Author Topic: Question about alarm clocks and transistors  (Read 3043 times)

0 Members and 1 Guest are viewing this topic.

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Question about alarm clocks and transistors
« on: October 30, 2011, 04:57:08 PM »
Hi everyone, I'm working on a project (which isn't strictly to do with robotics) just for fun, and I've run into a bit of a wall about what to do next.

Well, I have an analog alarm clock, which simply beeps at a given time. So I opened it up, removed the buzzer and connected the two wires to my multimeter (as I have no access to an oscilloscope). I saw that the reading alternated between 0V and around 0.250V (on the d.c. reading). What does this mean? I want to straighten out this voltage, and use it to 'turn on' a transistor; the transistor will then turn on a 3V d.c. motor. I plan on using a 9V battery as the power supply. I think I would need the motor to stay on as long as the clock gives the d.c. signal.

My questions regarding this are:
1) Can I simply use one capacitor to straighten out the d.c. wave? If so, how do I know what capacitance to use? If not, how would I straighten it out?
2) What type/make of transistor is appropriate for this?/How do I know which one to choose?
3) How should I connect the whole setup? A schematic would be helpful, but if not I can understand. I have some basic knowledge of electronics (though I'm still in high school) and I do know, as an overview, how transistors work.
Oh and experience I have: I've successfully built the $50 robot from the tutorial on this website, and I've also gone on to modify the board and use it for a different project - controlling the lights in my house with sound (by clapping). I've only worked with microcontrollers so far; I guess this will be my first real analog-only project.

I hope I'm making sense, and if you need any other information please let me know. I appreciate you taking the time to read this and hope you can help, thank you very much!

Vignesh R.

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #1 on: October 30, 2011, 09:04:24 PM »
Hi,

Well, I have an analog alarm clock, which simply beeps at a given time. So I opened it up, removed the buzzer [...]
By "buzzer" you mean...?
Electromechanical buzzer with integral drive electronics.
Electromechanical buzzer withouy integral drive electronics.
Piezo buzzer with integral drive electronics.
Piezo disk in resonator housing.

Is there a coil near the buzzer?


[...] and connected the two wires to my multimeter (as I have no access to an oscilloscope). I saw that the reading alternated between 0V and around 0.250V (on the d.c. reading). What does this mean?
Assuming you used a digital meter, with a display update of around 3Hz to 4Hz, it means that you cannot rely on what you measured, as the signal is most likely an audio frequency of between 500Hz and 4kHz.
If the voltmeter is True RMS, the 0.25V will be the average voltage of the waveform, if it isn't, all bets are off, but likely a 'scope would reveal a low duty cycle signal of several volts.

0.25V could be 2.5V pulses with 10% duty cycle or 5V pulses with 5% duty cycle.

Try something like an 1N4148 diode from the output (cathode) to a cap (anode) of eg. 100nF (with the other end of the cap to ground) and measure over the cap.


I want to straighten out this voltage, and use it to 'turn on' a transistor; the transistor will then turn on a 3V d.c. motor. I plan on using a 9V battery as the power supply.
Why would you want to drive a 3V motor with 9V - do you carry a grudge for this motor?  ;D


I think I would need the motor to stay on as long as the clock gives the d.c. signal.
You think? Well you better be a little more certain before you start building, as that choice influence how it should be done ;)

How did the buzzer sound - a continuous tone of a certain frequency, an intermittent signal (like beep-beep-beep) or what?


My questions regarding this are:
1) Can I simply use one capacitor to straighten out the d.c. wave? If so, how do I know what capacitance to use? If not, how would I straighten it out?
You need a diode for rectifying the signal and then a cap to integrate it somewhat.
If the output amplitude is low, a voltage doubler may be needed.


2) What type/make of transistor is appropriate for this?/How do I know which one to choose?
That cannot be answered without knowing the motor current.
Measure the motor current when running at 3V and with the intended load.
Don't try to run it at 9V!


3) How should I connect the whole setup? A schematic would be helpful, but if not I can understand. I have some basic knowledge of electronics (though I'm still in high school) and I do know, as an overview, how transistors work.
Emitter to ground, collector to one side of the motor, the other side of the motor to +3V, base through a resistor to the rectified signal - and place a reverse biased diode over the motor terminals, or the transistor will die soon.

If the signal is intermittent, you could use it to either trigger a 555 directly, or use a 555 as a missing pulse detector, to give it a constant output as long as the signal keeps going.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #2 on: October 30, 2011, 09:52:58 PM »
Thank you for such a quick and thorough reply!  ;D

By "buzzer" you mean...?
I'm not entirely sure what type of buzzer it is; I took two photos which I hope will help identify it:
Photo 1: https://docs.google.com/open?id=0BzicjaDkK8wTNjFhMDAwOWEtYjAxYS00NmQ5LTllZWQtNTU2YTM2NGE4Zjhl
Photo 2: https://docs.google.com/open?id=0BzicjaDkK8wTZjAzYzA2YTctYmVkOC00NTJlLTkyNGQtNWMzYTBlMTg1ZGM4

Is there a coil near the buzzer?
No, as far as I can tell the only coil in the entire clock is the one that is connected to the second hand.

If the voltmeter is True RMS, the 0.25V will be the average voltage of the waveform, if it isn't, all bets are off, but likely a 'scope would reveal a low duty cycle signal of several volts.
Hmm.. Well we do have an oscilloscope at school. Do you think it's necessary that I ask permission to borrow/use it?

Try something like an 1N4148 diode from the output (cathode) to a cap (anode) of eg. 100nF (with the other end of the cap to ground) and measure over the cap.
When you say the other end of the cap should go to ground, you mean it should go to the anode of the 'buzzer' yes? (The buzzer, by the way, does have + and - markings on its terminals) It's because I haven't yet disconnected the buzzer from the wires..

Why would you want to drive a 3V motor with 9V - do you carry a grudge for this motor?  ;D
Haha, of course not. It's just that I found a 3V motor which is the perfect shape and size (I believe it's a CD-player motor) for my purposes, except the torque isn't high enough at 3V. Which is why I had planned on using a 9V battery as the source. But now I think I'll reduce the load and test it again at 3V. Would it still be bad to the motor if I gave it 9V say for 2-3 minutes (not in one go) per day?

You think? Well you better be a little more certain before you start building, as that choice influence how it should be done ;)
I only said "think" because I didn't want to seem too demanding by saying "I want this.." But yes, I do want the motor to stay on as long as the clock gives the signal.

How did the buzzer sound - a continuous tone of a certain frequency, an intermittent signal (like beep-beep-beep) or what?
It is an intermittent signal - every second, it goes beep-beep-beep. Basically 3 beeps per second.

