Let's Control It

I thought this might be useful to others here.

I was noticing that one of the Wemos D1 mini I have running had a fairly weak WiFi signal (-87dB) compared to another almost next to it which was about -76dB. I thought this is major difference given how close they were in distance to the Access Point (6m). I then recalled that the unit with the lower signal had shield directly on top of the D1 mini, while the other had a sensor board about 60mm away from the D1 mini. I changed the configuration of the weaker D1 mini around such that shield was placed beside the D1 mini using one of the double shield boards. The signal then went up to -76dB.

In both cases the boards were orientated such that the on board antenna of the D1 mini was on the far side of the board in relation to the access point. Thus the signal would have been degraded by the board itself. Having said that, it does show that physical configuration of systems and orientation of ESP8266 can have a significant impact on range.

Yep, also the "polarity" of the antenna can make a huge difference.
So try rotating the board to see which is best.

Not only metal will greatly impact WiFi performance.
Water does too, so holding it in your hand does also make a difference.

And some plastics also affect the resonance frequency of the antenna.
For example ABS can de-tune an antenna a few 100 MHz. The ESP does have compensation circuits, but those do have their limits.
Thus it can happen that it is working fine on channel 1 and not on channel 11. (or vice-verse)