THVD1450: Lowering Emissions Design Tips and Tricks

Hi Clemens,

That reply was lightning fast...

[1] Would differential signaling "cancel" each other out, and reduce emissions?

[2] Would a common-mode choke be a way to trim THVD1450 so it would work at 10Mbps while reducing emissions? I don't know too much about common-mode chokes except they are inductively coupled together, allowing differential signals to pass, but not common-mode...so this might not be a good option to reduce emissions?

[3] Are devices that operate at lower voltages, with the same speed, have inherently less emissions? Considering if this is a topic that could be used to promote lower voltage options...(5V being used now)

[4] Would the easiest approach to slow down edges be to add a series resistors to each line, and a diff. cap? What about a cap from each line to GND? (Common-mode caps?)

Regards,

Darren

1. Yes, the whole purpose of differential signals (and of twisting the pairs together) is to make the two lines look like a single line with a total current of zero. If the lines/traces are not near enough each other or not balanced (which typically happens near connectors etc.), some noise might escape.

2. A common-mode choke reduces both common-mode emissions and common-mode noise injected from outside. A perfect RS-485 transmitter does not emit common-mode noise, so in practice, common-mode chokes are used mostly to protect against external noise.

3. In theory, yes. In practice, all RS-485 transceivers use the same differential output voltage, so the supply voltage does not matter.

4. Yes, such an RC low-pass filter would be cheapest. I do not think that capacitors to ground make a difference as far as slowing down edges is concerned. (They do make a difference with common-mode noise, but I do not know if they are any better or worse than split termination. I think I would prefer the single differential capacitor so that I am able to control the LPF independently from the common-noise capacitor to GND. But the recommended load  of SAE J1708 uses capacitors to ground, so it probably does not make much of a difference.)

HI Clemens,

I misunderstood the point from my customer.

It's not emissions so much they are concerned about, but the Receiver functionality.

The THVD1450 is more sensitive to input changes, and is more likely to experience issues during EFT/B system testing.

Do you have any thoughts on how to increase the noise immunity for the Receiving implementation?

This may be a pretty basic question, but any pointers you can provide will help me out...

Thanks!

This actually does not make any difference. You can use the same mechanisms to slow down edges at the input.

Darren,

From my experience, you'd better slow down (filter) the input signal instead of processing the signal at the output.