That cannot be answered without knowing the motor current.
Measure the motor current when running at 3V and with the intended load.
Don't try to run it at 9V!
I will post the motor current once I get to measuring it - it's already late here, so I'll most probably do it tomorrow.
EDIT: I just measured the current of the motor running on two 1.5V batteries, and it comes to around 22-24 mA. Does that help?

...place a reverse biased diode over the motor terminals, or the transistor will die soon.
I understand that this is necessary, but would you mind explaining to me why? Also, why does the diode have to be reverse biased (which I take it means it is connected the opposite way)?

If the signal is intermittent, you could use it to either trigger a 555 directly, or use a 555 as a missing pulse detector, to give it a constant output as long as the signal keeps going.
I was in fact also thinking about using a 555 timer, but I'd like not to go down that path as far as possible - If, however, I do end up needing to, I'll have to do some reading up on the timers first, since I've never used one before.  :)
« Last Edit: October 31, 2011, 08:46:32 AM by TheDarkLord »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #3 on: October 31, 2011, 10:44:28 AM »
Hi,

I'm not entirely sure what type of buzzer it is; I took two photos which I hope will help identify it:

They sure did. I have some of the exact same types and they're crude "loudspeakers" with a flat steel disc membrane and a conventional coil (mine measures around 16 Ohm), so the modulated signal is generated outside the buzzer (on the clock PCB).

Your buzzer may have a different coil resistance, but it shouldn't matter for the circuit needed.


No, as far as I can tell the only coil in the entire clock is the one that is connected to the second hand.

Yes, with this buzzer, there wouldn't be - had there been one, it would point toward a passive piezo disc buzzer.


Hmm.. Well we do have an oscilloscope at school. Do you think it's necessary that I ask permission to borrow/use it?

No, not for getting your clock to run the motor, better save that opportunity for something that  you cannot do without a 'scope.


When you say the other end of the cap should go to ground, you mean it should go to the anode of the 'buzzer' yes? (The buzzer, by the way, does have + and - markings on its terminals) It's because I haven't yet disconnected the buzzer from the wires..

No, to the negative terminal of the buzzer, like this:

Try it without disconnecting the buzzer at first, then you'll know when some voltage should be on the cap.
What does the voltmeter reveals with this circuit?


Haha, of course not. It's just that I found a 3V motor which is the perfect shape and size (I believe it's a CD-player motor) for my purposes, except the torque isn't high enough at 3V. Which is why I had planned on using a 9V battery as the source. But now I think I'll reduce the load and test it again at 3V. Would it still be bad to the motor if I gave it 9V say for 2-3 minutes (not in one go) per day?

9V batteries doesn't have much capacity and a relatively high internal resistance, so will be a poor choice for this. 2 (or perhaps 3) AA cells would be a more sensible solution.

Running a 3V motor on 9V would mean 3 times the voltage and since the motors resistance don't change, the current will rise 3 times as well, resulting in the motor getting 3x3=9 times as much power (which its windings weren't designed for).
With 4.5V (3 AA cells), the motor will see 2.25 times the nominal power (actually a little less, as the transistor will have a small voltage drop) and that should be tolerated for short times.

A way to find out if the voltage is OK, is to run it for say 5 seconds and then let it rest for 10 seconds, then repeat, while keeping a finger on he motor housing. If it heats up quickly, you're hurting it, but if it only gets lukewarm, increase the running times (but not the resting times) and if you can get it running constantly without it getting much above ~40°C, it is OK at that voltage.
40°C is around the point where it is too hot to hold, but you can still touch it.


I only said "think" because I didn't want to seem too demanding by saying "I want this.." But yes, I do want the motor to stay on as long as the clock gives the signal.

Ah OK. Don't be so polite that you won't get your ideas across :)
Approximately how long would you guess the pauses between the beeps to be?


...place a reverse biased diode over the motor terminals, or the transistor will die soon.

I understand that this is necessary, but would you mind explaining to me why? Also, why does the diode have to be reverse biased (which I take it means it is connected the opposite way)?

The diode is connected in the non-conducting direction across the motor, as a motor is a large inductance and as such, when you cut power to it (when the transistor stops conducting), it will generate a large (up to several hundred volts, depending on the motor, its speed, how fast the power is cut etc.) opposite polarized voltage. The diode shorts this inductive kick (flyback), keeping it from killing the transistor.
This also explains why it's reverse biased (when the motor is running, it shouldn't conduct, or it would short out the motor).


I was in fact also thinking about using a 555 timer, but I'd like not to go down that path as far as possible - If, however, I do end up needing to, I'll have to do some reading up on the timers first, since I've never used one before.  :)

That you haven't used them before might be a very good reason to get your feet wet in a controlled way. I can draw you a schematic, so it would just be putting it together and perhaps measuring a bit.
While what you need can be made with transistors, going for a 555 solution would give a neater circuit and would probably make you want to experiment a lot with this multi-purpose IC.

If you feel like dipping into the 555, I'll recommend Tony's Tutorial, as it is quite digestible if you don't try to take it all in in one go.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #4 on: October 31, 2011, 02:58:45 PM »
Hi again,

No, to the negative terminal of the buzzer, like this:

Try it without disconnecting the buzzer at first, then you'll know when some voltage should be on the cap.
What does the voltmeter reveals with this circuit?

I'm going to try this in a little while (I seem to be all out of diodes) and I'll let you know what it measures. However, I have another question: why is the diode necessary? I mean I understand that it rectifies the signal, but I thought that was only for a.c. signals - that it would allow the positive voltages to go through but not the negative. Did I have a misconception of diodes? Because clearly here the signal from the clock PCB is a d.c. signal..

9V batteries doesn't have much capacity and a relatively high internal resistance, so will be a poor choice for this. 2 (or perhaps 3) AA cells would be a more sensible solution.

OK, I'll follow your advice and go with 3 AA cells, as just using 2 doesn't give me enough torque.

Approximately how long would you guess the pauses between the beeps to be?

It beeps really fast, but if I had to make an approximation, by looking at the second hand on the clock, every second something like this happens:
1. Silence
2. Beep
3. Beep
4. Beep
5. Beep
6. Silence
So continuously, there are four beeps with a third of a second's gap in between the set of three. The time between two beeps is tiny.

This also explains why it's reverse biased (when the motor is running, it shouldn't conduct, or it would short out the motor).

Ah, that makes perfect sense now I think about it. Thanks! :)

That you haven't used them before might be a very good reason to get your feet wet in a controlled way. I can draw you a schematic, so it would just be putting it together and perhaps measuring a bit.
While what you need can be made with transistors, going for a 555 solution would give a neater circuit and would probably make you want to experiment a lot with this multi-purpose IC.

If you feel like dipping into the 555, I'll recommend Tony's Tutorial, as it is quite digestible if you don't try to take it all in in one go.


