Hmm... Do I put this in General Electronics or Project Guidance? Hmm...
Thanks for taking the time to read this post. I am trying to do a project which requires some pretty hefty work as a web server. The first version used the WiFiEsp library and an ESP-01. It "works" but lacks in power. AT commands over serial just doesn't provide enough power.
So, the new version will attempt to use an ESP-12F over SPI using the WiFiSpi library. My "Arduino" uses an ATMEGA1284 at 20MHz and runs on 5V. The board is already completed and I don't want to redesign everything for 3.3V,
I have cobbled a schematic together from various sources. The logic level converter is from a SD card module. The WiFiSPI library suggests a safe CS circuit because the ESP won't boot if the CS pin is low. That left one channel open which Jiri suggested I should use for RST. I am trying to implement a power switch (EN pin) using a PNP transistor. I brought out the TX0 pin just because the library can send debugging info via serial. Building a prototype is no small task because the ESP has a 2mm pitch.
If you have read this far, please take a look at the schematic and let me know of any errors you may detect. I sincerely thank you.
Why the added complication of two processors.
The ESP-12 runs on 80Mhz by default, and has more than enough internal flash (SPIFFS) to store many web pages.
What is your project.
Several parts of your diagram don't make sense.
Leo..
That's one of the problems with this forum - there are simply too many topics. I often wonder how much of the moderators' time is spent just moving posts to the correct topic.
AT = Antique Technology.
When I see ESP and AT commands in the same post, I am reminded of Marvin's prophetic words: "This will all end up in tears. I just know it."
Why are you using an ESP board to provide WiFi connection to an Arduino when there are ESP boards that can already do the tasks of most Arduino boards? I am thinking NodeMCU or WemosD1 Mini.
Oh, on the schematics:
You are deigning a PCB and don't have a schematic for it?
Your schematic cobbled together from various sources is miles ahead of a Fritzing picture, but we generally don't have chip pinouts in our long-term memory. Label the pins.
@ Railroader: Did I say it failed? Actually, the entire first WiFi project works but lacks WiFi power due to being serial based. I didn't ask a single question about coding and the wiring is in the schematic.
My first WiFi project uses 10 libraries, a 128 x 64 LCD, WiFi, Real Time Clock, 16GB SD card, 5 I2C devices, acoustic alarms, has its own internal diagnostics, supports a web page for viewing status, logs, and reconfiguring over 40 parameters, synchs the RTC with the internet, is interfaced to a Victron solar charger, controls two separate 75 Watt loads, etc. etc. etc.
@SteveMann: Concerning AT commands, I agree but that is how the ESP-01 libraries typically function. I don't actually issue the AT commands, the library does it. I got it working, BUT... at 115200 plus the overhead of the AT commands and library, it just isn't very performant.
I am using an ESP board to provide WiFi because most of my projects include a multitude of sensors using I2C, a SPI SD card, a SPI monitor, multiple IO pins controlling LEDs, alarms, ... The 1284 has the power, but of course doesn't have WiFi. I suppose similar situations is exactly why Jiri wrote the WiFiSPI library.
Yes, I am attempting to design a 5V adapter on which the ESP, logic level conversion, power supply, the required "safe Reset", etc. If it works, I can plug it into a project as simple as I now plug in a SD card module or sensor. Creating a schematic is the first step to creating the PCB.
I could do that and maybe I should have said something about the labels I did use. Something like CS5 is the CS coming from the 5 Volt side. CS3 is the same signal on the 3.3 volt side. The logic converter only has VCC and ground other than the signals. The ESP pins I didn't use are things like the ADC, and unused IO pins.
Yes, really. Like I said, creating a schematic is the first step of creating the board.
Thanks a ton for your time and suggestions. Did my explanation clear up enough about the pins or should I go back and label all of them?
Normally port expanders are used, or I2C sensors which all use the same two pins of the ESP.
The BC807 NPN transistor should be a PNP transistor, The 1117 is missing decoupling caps,
that EN circuit should not be needed, because the ESP-12 has an EN pin.
The 2N2907 NPN is the wrong way round, no decoupling on the 74LVC supply, why the 3k3 resistors, should I go on?
Leo..
Oops, the BC807 IS a pnp but I used the wrong part while drawing the circuit. I will correct that.
Ok, I can add 10µF on the VIn of 1117.
I have used a similar circuit with the 807 (except it was throughhole, this time I wanted SMT) thinking that the best way to save power is to shut the circuit down. You are right, the ESP has the EN pin for standby. I haven't used the ESP-12 before and was planning to break the EN pin out to see how much power everything would use in standby vs. power off. In retrospect, maybe I am being fanatic.
I think the 2907 is correct but I again used the wrong symbol in the schematics. The circuit itself is taken from a suggestion Jiri Bilek and others had incorporated into the library docs to allow the ESP to boot if the SPI Master has the CS pin high.
I can add 100nF on the 74LV.
I don't know why the 74LV has 3.3K on the input pins. The circuit was lifted from a manufactured SD Card Module and they are there.
