Tips on designing project in IoT

[tomascrespo]

@ledsyn

How much current do the whole sensor package draw?

When awake aprox 150mA, but it fluctuates between 110-170mA.

[ledsyn]

IRLB8721PbF
This one is overkill for the small current, but it's a good jellybean MOSFET that works with 3V3 and you can stock them for future use. I have no experience in ESP, but just to be safe add a resistor between the ESP and the FET. Since you don't do fast switching you can make it ~1k.

[ledsyn]

@tomascrespo

Low side switch could be done three ways:

1. With a BJT NPN transistor
2. With a MOSFET N channel
3. With a BJT NPN driving a MOSFET N channel

I have found a MOSFET in my draw (a K3567) and I have test option 2, but there is a problem: this MOSFET requieres 5v at gate, and ESP8266 provides it 3.3v (through a GPIO). So I need another MOSFET (perhaps IRL540 or IRLZ44N) or... using a BJT transistor as a driver of the Mosfet (option 3).

I have found some NPN transistors SS8050 in the same draw, but I am not sure if I could use this NPN transistor to drive the K3567 MOSFET. This is I want to do:

2 is the best choice - BJT's have losses.

If you go with the BJT and FET you got there, don't make it a permanent solution. That FET have a high RDSon (1.7 Ohm) while IRLB8721PbF have 0.013 Ohm at 4.5 V!

[tomascrespo]

Hi again

I waiting to receive my IRLB8721PbF MOSFETs, so I have tried with a BS170 MOSFET, but with bad results.

These are the connections I have made:

As you can see I am trying to control (with the low side switch) the RS485 and the steep up booster, but all the voltage of the circuit raises to 4.3v and is not stable. Do I have done somethinkg wrong?

I have put two resistors, one to protect the GPIO (470 Ohm) and one to pull down the transistor (1K Ohm)

Thanks

[ledsyn]

Hey there!

Yeah, those FETs have horrible RDS(on).

This is pasted from the datasheet, not prettified:
RDS(ON) Static Drain−Source On−Resistance VGS = 10 V, ID = 200 mA All − 1.2 5

1.2 Ohm is the normal value, and 5(!!) worst case, when driven by 10 V. Then you decrease voltage down to 5, or even 3.3 ? which makes RDS(on) even worse. There's no remedy for this one I'm afraid. You could possibly get away with less resistance if you use 2 in parallel.

[tomascrespo]

Do you think that is the problem? An high Rds(on) does not explain why I am having a voltage increase in the rest of the circuit, isn't it?

My reads of the battery voltage (through ADC pin) were working great without the transistor, but with the transistor is always 1024, even if the battery were 3.4v, because with the transistor the voltage raises to 4.30v

Do you thing this will be OK with a IRLB8721PbF mosfet?

Anyway, the sensors support vcc from 4.5 to 35v. So I could increase the booster voltage to 12v to avoid the mosfet Rds(on), isn't it?

[ledsyn]

I'm sorry, I was not focused Thought you had a voltage drop

I think I've spotted at least one problem now. Did you have a multimeter? If so I want you to disconnect both the solar panel and the battery. Put the MM in continuity mode and measure between each GND (I say GND even if it's shown by a -) of the TP4054. Tell me if the MM beeps and in that case which points?

[tomascrespo]

I have disconnected all from TP4056, solar panel, battery, and my circuit. Now I realize you didn't tell me to disconnect the circuit.

Anyway the measurements are:

• from - to OUT- = 0 (it beeps)
• from - to B- = 1
• from OUT- to B- = 1

I will repeat the measurement with my circuit connected

[noobmastha]

I think that you have some kind of gnd routhing problem also. I can`t clearly see where is your main ground reference point, or star ground?

[tomascrespo]

Is battery powered. Main reference ground point is the negative pole of the battery, isn't it?

[ledsyn]

The circuit can stay, just disconnect solar and battery complete.

[ledsyn]

I noticed that too, but in this situation I believe there shouldn't be common GND between all 3 circuits. Look at a picture of TP5056 and you'll see what I mean. To be sure, I asked Tips on designing project in IoT - #28 by ledsyn

Result is back in #29, that the battery have it's own GND. Here it spells trouble; boost converter is connected to the battery, and in turn connected to other parts in the circuit, effectively @tomascrespo now have a circuit inside a circuit.

The solution is simple: Move the boost converter to the output from TP4056.

[tomascrespo]

Very good appreciation. Thanks, I am going to try

[tomascrespo]

Hi again,

I hace done some new test today. I looks that neither the TP4056 nor the RS485 has nothing to do here.

I have removed all I could, so now the circuit is only:

• Battery
• Wemos D1 mini Pro
• Step up booster (connected to 3.3v regulated output of D1 mini pro)
• Sensors

I have even removed the RS485 module

The results:

• all connected without low side switch: stable voltage
• with low side in step booster but without connecting the sensors: stable voltage
• with low side in the step booster and connecting the sensors to he output if the step up booster: unstable voltage (when the low side switch is activated)

So the problems comes when I power the sensors (even no connecting data lines) using the low side switch.

Some new though?

[ledsyn]

@tomascrespo

Some new though?

Yes and no, but first; when you talk about low side switch you mean the BS170 MOSFET right?

If it is - when just using the boost converter, the high RDS(on) doesn't matter because of the tiny current flow. Then you put load on the boost converter, and now you get a significant Vf (voltage drop) over the BS170. The voltage fluctuation you see, that's the boost converter being unstable now that there's not enough current to drive it. I guess the BS170 gets hot too.

[tomascrespo]

Hi again @ledsyn the BS170 does not get hot, but likely you are right about RDS(on).

Today I have replaced the BS170 N-Channel MOSFET with a IRLB8721PBF N-Channel MOSFET and now the low side switch is working perfectly.

So I have tested again with two differents BS170, but with any BS170 the voltage of the circuit fluctuates when gate is high and sensors are connected. With the IRLB8721PBF the voltage remains stable, so I am going to use this MOSFET.

Anyway I would understand why?!

Lots of thanks for your help and effort

[noobmastha]

It would be a LOT easier for everybody if you draw the schematic..

@tomascrespo

Step up booster (connected to 3.3v regulated output of D1 mini pro)

What are you powering with step up booster?

[ledsyn]

Hey!

Good to hear the new FET do the trick! About the BS170, even if it doesn't get hot (I guess the current is not enough) it acts like a resistor if you don't drive the gate high enough. To be honest, I don't understand the voltage fluctuation either, some feedback takes place. Either you just accept it, or .. I recommend you to accept it

I say we call it even, take a well earned beer or glass of wine or whatever you fancy. Do you happen to have some other questions; open a new thread to keep things tidy.

Take care!

[tomascrespo]

@noobmastha

would be a LOT easier for everybody if you draw the schematic..

These are the connections

But now I am using a 10K resistor for the GPIO gate, and a 100K resistor to pull down the gate. I have used this resistors with BS170 and with IRLB8721PBF

[tomascrespo]

Perhaps the BS170 MOSFET was opening and closing constantly, who knows...

Thanks! Of course I will enjoy it

Take care