Yes, I would be very grateful if you could draw up a schematic for me. Also, to aid my learning process, could you briefly explain what is happening at each stage of the circuit? Thanks ;D
And I just read through that website you sent me. I think I have a conceptual understanding of the timer now, though I'll have to go back and study it again, as the explanations for Pin 2 (Trigger) and Pin 3 (Output) completely lost me  :-\

And thank you once again! You're really very knowledgeable and I hope that one day I'll also be able to help others like you are now. :) I remember when I was having troubles with the $50 robot, you were really helpful as well. So thank you very much, I'm really grateful for the help! ;D

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #5 on: October 31, 2011, 04:30:28 PM »
Hi,

I'm going to try this in a little while (I seem to be all out of diodes) and I'll let you know what it measures. However, I have another question: why is the diode necessary? I mean I understand that it rectifies the signal, but I thought that was only for a.c. signals - that it would allow the positive voltages to go through but not the negative. Did I have a misconception of diodes? Because clearly here the signal from the clock PCB is a d.c. signal..
It's a pulsating DC and if you added just a capacitor it would charge and discharge with the signal, only delayed a bit by the RC filtering, where the output impedance would be the "R" and since it's very low, the cap would follow the signal pretty closely and you wouldn't have gained anything.

By adding the diode as well, only the peaks of the signal going above ~0.6V will be charging the cap, but as it will only discharge through leakage and the input impedance of your meter (which is high, probably around 1GOhm), you get a DC value with a smal amount of ripple and this value will be almost identical to the peak DC value which is what you want to find.


It beeps really fast, but if I had to make an approximation, by looking at the second hand on the clock, every second something like this happens:
[...]
So continuously, there are four beeps with a third of a second's gap in between the set of three. The time between two beeps is tiny.
So, to recap, the max. time without sound in the time span where the motor should be running is around 1/3 second?


Yes, I would be very grateful if you could draw up a schematic for me. Also, to aid my learning process, could you briefly explain what is happening at each stage of the circuit? Thanks ;D
Sure, but I'll need the value of the peak voltage from the clock and the motors current consumption with the load attached (at 3.0V or 4.5V).


And I just read through that website you sent me. I think I have a conceptual understanding of the timer now, though I'll have to go back and study it again, as the explanations for Pin 2 (Trigger) and Pin 3 (Output) completely lost me  :-\
Nobody can grasp each bit of that amount of information in the first go, but it helps to study fig. 3, with the two comparators referenced to the voltage divider (the 3 x 5k resistors) that gave the chip its name.

The 555 is really easy to use and when using it with logic values (eg. 0V and 5V) it doesn't hold too may surprises - and the output can sustain 200mA, making it a tough little bastard.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #6 on: October 31, 2011, 09:01:52 PM »
It's a pulsating DC and if you added just a capacitor it would charge and discharge with the signal, only delayed a bit by the RC filtering, where the output impedance would be the "R" and since it's very low, the cap would follow the signal pretty closely and you wouldn't have gained anything.
Oh, that makes sense now - thanks!

So, to recap, the max. time without sound in the time span where the motor should be running is around 1/3 second?
Exactly. If you like I could take a relatively low quality video showing the clock 'in action'...

Sure, but I'll need the value of the peak voltage from the clock and the motors current consumption with the load attached (at 3.0V or 4.5V).
I set up the circuit as in the schematic you posted earlier, and my multimeter, set on the d.c. voltage setting, read about 5V, give or take a few hunded mV. It fluctuated between that value and close to zero, in pretty much the same timing as earlier, without the capacitor or the diode.

I was just getting around to measuring the motor's current consumption when I ran into a mechanical problem - the 'blades' attached to the motor (basically its load) fell apart and right now I'm working on fixing that. But I will try to give you the current consumption as soon as I can.

Nobody can grasp each bit of that amount of information in the first go, but it helps to study fig. 3, with the two comparators referenced to the voltage divider (the 3 x 5k resistors) that gave the chip its name.
Yes, fig. 3 is pretty helpful. It all makes a lot more sense now.

Just another quick question; would using a 555 timer make the circuit be classified as a digital one? Not that I'm against creating a digital circuit, but I'm just curious.

Thanks again! :)

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #7 on: November 01, 2011, 04:00:07 PM »
Hi,

Exactly. If you like I could take a relatively low quality video showing the clock 'in action'...
A better idea would be to record just the sound and post an MP3, as I could then see the exact pulse- and pause times, but I don't think it's necessary, so let's save it for if the circuit acts up.


I set up the circuit as in the schematic you posted earlier, and my multimeter, set on the d.c. voltage setting, read about 5V, give or take a few hunded mV. It fluctuated between that value and close to zero, in pretty much the same timing as earlier, without the capacitor or the diode.
Great - plenty of signal to tap then ;D


I was just getting around to measuring the motor's current consumption when I ran into a mechanical problem - the 'blades' attached to the motor (basically its load) fell apart and right now I'm working on fixing that. But I will try to give you the current consumption as soon as I can.
So, it's a "fan alarm" you're making?


would using a 555 timer make the circuit be classified as a digital one? Not that I'm against creating a digital circuit, but I'm just curious.
Well, the 555 do use digital logic on the inside (the flip-flop), but is classified as a linear device and while the output, the reset pin and the discharge pin are digital in function, I wouldn't call the entire IC digital.

However, I gave the circuit some more thought and while it can be made with a 555, it would mean a constant drain on the battery, and it would need to be the CMOS version to run reliably when 3 AA cells dicharge over their life.
So, I made another circuit, with no current drain except whenl the alarm signal drives it and the circuit just awaits the motor current measure to get the output drive right and I can post it.

Don't think it's a waste to learn about the 555 though. It's a chip with a multitude of applications.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #8 on: November 01, 2011, 05:55:22 PM »
...so let's save it for if the circuit acts up.
Sure! :)

So, it's a "fan alarm" you're making?
No, not exactly. I'm making more of a "water alarm" - the motor drives a pump which is meant to spray water in my face. The mechanism was working great until yesterday, and now it's being a little bit of a pain to fix, but I will do it. I thought of making it because 1) I have huge problems waking up early for school and 2) Simply because I wanted to make something electronic and functional. The pump basically works on principles of air pressure; the motor has blades attached to it and is inside a small plastic can - when the blades spin, the air pressure is reduced and water is sucked in from the bottom end of the can, which is immersed and has a pipe attached to it. The water is then passed through a longer pipe, which ends at a small 'shower head' conveniently hanging over my bed :P The problem I'm having is that the blades keep falling off - I've tried hot glue, super glue, rubber cement. I also found that the pump isn't very powerful this way. Not completely relevant, but would you happen to know of any alternative ways of pumping water? I think I got the idea for this pump from Instructables..

