Very poor wifi reception?

Question

I have the "2.9″ ESPaper Plus Kit" and it seems to work well, except it seems to have extremely poor wifi range. I've got it on my desk with unobstructed line-of-sight to the router (which is on the ceiling about 40' away) and frequently when it tries to refresh itself, it instead shows me the "Could not connect to Wifi / Press LEFT + RIGHT button to enter config mode" screen.

if I move to where I'm right next to the router it always works.

When it works it's a really cool little device and I have a million ideas I want to try with it, but it's hard to hack on when I'm constantly fighting the wifi.

Is this just what I should expect from such a small chip or is there anything I can do to improve this situation?

Answer 1

We're sorry to hear you're seeing intermittent WiFi connection failures. While this is not a commonly observed behavior (to our knowledge) there are so many factors that influence WiFi reception that it's hard to give specific advice.

Of course, it would most likely help if you could move the device closer to your AP but I'm sure you already tried that. Also, can you rule out that too many WiFi devices compete on the same channel?

The Arduino WiFi library has a function to control RF output power but by default it's already set to the max value. Can you maybe yank up the WiFi volume of your access point?

Comments

this is at my work so I can't really move where I sit or where the AP is. I'll just keep experimenting.

---

I have the same issue, received my device today but the wifi is unstable. It does not run for over an hour without loosing WiFi. It does not re-connect when it receives WiFi again, the device is 1 foot from the AP. I run a couple other ESP's on this access point 50 feet away so it is an ESPaper specific problem.

---

I have noticed that the close proximity of the ESP8266 to the display driver of the epaper causes RFI issues, separating them by a few 10's cm helps, but it's clear the display driver chip degrades the Wi-Fi reception. I have been experimenting with only powering the display when it's time for reception using a JFET controlled by an I/O Pin and that helps a lot, but it adds complication to the designs.

I have also noted that inadequate power supply decoupling / smoothing is another factor and although increasing the size of storage capacitors helps (but they are very inductive and so slow to respond due to -di/dt effects) including the use of small low-inductance capacitors (0.1uF) helps it does not solve the problem entirely.

---

@G6EJD that is an interesting observation, thanks.

"the close proximity of the ESP8266 to the display driver of the epaper causes RFI issues" - how about shielding the display driver somehow?

"separating them by a few 10's cm help" - do you happen to have a post with pictures a/o video that demonstrates this setup?

---

It’s an observation I’ve had for a while in that I never had WiFi reception problems with either an ESP32 or ESP8266 when using an OLED or TFT display but as soon as I switch over to e-paper displays the reception/connection problems started. I’ve noted that antenna orientation affects the link and by turning the ESP by trial and error the drop outs can be reduced. Then I discovered if I use 20cm dupont cables to separate the ESP from the display the problem is almost eliminated and that figured to me because a wavelength at 2.4GHz is 12.5cm and in general when you separate a device from an RF source by 1-wavelength it’s starting to enter the far field and RF energy is much reduced. If I can find time, I’ll wrap a display in a plastic bag and then cooking foil and see if that makes a difference. I don’t know if your board design has any low inductance ceramic decoupling capacitors say 0.1uF in addition to a larger smoothing capacitor because the low inductance will help reduce any RFI problems. Larger capacitors with their highe inductance will tend to very high impedance at RF frequencies and become ineffective.

---

More observations: the pcb layout places the ESP at about 2.5cm from the display driver chip, if ever the pcb was redesigned the FTDI connector and ESP positioning could usefully be swapped increasing the clearance to about 7 to 8 cm.

Today I’ll do some RSSI measurements in close proximity to the display with and without foil. My problem will be keeping the display active for long enough to see any effects, I’ll probably keep clearing the display in a loop.

---

I have undertaken some measurements of RSSI (received signal strength) using an ESP8266 and an e-paper display that is being refreshed continually in a loop, very simple, clear the screen read the RSSI, then repeat.

The first set of results are with the e-paper unit ~20cms away from the ESP8266, there is a distinct 6dBm advantage to having the antenna pointing directly at the Router to having the ESP8266 being flat face towards the router if that makes sense. This result is to be expected.  The reduction in power is 75% when not orientated towards the Router.  Each -3dB is a loss of 1/2 power at the receiver input. So -6db is 1/2 * 1/2 = 1/4 (1-1/4)=0.75

With the display (the end where I believe the display driver chip is located - the thick white section) now placed in close proximity to the ESP8266, the signal drops from the freer space reading of -60 to -64dBm and -66dBm to -69dBm with antenna pointing at and at 90° respectively.

So all this means there is some correlation with received signal strength and display proximity, it degrades Wi-Fi reception.

If your customers experience reception problems you should advise them to point the longest edge of the PCB towards their Router, this will ensure they are getting maximum signal strength.

You could add an RSSI display to the Setup screen to give an indication of signal strength at the instant it is drawn. Anything less than -70dBm e.g. -75dBm is beyond a usable limit. I'm already at the limit of my Router range in these results and I have a nearby extender that I normally use. See my G6EJD YouTube Tech Note 053 for a chart of usable Wi-Fi signal strengths.

Also you might ask your customers to try a simple Wi-Fi signal checker and if they report Wi-Fi signals below - 60dBm e.g. -65dBm then they have reception problems, this is a common problem.

Hope this helps.

Antenna pointing directly at Router (antenna square-wave like element longitudinally pointing at Router)
E-Paper screen 20-cms away from ESP8266
SKYAE43F    Signal: -60
SKYAE43F    Signal: -61
SKYAE43F    Signal: -60
SKYAE43F    Signal: -61

Antenna at 90° to Router
E-Paper screen 20-cms away from ESP8266
SKYAE43F    Signal: -66
SKYAE43F    Signal: -67
SKYAE43F    Signal: -67
SKYAE43F    Signal: -66

Antenna pointing directly at Router (antenna square-wave like element longitudinally pointing at Router)
E-Paper screen 1-cms away from ESP8266
SKYAE43F    Signal: -64
SKYAE43F    Signal: -64
SKYAE43F    Signal: -65
SKYAE43F    Signal: -64

Antenna at 90° to Router
E-Paper screen 1-cms away from ESP8266
SKYAE43F    Signal: -69
SKYAE43F    Signal: -69
SKYAE43F    Signal: -69
SKYAE43F    Signal: -69