Why Is My DIY Lie Detector Not Working?

[Dan90]

Sorry if I post this in the wrong section.

I am attempting to build a "simple lie detector" circuit which is floating around on the internet with a simple google search, but I am having trouble physically building the circuit and obtaining results.

I've drawn the circuit in a simulation program to have an idea what values to be expected, yet I can't obtain the same results.

The list of components I am using are

- 33k Resistor
- 5k Pot
- 1k Resistor
- 2x 3k Resistor
- 2N3565 Transistor
- 1uF 63V Electrolytic Cap
- 9V battery
- 1ma Analogue ammeter

I am connecting all the components via a mini bread board with 30 rows and 5 columns each side of a breaker.

The following is a list of the way each component is connected and the position on the breadboard

http://img529.imageshack.us/img529/9174/picture1ik.png

The way that everything is connected looks fine to me, but I've never tried to build any electronics before, so I'm not sure if I have done everything correctly or have even plugged everything into the right spots. When setup the meter should move when the subject is 'lying' by taking into account the change in skin resistance.

I just used a small resistor in parallel with the capacitor in an attempt to have the meter move, but to no avail. No change in amperage at the meter spot was detected with a multimeter either.

Any feedback would be appreciated, especially on where to place the components on the breadboard.

Cheers,

Dan

Schematic of the circuit

 

[vk6kro]

Can't help you wire it up. You might have to do that. Just check that the components are connected OK. Make sure you can see that they are actually connected to each other where they should be.

The circuit looks OK. The meter should be a centre zero type, though. It should work well enough to test the circuit.

As a test, try joining the two electrodes together with the pot at each end of its travel.

Is the meter reading the same when you remove the battery as when it is running?

Make sure the transistor is connected the right way. Each wire must be connected properly.
The battery must be connected the right way around too.

There is a lot more to try. Let us know how you get on with the above tests.
Can you borrow a multimeter and a soldering iron?

 

[negitron]

This is just a simple differential current amplifier.

There are three components (not counting the battery) which must be connected the right way or it won't work. There is the meter itself, of course. Mind the + and - markings. The capacitor is also polarized. Electrolytics are usually marked with a black stripe or a - marking on the negative terminal. The transistor has three terminals and each must be correctly connected. The base (the one to the left of the vertical bar in the schematic symbol) must connect to the - terminal of the cap. The collector (the topmost terminal) must connect to the - terminal of the meter. And finally, the emitter (bottommost terminal with the arrow on its end) must connect to the bottom terminal of the lowermost 3k resistor.

 

[waht]

The most difficult part is to connect the leads of the transistor right. Here is a data sheet that shows which lead is Base, Emitter, and Collector.


http://gifpdf.ic-on-line.cn/0032/2n3565_25548000001.gif

 

[vk6kro]

I found a numbering system for breadboards.

It looks like they are numbered from top to bottom. Then they have on each row, two sets of 5 holes that are joined together. These are labelled a,b,c,d,e and f,g,h,i,j.
So, on any row, say row 5, a,b,c,d,e are all joined together

I just tried to adapt this to your numbering system. The transistor for instance is connected to rows 5, 6 and 27. Nothing else connects to any of these rows.
R1 is connected between rows 2 and 10 but nothing else connects to any of these rows.

So, I doubt if it is wired correctly.

Have a look at the attached diagram and see if you can understand what is happening there.

Edit: repaired an error in the diagram. Tempting to simplify it a bit, but I'll leave it as is for now.

 

[Dan90]

Thank you for the help guys. I will have to try your suggestions after I finish work this afternoon.

With the breadboard, I tried to connect through the columns rather than rows, I'm not sure if they can work this way tho. I will try to connect them similar to the diagram that vk6kro has posted of the breadboard.

I do have access to a multimeter and soldering iron. I tried taking the ammeter out, and connecting the multimeter there set to measure amps. I never did get a stable value, nor one which was expected. Turning the pot didn't affect the reading, so something wasn't working correctly, then again I wasn't sure if I had it wired correctly. Setting the multimeter to measure voltage gave me a value so small I wasn't sure it was correct either.

If i can't make the components work in the breadboard I was going to attempt to solder the joins together to see if they work then, but I'm not sure if that will work either.

Again thanks for the help guys. I'll post again what my results are later tonight once I've had another go at testing the circuit.

 

[Dan90]

I found a numbering system for breadboards.

It looks like they are numbered from top to bottom. Then they have on each row, two sets of 5 holes that are joined together. These are labelled a,b,c,d,e and f,g,h,i,j.
So, on any row, say row 5, a,b,c,d,e are all joined together

I just tried to adapt this to your numbering system. The transistor for instance is connected to rows 5, 6 and 27. Nothing else connects to any of these rows.
R1 is connected between rows 2 and 10 but nothing else connects to any of these rows.

