Lowest cost ESP8266

I recently ordered a couple of ESP-12e modules, and (from a separate supplier) some adapter boards, so I can plug in “du-pont” wires, rather than having to mess around with the tiny contacts on the ESP-12e.

This is the thing I’m talking about.

The ESP-12e modules were pretty cheap, and the adapters came in at £0.49 each, shipped. Not bad I thought.

When the adapters arrived they came complete with… a free ESP-12 …

I checked my original order… yes, I had only ordered the board, and yes they only cost £0.49… so that left me with a problem… I now had three ESP-12e modules with no board (having married up the ESP-12s with the boards they came with).

Now what to do… I ordered another three adapter boards from the generous supplier of course. Now I don’t know if that makes me a bad person but if I get another three free ESP-12 modules I will reveal the supplier and you can try for yourselves. Now I could of course have gone back to the supplier and point out their mistake… so the question is… what would the rest of you have done in the circumstances…

You are of course under no obligation to answer, and/or call me out for the thieving cad that I am.

Hi Andy,
Lucky guy !
Maybe you should share the link of the seller …
Personally, I added 2mm pins on all my modules, and I’ve made a small proto-programmer with female header, so I’m switching modules to upload sketches.
I did the same with real PCBs that I have designed.

martinayotte wrote:Hi Andy,
Lucky guy !
Maybe you should share the link of the seller …
Personally, I added 2mm pins on all my modules, and I’ve made a small proto-programmer with female header, so I’m switching modules to upload sketches.
I did the same with real PCBs that I have designed.

ahull wrote:<…>
Now I don’t know if that makes me a bad person but if I get another three free ESP-12 modules I will reveal the supplier and you can try for yourselves. Now I could of course have gone back to the supplier and point out their mistake… so the question is… what would the rest of you have done in the circumstances…

You are of course under no obligation to answer, and/or call me out for the thieving cad that I am.

I ordered some ESP8266s a few months ago and it was cheaper to order ones with breakout boards than it was to order the ESP8266 module on its own.

So I just put the breakout boards in my component sock boxes, just in case they come in handy for something else.

I just spent a good hour trying to program one of these ESP-12 boards with a CH340G USB to RS232 TTL module. Everything looked fine, and the ESP board uploaded perfectly with my 3V3 FTDI USB cable.

The solution (after much head scratching) was a 10k pull down on the TxD pin. It seems the ESP doesn’t like 3v6 input voltages if you are running it from a 3v3 regulator. Well at least this time, no smoke.

ahull wrote:
<…>
The solution (after much head scratching) was a 10k pull down on the TxD pin. It seems the ESP doesn’t like 3v6 input voltages if you are running it from a 3v3 regulator. Well at least this time, no smoke.

@Ray, you didn’t look thoroughly to the schematic, it doesn’t use voltage regulator but 2 diodes in series …
Therefore 2 x 0.7V drop, 5V – 1.4V = 3.6V …

microscopic would be more correct, someone remind me of the tx current?
istr >300mA or am i thinking nrf24l01’s?

you got the -12E’s pretty cheaply, but have you tried finding an adapter plate for the -12E device?
millions listing as suitable for 12/12E, but product shown is only for the -12.

my feedback on some extended base plates i got was ‘excellent for the 0.9 dev kit’. definitely positive feedback
the 1.0 isn’t wide enough.

the 0.9 exposes 4 holes each side if spanning a pair of ‘supply rails’ bread board
the 1.0 exposes a single line of holes each side when used on bread board, doesn’t make
across the double psu rails, it would if it was a single ‘supply rails’ block

what’s worse is trying to work out the search strings!