I wouldn't call the entire IC digital.
I know that "digital" means discrete values and "analogue" means continuous ones - so would ANY circuit that has a programmable chip be considered digital? And would any circuit that does not have programmable parts be considered analogue in general?

So, I made another circuit, with no current drain except whenl the alarm signal drives it and the circuit just awaits the motor current measure to get the output drive right and I can post it.
I will fix the mechanical bits of my project and get back to you on the motor current within the next few hours. *crosses fingers*

Don't think it's a waste to learn about the 555 though. It's a chip with a multitude of applications.
No, of course not! In fact from searching for more information on the 555, I came across some simple but interesting circuits. Once I finish this project, I'll probably build a continuity tester or try to achieve the fading LED effect using the 555.

Thanks!

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #9 on: November 01, 2011, 08:25:05 PM »
Hi,

No, not exactly. I'm making more of a "water alarm" - the motor drives a pump which is meant to spray water in my face.
Radical!  ;D


I thought of making it because 1) I have huge problems waking up early for school and
I think it's a general problem among experimenters and I'm sure you know the real cure for that one (and I don't mean dropping out of school) :P ;D

Beware of any side effects, like mold and such and make sure your bed sheets dries out completely each day, or you may face serious respiratory issues in a fairly short time!

The only side effect of getting enough sleep is less time to experiment, but your brain will pick up more and faster, when you're not partially zombified.


If only I followed my own advice...  ;)


2) Simply because I wanted to make something electronic and functional.
A very good reason IMO ;D


The pump basically works on principles of air pressure; the motor has blades attached to it and is inside a small plastic can - when the blades spin, the air pressure is reduced and water is sucked in from the bottom end of the can, which is immersed and has a pipe attached to it. The water is then passed through a longer pipe, which ends at a small 'shower head' conveniently hanging over my bed :P
Add some shampoo and you can call it Bed&Shower  :D
There's a large range of other ways I'd rather wake up, but each to his own.


The problem I'm having is that the blades keep falling off - I've tried hot glue, super glue, rubber cement. I also found that the pump isn't very powerful this way.
Hot glue is a lot of different things, but the generic type is best used on cardboard and wood and won't be very good on plastic (which I imagine you're using) of any kind (well, depends on the area the glue will cover, but hot glue is a very misapplied substance.
Super glue (Cyanoacrylate) is good on hard plastics if the parts going together is a very close fit, as the glue cannot fill gaps of more than a around a tenth of a millimeter max.
Rubber cement, if it's anything like contact adhesive needs a certain area to create a strong bond.

Not knowing what exactly you want to join, but you may consider some two-component epoxy glue (Araldite or similar) - it's good for a lot of materials.


Not completely relevant, but would you happen to know of any alternative ways of pumping water? I think I got the idea for this pump from Instructables..
There's plenty of ways to shave that goat, but I assume you don't wanna build it from raw material?
I have a couple of small pumps that's not much more than a small cylindrical electromagnet with a tight coiled spring each end. They take AC to work and they're very silent. I got them as surplus some 20 years ago and I wouldn't know where to get them today, but perhaps somebody else wil know of a source.
Automotive window washer pumps can deliver quite an amount, but they are a little power hungry for a battery application.
There's a some plastic toys that could maybe be hacked, like automatic soap bubble machines, toy fountains etc. - always be on the lookout for usable parts.

Perhaps a gravity feed with a solenoid valve would be a better way to go. Install a tank/bottle up high with a tap and have a solenoid valve open. when you need the shower.


What about an alternative to water?
Flashing a bright lamp (or high power LED/LED cluster) pointed towards you at up to 10..15Hz will reach your brain through your eyelids and eyes, possibly forcing your brain waves into a rythm that wakes you up.
A vibrating motor in a box tucked under your pillow (used by many deaf people).
One of the more radical (even worse than water) ideas was posted in a magazine some 20 years ago, aimed at lucid dreaming triggering. It was a shocker circuit with two electrodes mounted on a leg and pumping the battery voltage up to some hundred volts of AC - quite effective, but not a good first (or tenth) project for safety reasons of course.
A mechanism dropping ping pong balls on your head might be sufficiently annoying as well.


I know that "digital" means discrete values and "analogue" means continuous ones - so would ANY circuit that has a programmable chip be considered digital? And would any circuit that does not have programmable parts be considered analogue in general?
There's programmable analog circuits as well and programmable may mean different things (like the PUT, which is a Programmable Unijunction Transistor, programmed by changing a resistor value).
And most logic chips isn't programmable at all, so programmability isn't a good criteria.

There's a shadow area, where a chip is neither/nor (or rather both) that falls under the term Mixed Signal, indicating that it has both analog and digital circuitry.
Other than that, if it ain't 100% digital, it's probably classified as a linear (analog) chip


I will fix the mechanical bits of my project and get back to you on the motor current within the next few hours. *crosses fingers*
Don't loose any sleep over it ;D
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #10 on: November 03, 2011, 10:51:24 AM »
Radical!  ;D
Why thank you  ;)

I thought of making it because 1) I have huge problems waking up early for school and
I think it's a general problem among experimenters and I'm sure you know the real cure for that one (and I don't mean dropping out of school) :P ;D
The real cure.. Not experimenting? In that case I don't think I'll ever be cured :P

If only I followed my own advice...  ;)
Well, you certainly seem to have picked up enough to help yourself and others ;D I really appreciate all the help, thanks!

Not knowing what exactly you want to join, but you may consider some two-component epoxy glue (Araldite or similar) - it's good for a lot of materials.
I managed to fix it - still using hot glue, but now made the 'blades' out of aluminium from a soda can; the curves really help the air flow!

There's plenty of ways to shave that goat, but I assume you don't wanna build it from raw material?
No, I'm afraid you assumed wrong - I did want to build it out of raw materials, because otherwise the mechanical part of my project would be very little and I really do need to improve my 'mechanical' skills :)

Perhaps a gravity feed with a solenoid valve would be a better way to go. Install a tank/bottle up high with a tap and have a solenoid valve open. when you need the shower.
I thought of that, but again it seemed too easy; I did want a bit of a challenge to keep me engaged.

Flashing a bright lamp (or high power LED/LED cluster) pointed towards you at up to 10..15Hz will reach your brain through your eyelids and eyes, possibly forcing your brain waves into a rythm that wakes you up.
A vibrating motor in a box tucked under your pillow (used by many deaf people).
One of the more radical (even worse than water) ideas was posted in a magazine some 20 years ago, aimed at lucid dreaming triggering. It was a shocker circuit with two electrodes mounted on a leg and pumping the battery voltage up to some hundred volts of AC - quite effective, but not a good first (or tenth) project for safety reasons of course.
A mechanism dropping ping pong balls on your head might be sufficiently annoying as well.
Haha, I like that last idea! Maybe once I get into the habit of sleeping with a towel and start drying myself off after the alarm to go back to sleep ;D I also like the idea of the LED cluster - maybe I could develop into that when I finish with this.