So, I doubt if it is wired correctly.

Have a look at the attached diagram and see if you can understand what is happening there.

Edit: repaired an error in the diagram. Tempting to simplify it a bit, but I'll leave it as is for now.

I can follow the majority of the diagram. The only thing that has me confused is the placement of the pot on your diagram.

Are all three wires of the pot connected right where the dark black squares are? I'm not sure what the arrows down from there mean, is it something similar like the electrodes?

It is a very detailed diagram btw, thanks for posting it.

 

[vk6kro]

The pot is only used as a variable resistor, so one of the ends is not connected to anything.
There is no need to even connect it if you don't want to.

The pot would be used a lot, so it would be a type you could mount on a panel of a box, not on the breadboard.

I showed connections to external components like that. Those black things are header pins and you can get plugs for them, or you could just poke wires into the breadboard.

The battery, meter, pot and electrodes have to be external to the breadboard.

 

[Dan90]

Gave the circuit another test using your diagram, but unfortunately I didn't get any results. I would think that I still have something connected up wrong so far.

Instead of the analogue ammeter I just set my multimeter to amps mode and used it in place.

This is another list of the places which I connected the components. I used a total of 8 jumper leads on the breadboard. I can't quite tell where I have gone wrong, I feel like I am getting closer though :)

http://img504.imageshack.us/img504/6073/onlinepartslist.png

 

[vk6kro]

There are a couple of mistakes in the jumper leads.

Jumper 3 should go from E7 to F7
Jumper 8 should go from J8 to I14

AND there should be a new jumper from C9 to C11 which I forgot to include.

Your multimeter should be set on about 10 mA full scale.

If you still can't see anything, measure the voltages at all points, measured relative to the negative terminal of the battery.
Especially at H7 H8 and H9 and at A5 without the meter connected.

You should be able to trace this now. Check the layout against the circuit so you can see how it works.

 

[Dan90]

Cheers, i'll check that tomorrow morning. I guess I picked a somewhat difficult circuit for a first project.

This will feel like the greatest achievement for me to date if I can get it working.

Will post results tomorrow morning.

 

[vk6kro]

OK,

You may not have noticed that this is a bridge circuit, so the current in the meter will reverse direction.
So, if you are using a digital meter, you need to be watching the polarity of the current.

You can probably use the meter on the voltage range and set it to 0 to 10 volts. All it is really doing is to compare the voltage at the collector of the transistor with that at the junction of the two 3.3K resistors. So, if this was equal, the meter would read zero. Otherwise it would read positive or negative.

Your analog meter would try to read backwards if the current was negative, so that is why it really should be a centre zero meter. For now, you should be able to see it trying to read backwards.

 

[Dan90]

Thanks for picking up on those errors. I made the changes and added in the extra jumper wire. I connected the multimeter instead of the analogue meter and actually obtained a ampere reading. I took the multimeter off and reconnected the analogue meter and I have a reading.

All I need to do now is obtain some decent 'electrodes' rather than bare wire to do some testing. So far when I adjust the pot I can't get the resistance high enough to "zero" the ammeter. I'm putting this down to my cheap and quickly made electrodes from 2 pieces of wire.

Thanks for your help so far Vk6kro

 

[vk6kro]

What happens if you join the electrodes together?
Should be able to get a balance if you do that, because the transistor should start conducting.
The bridge only has a limited range of balance, though. The transistor has to have a resistance between 1K and 6 K.

A source of electrodes would be teaspoons. Just wrap the bare wire tightly around the metal handle of the teaspoon many times and twist it so that it is secure. Then tape the spoon to your "victim" with the convex side towards the skin.

 

[Dan90]

You are right, when the "electrode" wires are connected together the meter does reach zero. What I wanted to try was to connect another pot where the electrodes were to see if I could balance it that way, just didn't have one because I wasn't at my house when I got it working.

The spoons are a good idea, will see what kind of results I can get with those tomorrow night.

In theory, wouldn't skin resistance be much higher than 1-6k ohms. I was under the impression that it would be around 1M Ohms? Though that is not right over the transitor though is it?

While on the electrodes, if 1 inch is roughly how far apart they should be, is that measuring from say the centre of the spoons themselves? Or from say the closest edges to one another?

 

[vk6kro]

When you adjust the pot with the electrodes joined together, does the meter go through zero and try to read backwards?
Sometimes analog meters have a screw adjustment on the front that let's you set the needle position. If it does, try to get the meter reading as far towards the centre of the scale as it can. This way, you will be able to see negative movement of the needle.

Yes, skin resistance would be much higher than 6 K (maybe 250K) but this is only supplying base current to a transistor which will then have a much lower resistance from collector to emitter.

