How to eliminate transient state of I/O pins during ESP8266 module boot?

Question

Welcome, I am designing a controller for a WiFi controlled AC motor based on an ESP8266 12E module. The motor has left/right rotation. The actuators are relays. When the controller is connected to the power supply, both relays switch on for a moment, which causes the motor to want to run in both...

Brivido · Answer

Show an executive diagram? What kind of motor is it, etc. For induction motors and contactors, they should have a mechanical interlock.

Maryush · Answer

This is a motor for a roller shutter. Standard relay control, with a BC848 transistor from the I/O port of the ESP module.

AIIoT · Answer

Maryush I read about this problem once, but I didn't think it posed any real problem, maybe you should try treating the microcontroller as a regular button and use so-called hardware debounce'ing (to eliminate contact vibration) before the base of the transistor.

using a program I wrote to test hardware emulation of contact vibration available at:
http://aiiot.c0.pl/web/2018/02/01/program-do-testowania-sprzetowej-eliminacji-drgania-stykow/ .

.
I have noticed that this transient does indeed occur, and it is possible to get rid of it with the MAX6818 circuit, but whether this option is viable for this purpose is another question, maybe it is enough to analyze the operation of this circuit and reproduce it only to the extent where the oscillation will not occur (just adding a capacitor and resistor will not help)

mpier · Answer

. With standard relay control the motor will not behave as you wrote, if the relays are working. Show the schematic. Have you tried on other pins?

Maryush · Answer

The transient problem when the ESP is powered applies to all I/O pins, just as colleague AIIoT has shown in the oscilloscope diagram. I don't know if applying left and right turns to the motor at the same time can hurt it, if not then it won't be a problem in my case, but if it is, then the problem is already there. The specification for this drive does not give details as to what the protections are.

mpier · Answer

Are you sure? If so, the relays are broken .

Maryush · Answer

Why would the relays be broken? The relay coils are switched, exactly as the diagram shows.

Loker · Answer

I see several possibilities:
1. Hardware elemination of transients on I/O pins - e.g. some RC filters - it would be useful to know the frequency and duration of these transients; besides, this type of filter will introduce a delay during normal operation.
2. some logic to prevent the relays from switching on simultaneously - this will not change the fact that during transients the relays will switch on for a while (not simultaneously) - this will cause the motor to "jerk", which is not desirable.
3. delayed activation of the relay power supply - the ESP power supply is applied first, the relay coils power supply with some (e.g. 1s) delay.

Maryush · Answer

Colleague Loker has analysed and worked out the subject very nicely.
The best solution would be option 3, i.e. first the ESP will turn on and then the power supply will be applied to the relay coils. It has to be some small IC/circuit because I have practically no space on the board at the moment. Maybe something on an NE555?

krzysiek_krm · Answer

Hello, . maybe a typical reset circuit - small, cheap, etc. Regards

Loker · Answer

Could be 555, could be on some kind of counter, could even be something like this: https://www.elektroda.pl/rtvforum/topic295587.html
Just count if there is still enough voltage left after the voltage drops across the two transistors to switch the relay on reliably.

mpier · Answer

. If you have wired everything up correctly, and they are triggering both directions at the same time, then they may be faulty. This is independent of the ESP8266. Are these transients for real? On every pin? Even if you were controlling a low state on the output? Without a schematic, it's difficult to give you sensible advice.

Maryush · Answer

They are not faulty because I have a diode on one of the outputs and it too lights up momentarily when the ESP is switched on, this module has that. I will be testing a solution with a transistor controlled by an RC circuit.

krzysiek_krm · Answer

On a slightly different note - the author probably made a minor conceptual error: he used one relay left and another right . Should have used one relay left - right and another start - stop . Which doesn't change the fact that transients will cause uncontrolled switching of these relays during the CPU ramp up.

mpier · Answer

This is hardly possible. After all, the author wrote that he connected in a standard way. That is, as you wrote, either the NO contacts of the second in series with the NC of the first, or yet another way. Which pins does the author use for control?

Maryush · Answer

If I had known there were transients, I would have given one relay to change the direction of rotation and another to turn on/off. Then the drive would only spin clockwise or counterclockwise when the ESP is activated.
Control on pins 14 and 16.

AIIoT · Answer

Maryush At the very beginning I would suggest you try to change the control method, during start-up the ESP has mainly a high state.

Slawek K. · Answer

And what is your colleague writing the program in ? in arduino ? do the pins have states (LOW or HIGH - depending on the logic of the circuit controlling the relays) set at the start of the programme ? for this, appropriate pulldown or pullup resistors and there should be no problems.
Pin 13 is the RX pin of uart1, so maybe this is the problem.

https://i.pinimg.com/originals/0c/8c/e8/0c8ce8bc22ee6ba0e2ed830bcbb28454.png

Greetings

krzbor · Answer

See the topic I am promoting (in the footer). This solution certainly works.

Maryush · Answer

. Very well described, if there is a clear on any of the ESP8266 12E pins at startup, this control can be done. I'll also try with adding pull-up resistors to zero, but I don't think that will do anything.

krzbor · Answer

I have given the waveforms - you can see that there is a stable signal on Rx (GPIO3). At my place this circuit controls the garage door and the entrance gate. It has never opened by accident yet.

Maryush · Answer

The resistors were of no use. I will do the control indirectly through an optocoupler and it should be fine. Does the ESP8266 12E have a second UART? I would like to use the first one for communication with the Atmega and the second one for debugging. I am writing the programs in arduino.

krzbor · Answer

Regarding the second UART you didn't quite read my promoted topic Someone asked this question there too. I don't understand why you didn't use the way you did with the Rx line - when connected via the 3k3 resistor to ground, was there any signal on that output during startup?

mpier · Answer

The first resistor from the left you have connected in reverse. The second is too large or the internal pull-up is too powerful. Count for yourself what came out at the base of the transistor. If pin 16 is in the high state (google it) you will sooner burn it out than pull it down to ground. You probably don't want to do that very much.

How to eliminate transient state of I/O pins during ESP8266 module boot?

Post #1

Post #2

Post #3

Post #4

Post #5

Post #6

Post #7

Post #8

Post #9

Post #10

Post #11

Post #12

Post #13

Post #14

Post #15

Post #16

Post #17

Post #18

Post #19

Post #20

Post #21

Post #22

Post #23

Post #24

Post #25

Topic summary