<edit>read last post first, so the tiny 3 pin smd is a double diode?
</edit>

stephen

Euhh ? I was talking about the 2 diode in series on the USB-CH340, they are normal SMD 1N4148 …

i was commenting about these

or if picture doesn’t work as in
http://www.stm32duino.com/posting.php?m … 756#pr8431

stephen

Ah ! Ok !
But, then, if I remember, the 3pins SMT on the back is simply an AMS1117 3.3V regulator.

mrburnette wrote:ahull wrote:
<…>
The solution (after much head scratching) was a 10k pull down on the TxD pin. It seems the ESP doesn’t like 3v6 input voltages if you are running it from a 3v3 regulator. Well at least this time, no smoke.

check the middle resistor on those boards. The boards were designed to have a 3.3V LDO and to be powered by 5V. Then they removed the LDO and bridged the trace with a 0Ohm resistor (basically a short). The two resistors on the side are a pull up and pull down, 10K Ohm each. On some of the cheap boards, they use 3 10K resistors, instead of 2 10K and a 0Ohm… needless to say, the ESP8266 module does not like a 10K resistor before VCC…

Also, most USB/Serial adapters do no provide enough mAh for the ESP8266… you really need to power an ESP8266 with its own 3.3V supply

@Andy, those additional pins at the bottom of ESP-12E are not really useful since they are the SPI interface and QIO for the Flash.

robca wrote:check the middle resistor on those boards. The boards were designed to have a 3.3V LDO and to be powered by 5V. Then they removed the LDO and bridged the trace with a 0Ohm resistor (basically a short). The two resistors on the side are a pull up and pull down, 10K Ohm each. On some of the cheap boards, they use 3 10K resistors, instead of 2 10K and a 0Ohm… needless to say, the ESP8266 module does not like a 10K resistor before VCC…

Also, most USB/Serial adapters do no provide enough mAh for the ESP8266… you really need to power an ESP8266 with its own 3.3V supply

martinayotte wrote:@Ray, you didn’t look thoroughly to the schematic, it doesn’t use voltage regulator but 2 diodes in series …
Therefore 2 x 0.7V drop, 5V – 1.4V = 3.6V …

mrburnette wrote:martinayotte wrote:@Ray, you didn’t look thoroughly to the schematic, it doesn’t use voltage regulator but 2 diodes in series …
Therefore 2 x 0.7V drop, 5V – 1.4V = 3.6V …

ahull wrote:
Do I saw the 1/2 diode length wise or width wise?
I think my 10K resistor is the lowest cost option, and since this is a Cheapskates Incorporated Production, the 10K resistor wins.

mrburnette wrote:ahull wrote:
Do I saw the 1/2 diode length wise or width wise?
I think my 10K resistor is the lowest cost option, and since this is a Cheapskates Incorporated Production, the 10K resistor wins.

stevech wrote:mrburnette wrote:ahull wrote:
Do I saw the 1/2 diode length wise or width wise?
I think my 10K resistor is the lowest cost option, and since this is a Cheapskates Incorporated Production, the 10K resistor wins.

ahull wrote:<…>
The quality of the boards is low (obviously), with some hand soldering on the USB connector for example, where they have only bothered to solder one of the retaining lugs. …

mrburnette wrote:ahull wrote:<…>
The quality of the boards is low (obviously), with some hand soldering on the USB connector for example, where they have only bothered to solder one of the retaining lugs. …

mrburnette wrote:
A perfect example of the pressures and issues on a production line that causes chaos: https://www.youtube.com/watch?v=HnbNcQlzV-4

stevech wrote:
Two silicon + one Schottky diode in series to get closer to 3.3? Geesh, what did that board seller save? $0.10?

Here’s a challenge, could we design it cheaper?
I’m thinking single layer board, no leds, smd usb socket, 12mhz resonator to replace the crystal.. don’t solder on the right angled header (or indeed include one at all). Replace the jumper with a solder bridge on the board to select 5V/3V replace the two diodes with a resistor voltage divider. Use a bare die chip in an epoxy blob…

ahull wrote:
<…>
12mhz resonator to replace the crystal..

ahull wrote:Use a bare die chip in an epoxy blob…

mrburnette wrote:
…
Thinking out loud on an empty stomach … so this could be a crap idea:
For full-duplex use, I wonder if “auto selection” could be utilized; that is, the unit starts at 3.3V but converts to 5.0V on Send if it gets a 5.0V on Receive. This is obviously a bit problematic in simplex mode, but perhaps could be overcome if the Receive line was tied to the Send line for simplex.

Ray

ahull wrote:
An interesting idea, but I suspect it would be a little difficult to implement, since you don’t know ahead of time what the receiving end expects.

Furthermore there is a risk that your device will misread the incoming data peak voltage (due to surges, software glitches or whatever) and select the wrong reply voltage, thus cooking the other device….