Don't loose any sleep over it ;D
Well, after pulling two all-nighters (sarcasm) I finally managed to finish it, and I must say, running the motor for 10 seconds with a 9V battery, it works even better than before. I'll post a video when I've finished the electronics too. Right now I'm waiting on some AA batteries, since I'm apparently all out of new ones - so I can now safely say the current reading will be present in my next post! :)

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #11 on: November 03, 2011, 11:37:37 PM »
Hi, so I just measured the motor's current draw.

Just so you can make sure my setup was correct, this is what I did:
I connected the negative wire from my battery holder to one lead of the motor.
The positive wire from the battery holder I connected to the positive lead of the multimeter.
I touched the negative lead of the multimeter to the remaining motor lead.

I got readings of 30-40 mA without any water present between the blades of the motor. After adding water, the motor drew a current of around 197-199 mA, and at times it went even higher than 200 mA - my multimeter can only measure a maximum of 200 mA of current..  :-\ Oh, and I was using 3 AA batteries (1.5V each) connected in series as the power source; the batteries are brand new.

Hope this isn't too late, and that you can still help! :)
Thanks!

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #12 on: November 04, 2011, 08:19:34 AM »
Hi,

The real cure.. Not experimenting? In that case I don't think I'll ever be cured :P

No, I meant getting to sleep an hour or so earlier  ;D


Well, you certainly seem to have picked up enough to help yourself and others ;D I really appreciate all the help, thanks!

You're welcome :)


I managed to fix it - still using hot glue, but now made the 'blades' out of aluminium from a soda can;

Heh, you've got to post a photo of that contraption ;D

Judging from your time zone, you're either in the US or somewhere in South America, but this sound more like how they'd do it in Africa, where soda cans and beer caps is turned into whatever - sometimes with great ingenuity.
Hmm. "aluminium" is how we spell it this side of the Pond... Where do you live?


the curves really help [...]

Ain't that always the case  ;)


No, I'm afraid you assumed wrong - I did want to build it out of raw materials, because otherwise the mechanical part of my project would be very little and I really do need to improve my 'mechanical' skills :)

OK then, I was more thinking in terms of what tools are available to you. if you have access to a drill pre4ss, a lathe and perhaps a mill - and chunks of brass or aluminum, you can make almost any kind of pump, but i didn't get that impression?


I thought of that, but again it seemed too easy; I did want a bit of a challenge to keep me engaged.

When you want to learn, a challenge is great, as it force you to think laterally.
When you need to make something that works reliable, the less challenges, the better.


Haha, I like that last idea! Maybe once I get into the habit of sleeping with a towel and start drying myself off after the alarm to go back to sleep ;D I also like the idea of the LED cluster - maybe I could develop into that when I finish with this.

I wasn't too serious about the ping pong balls ;D
Would you like having to collect the balls each day and reload?

The LED cluster OTOH is worth a try.


Well, after pulling two all-nighters (sarcasm) I finally managed to finish it, and I must say, running the motor for 10 seconds with a 9V battery, it works even better than before. I'll post a video when I've finished the electronics too. Right now I'm waiting on some AA batteries, since I'm apparently all out of new ones - so I can now safely say the current reading will be present in my next post! :)

OK, I changed the circuit to 6V to give the motor a bit more push. If it gets hot in use, back down to 4.5V or use some 1N400x diodes in series with the motor to lower the voltage.


I got readings of 30-40 mA without any water present between the blades of the motor. After adding water, the motor drew a current of around 197-199 mA, and at times it went even higher than 200 mA - my multimeter can only measure a maximum of 200 mA of current..  :-\ Oh, and I was using 3 AA batteries (1.5V each) connected in series as the power source; the batteries are brand new.

Super!


Hope this isn't too late, and that you can still help! :)

Too bad, it's just 5 minutes too late  :P

Nah seriously, why would it be too late?
I got an email recently, going something like "I got a new car, how could I change the circuit to work with that" - turns out I made him a design around 5 years back - a circuit for controlled re-fueling (en route) when driving with a caravan, but this guy expected me to remember all the details from around a thousand designs and a couple of PC's back and the files might not even exist anymore (real men(TM) doesn't do back-up 8)  They just cry when everything is lost :'(). Luckily he had kept the schematic, so I managed to reconstruct it (it was a quite elaborate analog controller, as he was allergic to digital design).
Turned out it wasn't too late either, although it took me a while to re-enter the thought process of when I originally made it.


I made it into this .PDF-file complete with PCB layout and circuit explanation.
If it acts up, just complain ;D
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #13 on: November 04, 2011, 05:27:38 PM »
No, I meant getting to sleep an hour or so earlier  ;D

Oh I see, but that would still mean less time to experiment :P

Heh, you've got to post a photo of that contraption ;D

Judging from your time zone, you're either in the US or somewhere in South America, but this sound more like how they'd do it in Africa, where soda cans and beer caps is turned into whatever - sometimes with great ingenuity.
Hmm. "aluminium" is how we spell it this side of the Pond... Where do you live?

Photo, once I complete it :)

Haha, well it does help to be "material-smart" - the end product does not always look professional, but it functions decently. I actually live in Mexico. I take it from your name that you're of Scandinavian origin, perhaps from Denmark? How come you've come about to have so much knowledge on pretty much everything? :)

OK then, I was more thinking in terms of what tools are available to you. if you have access to a drill pre4ss, a lathe and perhaps a mill - and chunks of brass or aluminum, you can make almost any kind of pump, but i didn't get that impression?

No, well I don't have access to any power tools, but I was thinking more along the lines of building a pump out of raw materials that I can find around the house - for instance I used a plastic film canister as the pump housing.

When you want to learn, a challenge is great, as it force you to think laterally.
When you need to make something that works reliable, the less challenges, the better.

Well in this case, I wanted to learn by overcoming the challenge of making something reliable ;D

OK, I changed the circuit to 6V to give the motor a bit more push. If it gets hot in use, back down to 4.5V or use some 1N400x diodes in series with the motor to lower the voltage.

All right.. but does that mean I need to measure the current again using 4 AA batteries?

Nah seriously, why would it be too late?
I got an email recently, going something like "I got a new car, how could I change the circuit to work with that" - turns out I made him a design around 5 years back - a circuit for controlled re-fueling (en route) when driving with a caravan, but this guy expected me to remember all the details from around a thousand designs and a couple of PC's back and the files might not even exist anymore (real men(TM) doesn't do back-up 8)  They just cry when everything is lost :'(). Luckily he had kept the schematic, so I managed to reconstruct it (it was a quite elaborate analog controller, as he was allergic to digital design).
Turned out it wasn't too late either, although it took me a while to re-enter the thought process of when I originally made it.

