ESP8266 Community Forum

Hello - I hope I'm in the correct forum and this has not been covered before. Please advise if so.

I read a lot of criticisms on the ESP8266 and the AT command set and wondered just how good or bad it would prove to be over time.

I bought an 'HUZZAH' module from Adafruit at the start of Sept 2015. I flashed it to be

AT version:0.25.0.0(Jun 12 2015 20:26:28)
SDK version:1.1.2
compile time:Jun 17 2015 22:11:51

I chose this firmware as, though I'm not sure, I think the HUZZAH is 1/2 a megabyte and won't fit the latest command set firmware which is bigger (anyone know better?).

So far it has performed flawlessly in simple connect to an access point tests. It is connected to my PC USB port via a Freetronics USB-serial converter. I'm not external powering it (and I was surprised how well this worked as I expected the USB power to be 'rubbish').

I've left it powered up for 5 days now.

I've written some code in Delphi to send AT commands to it, wait between each looking for a response;

AT+CWMODE=1
AT+CWJAP="Office Home Gateway","????????"
AT+CIFSR
AT+CWQAP

It works remarkably well and consistently.

My only criticism is that the AT+CWLAP command does not find _all_ access points each time. It takes a few tries to get them all. Some are far away (showing RSSI of -91 and worse) - maybe this is just luck of the draw as to what it has received error free resently. I will say that the module seems to be very sensitive. An iPod beside the module fails to find as many access points. Yay!

My reason for this post is to share what I've done so far - and ask if others have had any experiences I may benefit from with this module.

My next step will be to connect to a web server and download from it etc - then run this continuously and log the results.

Kind regards
RichardTXD

All features you mention are traits of the firmware, not of the hardware. Just flash a newer firmware.

Thanks for your feedback.
I feel I must give a "yes and no" response. Firmware sits inthe hardware. How do you determine which is problematic under use?
I've seen posts recommending a pullup on the RST pin and also other pins of the device should be properly terminated which makes sense to me - it's in a fairly high RF field stength environment so terminating unused pins would seem reasonable. Also, power is mentioned as being critical considering the high peak current required - the use of some filter caps being recommended. Again, sounds very reasonable and USB power is often cited as being 'noisy'.
In the above, I was very lazy. I did not do the right thing with power yet it behaved well which surprised and pleased me.
Yes you are quite right that the behaviour of the device is due to the firmware and if that is not stable then it will of course fail as you'd expect.

I suppose I was and am looking for a 'baseline performance' and from that, measure in some way, the effects of changes.

One other thing I've seen recommended is to use the AT+RST command to soft reset the module. I've not used that at all as I wanted to see if it really was necessary - does it make a difference for my testing? I'm thinking people will say yes - but for me so far it has not been an issue.
Using AT+RST does concern me. Is this required regularly? How often? Can I open a link then use it for how long before an AT+RST is required?
So far and now after about 8 or 9 days of testing, I've not power cycled or reset the module in any way. It does seem to work well especially considering my laziness in using it.
I was, and am, looking at it from a 'total picture' viewpoint and my questions being how reliable is it with the current firmware and with the HUZZAH board used 'as is' without going to any effort.
If I can get it to fail under repeated load then work on the hardware to find if that is causing it - or isolate fault to firmware then reload a new firmware - those are my current goals.
Thats where I'm coming from.
Kind regards - Richard

So the only hardware trait is the pullups and pulldowns, not quite rocket science