Hi all, I attached a Image of ESP32's datasheet where it say to add a series 470 Ohm resistor to Tx transmission line.. I wonder is it talking about which line, what are the pins ?

and why that specific value ?

here Is the link to the ESP32 datasheet https://www.espressif.com/sites/default/files/documentation/ESP32_FAQs__EN.pdf

 

the single resistor on circuit tracks is a popular and cheap quick fix that attemps to reduce currents
bouncing back, out of mismatched impedances along circuit tracks that on DC are pretty good, but as soon
as frequency increases, capacitance and inductance highlighted by 1/Cw Lw (w=2*pi*f, f[Hz])
and more complex models of DC resistive only components.

On complex boards that must meet strict design budgets, once the prototypes are assembled and everything else works ok, one cannot always start all over again looking for alternative components with ideal impedances
because after all.

Oonce the EMC compliance test is passed, the product is ready for sale
and a reduced time to shelves is critical for new products to be successful.

I checked
https://www.espressif.com/en/esp-wroom-32/resources
the recommended resistor on TXD is pin 41 (U0TXD)

it just reduces current therefore lowering the mentioned:

'.. harmonics of 80 MHz (160 MHz, 240 MHz, 320 MHz ..'

that probably come from the crystal U1 at 400MHz and the small size of the board with also reduced
grounding plane, don't think much metal is used either precisely because of keeping the price as low
as possible for IoT devices.

Have you tried to read Herr Bartmann's new book?

https://www.espressif.com/en/news/New Book on ESP32 by Best-Selling Author Erik Bartmann

John BG

 

Thanks for the reply John, but wonder why only the TXD line, and UART are low frequency signals...

How does crystal clocks relating to that line ? Hope you can direct me well

the single resistor on circuit tracks is a popular and cheap quick fix that attemps to reduce currents
bouncing back, out of mismatched impedances along circuit tracks that on DC are pretty good, but as soon
as frequency increases, capacitance and inductance highlighted by 1/Cw Lw (w=2*pi*f, f[Hz])
and more complex models of DC resistive only components.

On complex boards that must meet strict design budgets, once the prototypes are assembled and everything else works ok, one cannot always start all over again looking for alternative components with ideal impedances
because after all.

Oonce the EMC compliance test is passed, the product is ready for sale
and a reduced time to shelves is critical for new products to be successful.

I checked
https://www.espressif.com/en/esp-wroom-32/resources
the recommended resistor on TXD is pin 41 (U0TXD)

it just reduces current therefore lowering the mentioned:

'.. harmonics of 80 MHz (160 MHz, 240 MHz, 320 MHz ..'

that probably come from the crystal U1 at 400MHz and the small size of the board with also reduced
grounding plane, don't think much metal is used either precisely because of keeping the price as low
as possible for IoT devices.

Have you tried to read Herr Bartmann's new book?

https://www.espressif.com/en/news/New Book on ESP32 by Best-Selling Author Erik Bartmann

John BG

 

how?

it's in the 1st link I mentioned to you in previous post. Go to

https://www.espressif.com/en/esp-wroom-32/resources

open the Hardware design point, then download the .zip

inside the .zip there's the PCB and schematic so you can build yourself the hardware.
All needed to program the device is also available from the same ESPR page.

Have a look

If you really want to know how inside the chip, you will have to ask the manufacturer, but it may be IP protected,
which is why pros put the circuits in really small packages, the smaller the safer, because fewer have
the tools to open the chip and 'have a look'.

But for an application engineer as you seem to be, do you really need to into such depth?

Regards

John BG

 

Yes I am just a field engineer.. but I was just curious. I design complex prototypes but since it's prototype i never looked into EMI/EMC before.

how?

it's in the 1st link I mentioned to you in previous post. Go to

https://www.espressif.com/en/esp-wroom-32/resources

open the Hardware design point, then download the .zip

inside the .zip there's the PCB and schematic so you can build yourself the hardware.
All needed to program the device is also available from the same ESPR page.

Have a look

If you really want to know how inside the chip, you will have to ask the manufacturer, but it may be IP protected,
which is why pros put the circuits in really small packages, the smaller the safer, because fewer have
the tools to open the chip and 'have a look'.

But for an application engineer as you seem to be, do you really need to into such depth?

Regards

John BG

 

I think it is very likely that the resistor simply acts with the capacitance of the cable to form a low pass filter acting on the transmit signal itself rather than anything to do directly with the clock frequency. It would reduce the slew rate of the edges of the signals and in consequence the high frequency content. If the problem were coupling of clock frequency into external connection, all external lines would need RFI countermeasures, as you suggest. A lossy ferrite "bead" around the cable would be one such measure. There is no way that 470 ohms does anything in terms of impedance matching since there is no practical cable used in electronics with an impedance anywhere near that. Single-ended signals such as used for async serial very rarely use cables with well-defined impedance since you can't achieve well defined impedance with a bunch of parallel conductors in a common jacket.

You sometimes see ferrites on things like USB cables. The cable vendors will tell you that is to keep noise from external sources out to improve data integrity. What it actually does is keep noise IN.