ESP 201 Development Board USB Cable Sensitivity Modification
Posted: Wed Jun 22, 2016 5:22 pmGreetings,
I encountered a problem with the ESP-201 Development Board that is equipped with a relay, temperature sensor, switches and LEDs. This board is available out of China on eBay, Aliexpress and other vendors. I am very pleased with this development board and ended up purchasing a second one.
However after I decided to give Arduino a spin after experimenting with LUA I noticed that the serial port monitor was not reading the expected program output data. After jumping through some hoops by trying different code I noticed that the monitor appeared to be loosing connection to the serial port each time I powered cycled the esp8266. If I closed the port monitor and re-started it then it would again read and display the data. The problem is significant as one looses serial port connection every time the module is turned on. This could result in unexpected loss of data for data acquisition applications and to say the least is confusing during development phases.
These esp201 development board employs a CH340 USB/UART data interface chip which receives power directly from the USB port. Whilst connected to the USB port the CH340 remains powered all the time, even when the esp8266 power is turned off. As the 8266 reset pin is connected directly to VCC the only way one can reset the esp8266 on this board is using the power switch to power cycle. As I have two boards I tried testing each and found the same results.
I tried both Windows 10 and then XP on an old computer and again found the same results. Device Manager showed that the serial port remained connected to the CH340 during the power off period, as expected.
Thinking that it might have been an Arduino issue I decided to test with different terminal programs and found they also had the problem. Putty reported a fatal serial port error when the power was re-applied to the esp8266. It appears the port connection was dropping momentarily during power up of the esp8266.
In the process of testing I decided to try different USB cables and discovered that with one cable the problem disappeared. I tried four different cables and found only one out of the four would work. All cables worked for regular data but with 3 out of 4 the serial port connection would become broken when the esp8266 power was cycled. This required closing and re-opening the port to restore the serial port connection.
With the realization that the problem was related to the re-application of the esp8266 power I decided to review the schematic. I noticed that there are tantalum electrolytic bypass capacitors (as there should be) on the input and output of the 3.3V regulator which is powered from the switch. The CH340 is powered by 5V directly from the USB port. The CH340 only had a small value ceramic bypass capacitor (C5) across its power pins. I concluded the current surge needed to charge the regulator capacitors during power up was probably causing a momentary voltage dip on the CH340 causing it to fail momentarily. I tacked a 22uF/6V tantalum capacitor across C5 to prevent the CH340 voltage dip during the power on cycle. This has solved the serial port issue. The serial port no longer drops out with the use of any of my cables.
From a circuit design viewpoint a better solution would have been to power the CH340 through a low voltage drop schottky diode to isolate it from the esp8266 power up serge. However as this would have involved hacking the PC board I chose the easier to implement tantalum capacitor solution.
I hope this information is found useful by others.
Best Regards,
tma
I encountered a problem with the ESP-201 Development Board that is equipped with a relay, temperature sensor, switches and LEDs. This board is available out of China on eBay, Aliexpress and other vendors. I am very pleased with this development board and ended up purchasing a second one.
However after I decided to give Arduino a spin after experimenting with LUA I noticed that the serial port monitor was not reading the expected program output data. After jumping through some hoops by trying different code I noticed that the monitor appeared to be loosing connection to the serial port each time I powered cycled the esp8266. If I closed the port monitor and re-started it then it would again read and display the data. The problem is significant as one looses serial port connection every time the module is turned on. This could result in unexpected loss of data for data acquisition applications and to say the least is confusing during development phases.
These esp201 development board employs a CH340 USB/UART data interface chip which receives power directly from the USB port. Whilst connected to the USB port the CH340 remains powered all the time, even when the esp8266 power is turned off. As the 8266 reset pin is connected directly to VCC the only way one can reset the esp8266 on this board is using the power switch to power cycle. As I have two boards I tried testing each and found the same results.
I tried both Windows 10 and then XP on an old computer and again found the same results. Device Manager showed that the serial port remained connected to the CH340 during the power off period, as expected.
Thinking that it might have been an Arduino issue I decided to test with different terminal programs and found they also had the problem. Putty reported a fatal serial port error when the power was re-applied to the esp8266. It appears the port connection was dropping momentarily during power up of the esp8266.
In the process of testing I decided to try different USB cables and discovered that with one cable the problem disappeared. I tried four different cables and found only one out of the four would work. All cables worked for regular data but with 3 out of 4 the serial port connection would become broken when the esp8266 power was cycled. This required closing and re-opening the port to restore the serial port connection.
With the realization that the problem was related to the re-application of the esp8266 power I decided to review the schematic. I noticed that there are tantalum electrolytic bypass capacitors (as there should be) on the input and output of the 3.3V regulator which is powered from the switch. The CH340 is powered by 5V directly from the USB port. The CH340 only had a small value ceramic bypass capacitor (C5) across its power pins. I concluded the current surge needed to charge the regulator capacitors during power up was probably causing a momentary voltage dip on the CH340 causing it to fail momentarily. I tacked a 22uF/6V tantalum capacitor across C5 to prevent the CH340 voltage dip during the power on cycle. This has solved the serial port issue. The serial port no longer drops out with the use of any of my cables.
From a circuit design viewpoint a better solution would have been to power the CH340 through a low voltage drop schottky diode to isolate it from the esp8266 power up serge. However as this would have involved hacking the PC board I chose the easier to implement tantalum capacitor solution.
I hope this information is found useful by others.
Best Regards,
tma