1

I have an RGB LED present in an EV charging station. The RGB LED uses a digital serial protocol to receive colour information from the MCU. The cable length between the two is 1.5 m and it's all enclosed within the charging station.

The RGB LED protocol uses a fixed frequency (with a tolerance) to signal a '1' or '0'. The frequency mark/space times are: 0.8/0.45 us for a '1'; 0.4/0.85 us for a '0'.

My questions are:

  • Will this 1.5 m wire cause any EMI/EMC issues?
  • Do we need to take any precautions?

EDIT:1 My calculations

I calculated the frequency of the serial protocol waveform and it is 833 Khz. I calculated its wavelength to be approx. 360 m.

The (wavelength/4) value is 90 m. My maximum length is 1.5 m only. So I believe this wire won't make any issues.

TonyM
  • 22,898
  • 4
  • 39
  • 64
Hari
  • 2,207
  • 14
  • 29
  • Any reason to PWM a LED at 800kHz when 5kHz would work fine? – bobflux Mar 14 '24 at 08:57
  • 2
    @bobflux it is a LED with digital bus where data is transmitted at 800 kHz and 5 kHz would not work. – Justme Mar 14 '24 at 08:59
  • 2
    Ah so it's not PWM, it's a digital control signal then. – bobflux Mar 14 '24 at 09:17
  • 2
    @bobflux, yep, you're right, it's bad terminology from the OP. What they're calling PWM is the comms protocol. It does indeed vary the pulse width of a fixed frequency but only to signal two binary levels. And the waveform is to communicate data, not deliver an average voltage or power like PWM. I'll rephrase their question. – TonyM Mar 14 '24 at 10:02
  • if you don't mind could you please explain the difference between digital serial protocol and PWM.PWM is also a digital signal, correct me if I am wrong – Hari Mar 14 '24 at 10:31
  • Hi Hari, have you read my comment above? It does explain that already. – TonyM Mar 14 '24 at 11:07
  • I read that but I don't know why digital signal is not able to provide that .Both looks same – Hari Mar 14 '24 at 11:13
  • A digital signal does indeed carry out the serial communications. Look at the datasheet, no mention of PWM. Instead, there's a protocol spec'. PWM does not vary its duty cycle once per cycle to encode binary data. PWM uses a continuous waveform to deliver an average voltage or power. This protocol doesn't do these things. Instead, it uses a fixed number of bit periods (24?) to deliver a binary message then stops. Sorry to ask but do you really understand the communications protocol, serial communications in general and the purpose of PWM? (I'm afraid "they look the same" isn't any of those.) – TonyM Mar 14 '24 at 11:40
  • Thank you. I need to revisit pwm and protocols again – Hari Mar 14 '24 at 12:28
  • 1
    You're welcome, glad it helped :-) You'll find plenty of much more detailed and better-explained information about this already on the internet. – TonyM Mar 14 '24 at 13:01
  • Your calculations are wrong. The signal does not travel at the speed of light, but at about 0.66*c. And it really has nothing to do with this. You have square wave edges. Infinitely fast edges need infinitely fast bandwidth, but it is likely that you can slow down the edges so they have a sensible bandwidth requirement and only radiate by a sensible amount. – Justme Mar 14 '24 at 13:08
  • If 833kHz is the baudrate of a serial protocol, then I'd say that's way too fast. The human eye is super slow in comparison and surely there's not a lot of data that needs to be transferred. You could likely use <50kHz baudrate and that won't affect anything. Except you'll get way better EMC performance. – Lundin Mar 20 '24 at 11:38
  • @Lundin, you've not read the question or its datasheet properly, I think you're imagining its a UART or something. Many serial protocols have limits on bitrates. The datasheet states that this serial protocol's bitrate is 1.25 µs ±600 ns. – TonyM Mar 20 '24 at 12:29
  • @TonyM Still, that is probably unsuitable for an application with automotive EMC requirements. The OP is having problems with a failed EMC test, how I ended up commenting here. The hobbyist shop Adafruit doesn't exactly help building confidence about part suitability either. – Lundin Mar 20 '24 at 13:42
  • @Lundin, "The OP is having problems with a failed EMC test" Unfortunately you've again not read the question properly: where is any mention of a failed EMC test? Automotives use fast comms, like CAN, anyway but this is single-ended signalling within a stationary charger. With the correct driver, and the good attitude to EMC that the OP's showing here, I'm sure they'll get a suitable EMC-friendly design. – TonyM Mar 20 '24 at 14:10
  • @TonyM https://electronics.stackexchange.com/questions/706696/emc-re-test-failure-how-to-approach – Lundin Mar 20 '24 at 14:17
  • @Lundin, this question is about a charger. The question you link to is by the same OP but about electric bikes instead. – TonyM Mar 20 '24 at 16:29

0 Answers0