No idea about electrode spacing. You might have to fiddle with that. Try it on your own arm.
Do not put it on opposite arms because of possible shock hazards.

I'm starting to think that the pot should be there instead of the two 3 Ks. Might give you a better range of control.

 

[Dan90]

I'm not too sure if the meter goes through zero, I haven't had a chance to play with it since Monday night, will get a chance this arvo.

I do know that I can touch the wires together making the meter reach zero then scale the pot so the meter will go up positive away from zero.

The pot where the two 3ks are might be a good idea. The original design of the circuit used two 4.5 volt batteries rather than the one 9 volt and two resistors, so I'm going to guess that they lower the voltage down in a similar way to having 2 batteries.

Still, if I was to try a second pot would say a 6k pot be adequate? And would that be connected with the 2 outer pins where the ends of the 3k resistors are with the center pin being where the closer ends of the 3k resistors meet?

 

[vk6kro]

Pots don't usually come in small steps. If you can find a 5 K LINEAR pot it would be OK.
That should give you a much greater range of adjustment, especially if you leave the old pot in place.

Yes, you would connect it as you describe.

Have a look for a centre-zero meter as well.

"ARVO".... Are you Australian?

 

[Dan90]

Yes, I'm an Aussie. I live about 40 minutes from Brisbane.

I'll take a look for a center-zero meter, It would be helpful, but I should be able to make do with my existing one. If the current goes below zero it does drop down so you 'can' see the change in amperes, but you don't get an actual value.

I'll try to get a pot tomorrow night so I will be able to have a play around with tomorrow night.

 

[vk6kro]

Hi, I'm in Perth.

Jaycar have some pots for $2.45. Catalog No RP-7508 for a 5K one.

They don't have any centre zero meters but their normal ones are about $17 so maybe you can make do with the one you've got. DSE and Altronics are no better.

Your device could be used as a soil moisture detector. This may be handy after all your relatives refuse to be tested on a lie detector!

 

[Dan90]

Will head down to jaycar tomorrow and see if I can pick one up. It should be the same, if not similar to the pot I already have.

The meter I am using did come from jaycar. I would imagine someone would have the center-zero meters, which would be much more handy in providing detail, but as long as the meter 'moves' shows the subject is 'lying.'

Then again its not going to be super accurate, but none-the-less a fun little project.

 

[vk6kro]

Why not use the pot you already have? Especially if it is a linear one.


Just move it to the position of the two 3.3K resistors and put one of the 3.3K resistors where the pot was.

The centre position of the pot will balance the bridge when the transistor has a resistance of 4.3 K (1 K + 3.3 K ).

 

[Dan90]

I did try replacing the 2 3kohm resistors with the pot then moving one of the resistors to where the pot was. I didn't appear to have anything happen with adjusting the pot, so I changed it back because I had someone helping me to try to test skin resistance. Some meter movement is a bit better than none.

I will try swapping the two around later once I get home. I am really thinking about going with the moisture detector idea just as something more practical. Something in say a house flower pot, or something like that.

I'll post the results after I try swapping the resistors and the pots placement again.

 

[Dan90]

Sorry for the slow reply, went on a brief holiday.

Anyways, I tried the soil moisture monitoring this arvo and it seems more responsive to play with. I tried it at first with the original setup of the circuit. So far I've only dumped a handful of dirt from the garden in an ice-cream container and added water. What happened was the resistance went down slightly from 0.6mA to 0.5mA on the meter with the pot adjusted all the way to one side (sorry I'm still not sure which is high and low). I then adjusted the pot so the meter read 1mA then proceeded to fill the container with water. The meter went down to around 0.2mA which showed the resistance definitely lowered.

I am going to try it again with the pot in place of the two 3k ohm resistors. I have moved the pot to the resistor places and one 3k ohm resistor to the place of the pot. I haven't tested this with dry soil yet, but I have noticed if I have the wires apart without touching together in any way, when the pot is adjusted down at a constant rate, the meter will 'drop' to the side once it reaches below 0.2mA at a much higher rate than falling previously. This also is true when raising the amperage by adjusting the pot in the other direction. This doesn't happen while connected togher (like through the water). What would cause this?

I'm going to go try with dry soil now, I will post the results from this with a smaller container (plastic cup).

 

[vk6kro]

As a water moisture meter, it could be used without the transistor. Just remove the transistor and take probes from the place where the collector was and from the ground (-ve lead of the battery ).

Rather than using the actual meter readings, you could use this circuit as a bridge. So, you would be always trying to get the meter to read zero by varying the pot. This is why you need to have the pot where the two 3.3K resistors were.

You can put a knob on the pot and mark the position for different values of resistance and then compare these markings with dry soil, slightly moist soil, to overwatered house potplants.
If the pot is right up one end, the sample is very wet (one way) or very dry (the other way).