Well, I did make you hold on to that schematic for a few days..so my apologies for that. Wow, that must feel really great - coming across someone you helped 5 years ago. What's amazing is that the person still remembers you! :P

I made it into this .PDF-file complete with PCB layout and circuit explanation.
If it acts up, just complain ;D

Thanks! The explanations are really helpful.
I've a few questions regarding that schematic:
1. Is it important what resistors I use? Would it make a difference if I use the carbon ones as opposed to the metal film ones?
2. Is the capacitor an electrolytic one?
3. The shop from where I usually buy all my components doesn't seem to hold the BD140 transistor (Q2) so what would a suitable alternative be? It doesn't matter if there aren't any alternatives; I could go into the city and they have hundreds of component shops all lined up, at least one of them would have it ;D
4. Should I disconnect the buzzer from the terminals now?

Thank you so much for helping me out! I'll be sure to mention you in the video that I'm going to make of my contraption in action :)

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #14 on: November 05, 2011, 01:06:13 PM »
Hi,

Haha, well it does help to be "material-smart" - the end product does not always look professional, but it functions decently. I actually live in Mexico. I take it from your name that you're of Scandinavian origin, perhaps from Denmark? How come you've come about to have so much knowledge on pretty much everything? :)
As long as your supply of materials last... ;D
Yes, I'm a Dane and living in Denmark (well googled ;))
I started at an early age, picks up something each year and a whole lot of years have passed since then.


No, well I don't have access to any power tools, but I was thinking more along the lines of building a pump out of raw materials that I can find around the house - for instance I used a plastic film canister as the pump housing.
There's a lot of fairly simple DIY centrifugal pumps on the web, so there should be some that would suit the materials that you can come by.

If you need more rigidity out of soda can aluminum, "bead rolling" can be done with simple hand tools - a technique you should test, if soda cans are a preferred material. It really stiffens up a thin sheet (can even be tested with paper).

Perhaps you should start a web site dedicated to "recycling techniques"?


All right.. but does that mean I need to measure the current again using 4 AA batteries?
No, the BD140 can handle 1.5A, so it should be fine.


What's amazing is that the person still remembers you! :P
To be honest, I think he just kept the email correspondence.


1. Is it important what resistors I use? Would it make a difference if I use the carbon ones as opposed to the metal film ones?
Not a bit and a fair tolerance is allowed, so if there's a value you don't have, just try one close.


2. Is the capacitor an electrolytic one?
Yes.


3. The shop from where I usually buy all my components doesn't seem to hold the BD140 transistor (Q2) so what would a suitable alternative be? It doesn't matter if there aren't any alternatives; I could go into the city and they have hundreds of component shops all lined up, at least one of them would have it ;D
TIP42 springs to mind, but any PNP transistor able to handle at least 1A and having a current gain (hFE) of at least 40 should do.
If you tell me what they have, I can point one out - do you have web links to any of the shops?


4. Should I disconnect the buzzer from the terminals now?
Only if you want it silent, but it shouldn't matter for the function.

You could add a louder alarm if needed (if if your family can bear it :D)
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #15 on: November 05, 2011, 01:32:03 PM »
Yes, I'm a Dane and living in Denmark (well googled ;))

Googled? No. :P I actually used to live in Sweden, and came across a few people with the same name as you, and I think most of them were Danish.

Perhaps you should start a web site dedicated to "recycling techniques"?

Perhaps. :)

2. Is the capacitor an electrolytic one?

Yes.

And on the .pdf it says the cap is rated for 10V. Would I be fine using one rated for 16V?

TIP42 springs to mind, but any PNP transistor able to handle at least 1A and having a current gain (hFE) of at least 40 should do.
If you tell me what they have, I can point one out - do you have web links to any of the shops?

Of course, here you go:
http://www.steren.com.mx/catalogo/search.asp?s=transistor&x=0&y=0&search_type=prod
That's all the transistors they have. But like I said, if there aren't any suitable alternatives, I can go to the city and they're very likely to have it there.

Only if you want it silent, but it shouldn't matter for the function.

You could add a louder alarm if needed (if if your family can bear it :D)

Alright! :) Thanks again, I'll try finishing the circuit by this weekend and let you know how it all goes ;D

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #16 on: November 05, 2011, 05:46:14 PM »
Hi,

Googled? No. :P I actually used to live in Sweden, and came across a few people with the same name as you, and I think most of them were Danish.

Yes that sounds likely. Swedes use the "ö" where we use the "ø", but I have never met a Swede with my name (no matter the spelling).
So, you're a real globetrotter then  ;D


And on the .pdf it says the cap is rated for 10V. Would I be fine using one rated for 16V?

Yes, you can always use one rated for a higher voltage.


That's all the transistors they have. But like I said, if there aren't any suitable alternatives, I can go to the city and they're very likely to have it there.

They do have the 45V BD136 (the BD140 is 80V), and that should do.
Alternatively, there's the TIP32 that can handle twice the current. It's unclear whether it's a 40V or a 60V version, as they write 40V on the page, but TIP32A which is a 60V version, but both would work.
They sure seems expensive though. Denmark is very expensive with components, but I can get a BD136 at around $1.20 locally - I'm sure you can get at least 5 for $1 at eBay.


Alright! :) Thanks again, I'll try finishing the circuit by this weekend and let you know how it all goes ;D

It would be gold if you could set up a video cam to start right before the shower hit you and the next few minutes onwards :D :D :D
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #17 on: November 05, 2011, 06:15:14 PM »
They do have the 45V BD136 (the BD140 is 80V), and that should do.
Alternatively, there's the TIP32 that can handle twice the current. It's unclear whether it's a 40V or a 60V version, as they write 40V on the page, but TIP32A which is a 60V version, but both would work.

Great, thanks. I'm going to spend another day trying to find the BD140, but if I can't I'll settle for one of those.

They sure seems expensive though. Denmark is very expensive with components, but I can get a BD136 at around $1.20 locally - I'm sure you can get at least 5 for $1 at eBay.

Really? I don't find them that expensive. Oh wait, maybe it's because you're confused with the dollar sign. The prices on that site are not in U.S. dollars, they're in Mexican pesos - many people get confused since both currencies use the "$" symbol. 1 U.S. dollar is about 13 Mexican pesos, so I'd say they're quite affordable.
And no shop can beat eBay at prices and availability :P

It would be gold if you could set up a video cam to start right before the shower hit you and the next few minutes onwards :D :D :D

Ah, now that that idea has entered my head, I'll find a way to do it ;D Though when it comes down to it, it might just be easier to fake sleep and record it.. :)

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #18 on: November 05, 2011, 08:13:06 PM »
Hi,

Really? I don't find them that expensive. Oh wait, maybe it's because you're confused with the dollar sign. The prices on that site are not in U.S. dollars, they're in Mexican pesos - many people get confused since both currencies use the "$" symbol. 1 U.S. dollar is about 13 Mexican pesos, so I'd say they're quite affordable.
Ah, that's quite another story then - Didn't know that the $-sign were used for Pesos.


