents, thus no GND bounce and thus less channel crosstalk and better noise margin.

***
It depends on the application. Do we talk about on board signaling, short distance wires, long distance?
Single channel or multiple channels?
What data rate?
What type of cable? coax for single endend signals?

Also to consider: EMI / EMC, signal loss, distortion, cable production tolerances, cable cost, cable stiffness (size, diameter...)

Klaus

As we all know, impedance matched high speed links can be also implement single ended, e.g. 50 ohm. Cancelation of return currents, immunity against common mode noise and reduction of EMI are however a good reason why all recent high speed interface standards (PCIe, SATA, USB, Ethernet) use differential signaling.

Thank you all for your contribution.

After some research, I found that:

Because of Differntial signaling improved resistance to noise, differential signals can use lower voltages and still maintain adequate SNR (signal-to-noise ratio). Also, the SNR of differential signaling is automatically increased by a factor of two relative to an equivalent single-ended implementation, because the dynamic range at the differential receiver is twice as high as the dynamic range of each signal within the differential pair.

Differential signals usually have a much smaller signal swing (example LVDS 100-350mV) then single-ended signals. The faster you can switch between voltage high and voltage low, the faster the data rate. The primary advantage of using differential signaling is the fact that it reduces the voltage amplitude that you would see in a single-ended signal. A smaller amplitude leads to shorter rise/fall times, which translates to a faster data rate.

Below are some resources:

Ground considerations in differential signaling :



Regards, Dana.