ESP8266 Community Forum

- Sun Mar 20, 2016 5:40 pm #43654 I haven't got any unshielded parts so I can't scope this out. As for RF bypassing the GPIO's 1E-9Fx100R (100R as a conservative guesstimate of RS232 driver source impedance) is going to be OK up to 1Mb/s (1/10th bit period slope) but you're free to drop the capacitance all the way down. 2.4GHz wouldn't get far in a network with a few tens of nH and pF.

I can't picture how injecting current alone would help with an RF problem here unless the parasitic capacitance of your resistor is more significant than usual (what type is it? SM?) .

As for heat, your flood ping will be exercising the PA to the max. with 3V3x0.17A~=0.5W in that little package, yes it will get toasty!

- Mon Mar 21, 2016 9:09 am #43706 Yes, I already found out that stopping the flood ping would make the chip cooldown again . Good to know it was not the strong pull-up that causes the heating up. That also suggests the ESP8266 has some sort of "halt" mode comparable to other microprocessors where no instructions are executed when nothing is left to do. No busy waiting. And also that this mode actually shuts some parts off. In contrary to older Intel microprocessors where it would remain full-on in halt mode.

Again, I think the cause is leakage of RF signal (maybe even not in the chip, but parasitic from a parallel print trace). Pulling the trace up will make the impedance go down by a fair amount. Otherwise the lines on my print are "open ended" and I guess we all know what happens if you insert RF into unterminated/high-impedance wires/trace

- Mon Mar 21, 2016 11:52 am #43720 AdrianM, I think your are spot-on!

I replaced the two resistors on TxD and RxD with capacitors of 1 μF instantly got perfect results. I didn't test if the UART still works this way, but at least the WLAN connection is perfect now.

Interesting is that if I change the mode of the pins (1 and 3) from UART to simple GPIO digital output, there is no change and I still need the capacitors. So it's not an UART "thing". Maybe it's simply something of the wiring/tracing on the ESP-01 (which I am using right now).

So indeed, apparently the ribbon cable to the CP201x and/or other wiring was acting as a small capacitor, but still big enoug apparently.

- Tue Mar 22, 2016 5:22 am #43771

I replaced the two resistors on TxD and RxD with capacitors of 1 μF instantly got perfect results. I didn't test if the UART still works this way, but at least the WLAN connection is perfect now.

I guess you really meant 1 μF because that's a non-standard ASCII symbol! In that case you wouldn't want to try driving it with any digital pin. Do you have any 1n or 100pF caps?