TheBackShed.com

Forum Index : Microcontroller and PC projects : Frustration with ESP8266

Author

Message

Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1664

Posted: 04:58am 23 May 2016

Copy link to clipboard

Print this post

Totally at a loss with the ESP8266 I have connected here.

It's an ESP-01, a black one, 1 Gig I think, marked "AI-Cloud Inside".

Flashed it with ESP-Link & it will work fine for periods.

Then it just goes off air.
powering up & down, resetting, disconnecting the CH-PD etc will randomly get it going again eventually.

It can be stable for hours on end, or be up & down like a Yoyo.

Could it be partially fried as I don't have a 33k resistor in series with the RX line?

Also is it required th have the rest pulled high?

Initially it wasn't connected, just VCC & CH-PD held high, but pulled reset high today & thought it may have made a difference, but it didn't.

Same situation on either of my access points, and the Esp-Link interface shows my two AP's with -60 & -61DB & it's Wifi Rssi at -64DB.

Just pulled the power on it while posting & it immediately sprang back to life.

Tried the same thing earlier & not a chance of getting it to come up again.

Could it be environmental interference?

Lots of wireless stuff in this place, but generally everything else works.

Thanks

Phil

matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 8605

Posted: 08:40am 23 May 2016

Copy link to clipboard

Print this post

Normally faults with the ESP8266 are power supply related. You need a solid 500mA 3.3V supply (not USB) with a decent size reservoir capacitor close to the module (I use 1000uF electrolytic + 10uF ceramic)

Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1664

Posted: 11:38am 23 May 2016

Copy link to clipboard

Print this post

Well it's been up for 6 hours now.
5% packet loss.

Power Supply it's running off is one of these, no additional caps though.

Feeding it with a 1A adjustable unregulated supply, plug pack, with its output at around 7.5V.

Did notice initially, that when powered with a 12V regulated wart, that the 5 volt rail seemed to go into a thermal runaway once 5V load approached 300mA.
Thought the chips protection should have prevented that.

Was resolved by lowering the input voltage.
the 300mA load was only experienced when all 4 of the relays on the relay module were pulled in, which will not be a real world situation.

Entire load is the birds nest in my other post.

So 1000uF electrolytic + 10uF ceramic, parallel I presume.
How close is what you suggest? Within cms? or 10 to 15 ok?

Both need to be close, or is there a preference for 1.

Also the 33k in the RX line which Geoff mentioned in his Silicon chip article.
Is that something that should just be done as standard practice.

Cheers.

matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 8605

Posted: 12:26pm 23 May 2016

Copy link to clipboard

Print this post

Quote parallel I presume.

Yes

Quote How close is what you suggest? Within cms? or 10 to 15 ok?

As close as possible to the module, certainly less than 2cm

Quote Also the 33k in the RX line which Geoff mentioned in his Silicon chip article.
Is that something that should just be done as standard practice.

I don't know the reason for this, not something I've ever done assuming both devices are at 3.3v.

Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1664

Posted: 01:04pm 23 May 2016

Copy link to clipboard

Print this post

matherp said

Quote Also the 33k in the RX line which Geoff mentioned in his Silicon chip article.
Is that something that should just be done as standard practice.

I don't know the reason for this, not something I've ever done assuming both devices are at 3.3v.

Geoff mentioned restarting the module by powering it down with a TPS2042B.
In his Silicon Chip article in Dec' '14.

[Quote=From SC Dec' 2014 Geoff Graham]
Note that if you use this technique, you should also place a 33k resistor in series with the signal going to the modules Rx pin. This is because the microcontroller will still be driving its Tx line at 3.3V when the module is powered down and the resultant current into the module's Rx input will destroy it.
The resistor will limit this current to a safe value (100�A).
[/quote]

I've been powering mine down by pulling the 3.3V supply, so that's why I wonder if I've done damage.

Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1664

Posted: 02:07pm 24 May 2016

Copy link to clipboard

Print this post

Added those 2 Caps Peter suggests.

Not as close as recommended, just on the breadboard for the moment, with the module still attached to that with 20cm fly leads, but seems much more stable.

Thanks.

Phil.

Print this page

To reply to this topic, you need to log in.