Question on the Plants Watering Watcher circuit

[JLB1]

I have built the Plant Watering Watcher circuit presented under Projects, "science related" circuits and works to a certain extent. Problem is that I find that the LED has only two states ( ON or OFF ), it does not dims gradually as the description states.

As a matter of fact I think that the circuit, as it is, will only put
the LED ON or OFF, as mine does.

I am mistaken ? It really should dim gradually ?
If so, I will really appreciate if you can explain why.

I plan to use it as a conductivity tester for salts disolved in water , in order to measure conductivity in micro Siemens, but not just High or Low. but continuously.

Helpwill be appreciated.
JLB.

 

[audioguru2]

The project is here:
http://www.electronics-lab.com/projects/science/008/index.html

JLB,
I see your problem. I think that a capacitor is missing, either from the right-side probe to ground, or across the rheostat. Can you try it with about 1nF? Please let us know.
The author's web-site URLs don't work to see if there is an update, so I e-mailed him with my query.

 

[ante1]

The URL works but there is no update. The circuit and description is the same.

Ante :

 

[audioguru2]

JLB,
The author of the project replied saying that an additional capacitor is not required, and that he has had several correspondents state that they are satisfied.
He also recommends using only Motorola ICs in order to correct a "failure to perform at 3V" problem.

I don't believe that the LED can dim without using Pulse-Width-Modulation which requires the time delay created by an additional capacitor in combination with the soil's resistance.
Relying on a single manufacturer's part to correct a problem that is caused by a poor choice of logic family (regular CMOS, not guaranteed to work using less than a 3V supply) is bad design.
I will analyse PWM to dim the light, and the use of a 74HC00 IC which is guaranteed to work down to a 2V supply, and will allow more output current resulting in a brighter LED.

Have you tried adding a capacitor to ground at the right-side probe yet?

 

[MP1]

audioguru, the circuit is a comparison of 2 square waves. A square wave coupled through the resistance of the probes and the untouched inverse signal. The difference gives you the dimming of the LED. It is not a single pulse which turns on the LED. Nor is it a gated action which turns on the LED. When both signals are the same strength, the LED is off (plenty of water between the probes), the signals cancel each other. When the inverse signal is the high signal (dry soil), the LED is on full. Any variance between these points will give a variance in the brightness of the LED. The pot will trim this reaction. Adding capacitors all over the place is not going to give you any result. Pulse width modulation would only complicate a simple circuit with unneeded hardware.

Time for JLB to go through his circuit and re-check things. I agree with the author that 3V is right at the edge for digital and some parts with loose tolerances will not perform. Does the LED work? Have you tried to raise the voltage to 3.5 or 4V just to see if it helps? What type of probes are you using? Have you re-checked the wiring for the NAND schmitt trigger? Checked values of the parts? Have you substituted any parts?

MP

 

[audioguru2]

MP,
You made an incorrect statement:
"nor is it a gated action which turns on the LED". Of course it is gated action, the LED driver is a Nand gate (actually 2 gates paralleled) which requires both inputs to be high in order to light the LED.
You are correct that the LED is on when the inverse signal is high (dry soil), because both inputs have the same signal (the lower input through the pot). As the soil becomes conductive (wetter) the lower input's voltage becomes lower due to the voltage-divider of the pot and the soil. Even though this voltage is lower, it is still above the threshold and therefore is a valid high signal.
So the LED remains at full brighness until the voltage-divider reduces the lower input's voltage to its threshold. No dimming.

When the lower input's voltage is slightly below or very much below its threshold, then it is a valid low signal and the LED is completely off.

If a capacitor to ground is added to the lower input, then its signal will be delayed (until the capacitor charges up to the threshold). If the capacitor is chosen so that it charges quickly with dry soil, then it will charge slower when the voltage divider has a lower voltage. This will create PWM of the LED.
Therefore the LED will dim.

 

[MP1]

The LED brightness is controlled by the differential between the gate circuits. Not a gate action directly turning on the LED.

But, whatever. Believe what you want. :


MP

 

[audioguru2]

Believe?
Isn't the LED driven by a logic gate (actually 2 paralleled to get more ouput current)?
Can't you see that the output of a logic gate is either high (LED off) or low (LED on) but never in between?
This gate is a Schmitt trigger, so even feedback (which the circuit doesn't have) won't make it linear.
There is no way that this circuit can dim the LED without Pulse-Width-Modulation, that it doesn't have, yet.

The author of this project states that the signal at the probes alternates, in order to eliminate a plating action at the probes which would be caused if the probes had DC (or assymetrical AC). But the original circuit uses assymetrical AC (causing probe plating) so that the LED is on most of the time (PWM) because the standard CMOS driver has a very low output current when operating with a 3V supply.