Ah, now that that idea has entered my head, I'll find a way to do it ;D Though when it comes down to it, it might just be easier to fake sleep and record it.. :)
Easier yes, but it would be acting and not half as fun (or realistic).
Well, I fully understand if you'd like to keep such a moment to yourself - I know I would ;D
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #19 on: November 06, 2011, 10:03:44 AM »
Hi

Well, I fully understand if you'd like to keep such a moment to yourself - I know I would ;D
Not exactly.. You were such a great help I would have to share it at least with you :)

Well I put together the circuit on a breadboard...and it doesn't work. So I started checking voltages with my multimeter and found that there is a fairly large voltage going into the BC547, but apparently a much smaller one coming out of it. Could the transistor be defective? I'm positive I've connected it the right way around - I referred to 3 different datasheets to be sure.
Also, I'm using the BD136-10 as Q2 since I couldn't find BD140 - I didn't buy it from Steren though. I've tried replacing the 330uF cap with a 470uF one but still nothing. I did notice that when I removed the capacitor, the motor gave off a tiny, almost inaudible beep similar to the buzzer.
I also noticed that the collector of the BC547 (Q1) goes to Q2 while its emitter goes to ground. Why is this? Shouldn't it be reversed? (And by the way I did try reversing it - still nothing)

What could I try now?

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #20 on: November 07, 2011, 03:26:21 PM »
Hi,

Well I put together the circuit on a breadboard...and it doesn't work.
OK, is it a solder-less breadboard?


So I started checking voltages with my multimeter and found that there is a fairly large voltage going into the BC547, but apparently a much smaller one coming out of it.
The BC547 doesn't put out a voltage, it pulls the base of Q2 towards ground (through R5)
If you have your voltmeter at the collector of Q1, it should be close to the battery voltage and when the buzzer beeps, it should be close to ground (less than 0.5V).


Could the transistor be defective?
While you cannot exclude a defect transistor, I think it's unlikely, if you didn't mistreat it .


I've tried replacing the 330uF cap with a 470uF one but still nothing.
That only changes the timeout - too small a value and it will follo the beeps more rigidly (and stop in between each beep)


I did notice that when I removed the capacitor, the motor gave off a tiny, almost inaudible beep similar to the buzzer.
So, at least something gets through. The reason the motor can beep is because the windings expand and contract a small amount at the frequency it's driven by.


I also noticed that the collector of the BC547 (Q1) goes to Q2 while its emitter goes to ground. Why is this? Shouldn't it be reversed? (And by the way I did try reversing it - still nothing)
Emitter to ground, collector to Q2 as it should be.
The base of an NPN must go ~0.7V positive relative to the emitter, to allow current from collector to emitter.


What could I try now?
First, you didn't swap the Buzz- and Buzz+ by chance?

What's the voltage from ground to the top of C1 when it beeps?

Next, remove at least one wire from the input (both and you don't have to worry over the clock interfering).
Then touch the positive input (K1, pin 1/Buzz+) to the positive side of the 6V battery (K2 pin 1) Does this get the motor going?

If not, try removing C1 and try again - did that help?

If not, measure each resistor, to check if they're what you think - it's sometimes hard to tell some colors apart (like orange/red, brown/red, brown/violet, green/blue).

Still no luck, then measure the voltage at base and collector of both transistors (with K1 pin 1 and K2 pin 1 connected together) and post it here - then we take it from there.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #21 on: November 12, 2011, 11:09:27 AM »
OK, is it a solder-less breadboard?
Yes, I actually wasn't aware that there were breadboards you could solder to.

If you have your voltmeter at the collector of Q1, it should be close to the battery voltage and when the buzzer beeps, it should be close to ground (less than 0.5V).
Just tested that, and it's how it's supposed to be.

First, you didn't swap the Buzz- and Buzz+ by chance?
I'm positive I didn't  :P

What's the voltage from ground to the top of C1 when it beeps?
I take it that is the same as the voltage across C1 when it beeps? Well it varies from around 0.9V to 1.1V.

Then touch the positive input (K1, pin 1/Buzz+) to the positive side of the 6V battery (K2 pin 1) Does this get the motor going?
Yup, it runs like a charm with the buzzer input removed.

If not, measure each resistor, to check if they're what you think - it's sometimes hard to tell some colors apart (like orange/red, brown/red, brown/violet, green/blue).
I actually did measure each individual resistor before plugging it in. But I just double checked and they're all the right values from the schematic.

I noticed that by removing R4, which connects the base of Q2 to +6V, the motor runs almost perfectly as I want it to - save for the fact that the motor runs in bursts of spin-slow down-spin-slow down. But I think I could just try a higher value capacitor value since you said that is what affects the timeout? My question is, would eliminating R4 have any effect on the circuit, e.g. blowing a transistor or something?

Thanks.

P.S. Sorry for the delayed post, I've been a little caught up in school work all week.

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #22 on: November 12, 2011, 01:28:51 PM »
Hi,

Yes, I actually wasn't aware that there were breadboards you could solder to.
A breadboard can be just about anything, from a wooden board with brass/copper nails, an unetched PCB that you glue tiny bits of the same to for soldering pads etc.
The solderless boards can be a pain with certain circuits, especially when they're some years old.


If you have your voltmeter at the collector of Q1, it should be close to the battery voltage and when the buzzer beeps, it should be close to ground (less than 0.5V).
Just tested that, and it's how it's supposed to be.
Sounds like Q2 is a bit marginal in its gain spec then, so it needs some attention.


I take it that is the same as the voltage across C1 when it beeps? Well it varies from around 0.9V to 1.1V.
Yes and that sounds reasonably with this setup.


Then touch the positive input (K1, pin 1/Buzz+) to the positive side of the 6V battery (K2 pin 1) Does this get the motor going?
Yup, it runs like a charm with the buzzer input removed.
Although the measurements on Q1 collector OK'd them, I think we should change R2 and R3 then.