You could also just use the OHMs range on a multimeter to measure this. However Bridge circuits are very important and this is a chance to learn about them.

As a source of probes, some car windscreen wiper blades have nice stainless steel reinforcement strips in them and these are discarded when a new wiper blade is fitted. Your local garage probably throws out lots of them.




Do you know about voltage dividers? If you put 3 resistors across a voltage source, one third of the supply will appear across each resistor. 1/10th if it was 10 resistors and so on.
These voltages add up to the supply voltage. A pot gives you almost continuous voltage variation.

If you have two such voltage dividers, the voltage between one point on one and one point on the other will be zero if the voltages at those points are the same relative to either power supply end.
This will happen if the ratio of the resistance above that point to the resistance below that point is the same for each divider.

When the meter reads zero, the voltages on both voltage dividers is the same at the measuring points. If one part of one divider is not known, you can measure it by using a bridge like this.

Have a read of the following, if this is all news to you:
http://en.wikipedia.org/wiki/Wheatstone_bridge

 

[Dan90]

I'll try making the circuit work without the transistor first. I don't fully understand where the probes would go if you removed the transistor. I'll post the placements of what I currently have and what I think I should do:

Currently working places with the pot in place of the two 3.3k ohm resistors.

http://img19.imageshack.us/img19/4309/currentgo.png


Potential places for probes without the transistor.

http://img503.imageshack.us/img503/1275/notransistor.png


Wouldn't taking out the transistor then make the capacitor useless? or would you just move the j8 i14 jumper to j9 i14?



Voltage dividers is tied with Kirchoff's Voltage law isn't it? Well because if you start at one point in a simple circuit and move around the sum of the voltages are = 0. Thats sort of how I understand it.

 

[vk6kro]

I think you could leave the transistor in place and just connect the probes to J7 and J11.
That way the old lie detector would still work if you wanted it to.

You could push the probes through a bottle cork so they are held a fixed distance apart.
A good trick is to insulate the probe all except for 2 cm of the tip. That way, you can test the conductivity at a depth in the soil. Heatshrink tubing is good for this.

Yes, this would mean some of the components wouldn't be doing anything, but they wouldn't be doing any harm either.

The pot and the new position for the 3.3 K resistor seem OK. This may need to be changed.

I have attached a version of this circuit. It may not follow your description because I only checked the new bits of the circuit. I couldn't fit the 3.3 K on the diagram between H4 and H5.

I added a resistor in series with the meter as a precaution. Since you have a 3.3 K, you could use that.

Yes, Kirchoffs Law explains it. Kind of.
Just so you know the voltages depend on the resistors and you can find the ratio of the soil resistance to 4.3 K by the position of the rotor on the pot.https://www.physicsforums.com/attachments/20761

 

[Dan90]

Been a while since I've looked at it. I did manage to create probes out of two old push bike wheel spokes. It works properly now. I've been using sand to play around with it, a fair bit cleaner than soil.

Anyway, I was just playing with it and I noticed that if the capacitor is removed, there doesn't seem to be any change in the readings that the circuit is giving without the probes connected, I checked this in simulation software and the circuit itself. Would that possibly mean the capicitor need not be included for the circuit to function?

Just curious about that, its keeping me awake :P

Here is an image of how I have it connected from the simulation software (the capacitor is backwards in the diagram, was trying to see if anything would happen if I connected it backwards):
http://img395.imageshack.us/img395/1010/circuitwizardscreenie.png

 

[vk6kro]

You can remove the capacitor, but it isn't doing any harm.
It was there when your device was a lie detector to make it less affected by interference when it wasn't connected to anything.
I'd leave it there but connect it the right way around.

Incidentally, there should be a resistor in series with the meter.
With the pot at the top end of its travel and the transistor conducting heavily, the meter could get destroyed.
Doesn't matter in the simulator, but the real meter needs a series resistor.

 

[Dan90]

I'll add one of the 3.3k Ohm resistors in series with the meter like you suggested earlier.

By destroying the meter do you mean bending the needle by passing to much current in either direction through making the needle hit the wall? Or damaging more then just the needle?

 

[vk6kro]

The first thing that would probably happen with a meter is that you bend the needle. You can straighten a needle but it never looks as good after that. :)
Then, depending on the current, you might burn out the moving coil or one of the spiral springs.
If the meter has shunts, these will burn out too, with lots of smoke. (Shunts are resistors in parallel with the meter movement to increase the current capability of the meter).

It is actually not a bad idea to put two Silicon diodes across the meter so they face in opposite directions. This stops the voltage across the meter rising above about 0.6 volts but doesn't affect the accuracy of the meter at low currents.
You would still need a series resistor even if you included diodes.

Moving coil meters are very expensive and worth looking after.