The 74HC132 which I propose using, is a quad 2-input Schmitt trigger Nand gate which is guaranteed to operate with a supply voltage as low as 2.0V. It has much more output current. It can deliver a perfect 50-50 signal to the probes and still have enough output current to drive the LED brightly when required.

Why a 3V supply? A single lithium cell, 2 AAA or 2 AA cells are more compact than a 9V battery and will last 10 times as long.

 

[MP1]

All you are suggesting is another circuit. Yes, a pseudo form of PWM is being used here from the cancelling operation of the two gates. The duty cycle causes the dimming effect.
When the circuit has a history of working, and you suggest to replace the main IC and several other components of the circuit, you are not fixing an inherent problem, you are designing another circuit similar to the other by using the original author's ideas. Not exactly a fix and not exactly original.
Why don't you just learn how to use a schematic draw program and post your own designs? You can call them "audiogurus XXX" where XXX is the name of someone else's project. You are free to post your circuits.

MP

 

[ante1]

Hi Guys!

Maybe it

 

[audioguru2]

Designing another circuit?
No. Just add one capacitor. That will fix JLB's complaint.

Replacing several other components?
No, the opposite. After replacing IC1, eliminate D1 and R2.

Replacing the main IC?
Why not? Most of the original ICs don't work with new batteries and they all won't work when the batteries are running down. The replacement IC is the same, but improved.

Pseudo PWM, cancelling operation and duty cycle? Where?
The oscillator has a permanent 90/10 duty cycle only so that the LED is brighter (it gets only 1mA).

Theory?
JLB saw it, I see it, why can't you see it?

 

[ante1]

Well when I think of it, why does my car only have two headlights? If there were four of them I would have twice the light in the dark! And why is it fitted with a 70liter petrol tank, a 200liter would carry me much longer? What the heck does these car manufacturers think of? But then again, my car works anyway!

Irrelevant? Oh yes!

Ante :

 

[audioguru2]

Ante,
JLB's constructed project does not work and he believes (as I do) that the circuit cannot dim its LED.
If the dimmer knob for your car's dashboard didn't work, wouldn't you complain and wonder why-not?

Your eyes' response to light is the same as your ears' response to sound: Logarithmic. Double the lights and it is just a little bit brighter. It is the same with audio amplifier power, 20W sounds just a little bit louder than 10W (only 3dB difference).
10 times the light looks twice as bright and 10 times the audio sounds twice as loud.
Since the LED in this project is operated at just 1mA, even though the author doubled the LED drivers and duty-cycled the on-time, then 10mA (if possible) would look more normal.

 

[MP1]

I was told once that practice is substance and theory is mostly air

Very true my friend. Theory is only a starting point.

If the dimmer knob for your car's dashboard didn't work, wouldn't you complain and wonder why-not?

If I made the dimmer switch myself, I would not be so quick to ask the manufacturer to redesign the switch, and especially when others have claimed it works.

MP

 

[audioguru2]

Do we have ANY projects that work properly? (2 inverters, 30V power supply, stethoscope, IR detector, 6-channel disco lights, batteries charger, roadrunner, etc. and now this one).

The authors of the above claimed that their circuit works.
I had fixes for them all.

 

[ante1]

audioguru,

My point is there are alternatives to

 

[audioguru2]

The point of this discussion is not about the LED brightening (dimming in reverse) when the soil gets drier and drier, as the project's description says that it will. When looking at the LED today, can you determine whether its brightness is more than it was yesterday (I noticed that the LED was very dim yesterday), without having a standard LED to compare it to? Therefore another improvement would be for it to have a numeric display!

The point is that JLB wants to use the project for a different function, but was relying on its description about dimming, which is extremely important to his function. He will probably have a bunch of circuits, side by side, and will see which one is brighter or dimmer.

 

[ante1]

audioguru,

Well then, this is where you come in to the picture. From scratch, design and post a circuit suitable for this job. Then it

 

[audioguru2]

Ante,
Didn't you see my post?:

Designing another circuit?
No. Just add one capacitor. That will fix JLB's complaint.

Replacing several other components?
No, the opposite. After replacing IC1, eliminate D1 and R2.

Replacing the main IC?
Why not? Most of the original ICs don't work with new batteries and they all won't work when the batteries are running down. The replacement IC is the same, but improved.

Why re-invent a new circuit when this one will work fine with the addition of only ONE capacitor?
For JLB and others who build this project, and want it to work at all with new batteries, or to have it continue functioning when its batteries are running down, and for a bonus have a much brighter LED, then simply repace its old IC with a newer and improved one.

Isn't that constructive and are simple solutions to the problems?

The author recently told me that he may update his circuit with my improvements when he has time in a couple of months.