I noticed that by removing R4, which connects the base of Q2 to +6V, the motor runs almost perfectly as I want it to - save for the fact that the motor runs in bursts of spin-slow down-spin-slow down. But I think I could just try a higher value capacitor value since you said that is what affects the timeout? My question is, would eliminating R4 have any effect on the circuit, e.g. blowing a transistor or something?
Good thinking and trouble shooting!
You cannot rule out R4, as it is used to pull the base of Q2 high when the buzzer is silent (to keep noise and radio signals from opening Q2), but it can be made larger.
Yes, a larger cap will get rid of the intermittent running (as long as the alarm clock fills it)


P.S. Sorry for the delayed post, I've been a little caught up in school work all week.
Don't be, your project is mostly for fun I assume - school is for life!  ;D


And before I forget...
R2  1k (to 1.5k
R3  10k (to 20k)
R4  10k
R5  100R

R4 is probably the most important, then R3, but changing them all should more than make up for Q2's low gain.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #23 on: November 12, 2011, 01:48:31 PM »
A breadboard can be just about anything, from a wooden board with brass/copper nails, an unetched PCB that you glue tiny bits of the same to for soldering pads etc.
The solderless boards can be a pain with certain circuits, especially when they're some years old.
Oh, I see. And this is only temporary to make sure everything works, I plan on soldering it all to a perforated PCB as the final step.

I noticed that by removing R4, which connects the base of Q2 to +6V, the motor runs almost perfectly as I want it to - save for the fact that the motor runs in bursts of spin-slow down-spin-slow down. But I think I could just try a higher value capacitor value since you said that is what affects the timeout? My question is, would eliminating R4 have any effect on the circuit, e.g. blowing a transistor or something?
Good thinking and trouble shooting!
You cannot rule out R4, as it is used to pull the base of Q2 high when the buzzer is silent (to keep noise and radio signals from opening Q2), but it can be made larger.
Thank you! :) I'm not sure I understand how noise/radio signals would open Q2.. Care to elaborate please?

Yes, a larger cap will get rid of the intermittent running (as long as the alarm clock fills it)
Up to what maximum value would it be safe/wise to try?

P.S. Sorry for the delayed post, I've been a little caught up in school work all week.
Don't be, your project is mostly for fun I assume - school is for life!  ;D
Ahh, wise words.

And before I forget...
R2  1k (to 1.5k
R3  10k (to 20k)
R4  10k
R5  100R

R4 is probably the most important, then R3, but changing them all should more than make up for Q2's low gain.
Okay, will try with these values in a little while and post back with results.

Thanks!

EDIT: I just put it all together with the new resistor values - and with a higher capacitance - and it works perfectly!  ;D I'm feeling on top of the world haha. Well thank you so much, I can't sufficiently express my thanks in words! I'm so glad you could help me, thanks again! :D

By next weekend I'll solder it all on a perf. board and maybe post a video here on SoR, I'll be sure to show it to you as well! :)
« Last Edit: November 13, 2011, 11:14:07 AM by TheDarkLord »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #24 on: November 13, 2011, 03:32:35 PM »
Hi,

Thank you! :) I'm not sure I understand how noise/radio signals would open Q2.. Care to elaborate please?
The floating base will act as an antenna, although the one you have doesn't appear to be very sensitive, but it's a cheap insurance. A lot of people building sensitive amplifiers end up with what is generally called "Radio Moscow", which is when a radio station (or more than one) breaks through the sensitive front end. "Radio 'Moscow" on Q2 wouldn't be sound of course, but rather a little tingle of water here and there :)

The higher the input impedance and the higher the gain of the device is, the more important it is to keep the base in check at all times.


Up to what maximum value would it be safe/wise to try?
The only safety concern would be whether the alarm output would be able to drive it without getting shorted, but R1 takes care of that, so the sky is the limit - it shouldn't be needed though.


EDIT: I just put it all together with the new resistor values - and with a higher capacitance - and it works perfectly!  ;D I'm feeling on top of the world haha. Well thank you so much, I can't sufficiently express my thanks in words! I'm so glad you could help me, thanks again! :D

By next weekend I'll solder it all on a perf. board and maybe post a video here on SoR, I'll be sure to show it to you as well! :)
Great  ;D

'Then you're ready for trying with light :)
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #25 on: November 18, 2011, 07:21:23 PM »
Thank you! :) I'm not sure I understand how noise/radio signals would open Q2.. Care to elaborate please?
The floating base will act as an antenna, although the one you have doesn't appear to be very sensitive, but it's a cheap insurance. A lot of people building sensitive amplifiers end up with what is generally called "Radio Moscow", which is when a radio station (or more than one) breaks through the sensitive front end. "Radio 'Moscow" on Q2 wouldn't be sound of course, but rather a little tingle of water here and there :)

The higher the input impedance and the higher the gain of the device is, the more important it is to keep the base in check at all times.

Oh I see, well that makes sense. Thanks!

Up to what maximum value would it be safe/wise to try?
The only safety concern would be whether the alarm output would be able to drive it without getting shorted, but R1 takes care of that, so the sky is the limit - it shouldn't be needed though.

Alright, I'm using 470uF and it works great so I'm going to stick with it.

'Then you're ready for trying with light :)

I'll save that for when I get hold of some free time.

I have another question regarding the project: What would I need to change in the circuit so that the motor turns on and spins for a set period of time when the alarm rings (let's say 40 seconds) before it turns off by itself regardless of the length of the alarm tone? Using a 555 timer perhaps?
« Last Edit: November 18, 2011, 07:35:56 PM by TheDarkLord »

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #26 on: November 18, 2011, 11:54:15 PM »
Hi,

What would I need to change in the circuit so that the motor turns on and spins for a set period of time when the alarm rings (let's say 40 seconds) before it turns off by itself regardless of the length of the alarm tone?

A non-retriggerable monoflop (MF), which is not able to start a new timing period, until the signal have been returned to the resting value.
A retriggerable MF, when timing out, will either retrigger or just keep its output active, if the input signal is still active.


Using a 555 timer perhaps?

That could be used, as there's no demand on high precision in the timing.
Since it would need to use an electrolytic capacitor and since they're very temperature dependent, timing would drift some with temperature.

I have attached a new schematic with 2 alternative output stages. The first uses the transistor you already have and can be used with a CMOS-555, which will ensure that it works all the way down, while the battery discharges.
The regular 555 will cease to work before the battery is run down all the way (to 0.9V/cell or 3.6V total), but since there are differences between the versions from different manufacturers, I cannot say at what voltage your specimen will give up.
The CMOS version will work all the way down (it's guaranteed to work at a voltage as low as 1.5V).
However, the CMOS version is expensive compared to the regular version, so you decide - no harm in trying the cheap one, as long as you test how far down it will work.

If you haven't got a 1M trimmer resistor or don't want it to be adjustable, use a single resistor of around 820k for VR1 and R6 to get roughly 42s - that said, you might need to experiment a bit, as electrolytic caps have wide tolerances. The formula for the timing is:

 1.1*Ct*Rt = 1.1*C3*(R6+VR1)

Plugging in some numbers:
  1.1*47µ*820k = 1.1*0.000,047*820,000 = 42.394s


Wishing you a nice cold morning shower   :P ;D
(Later you could add a circuit to automatically call in sick to your school, if you sleep through all 40s of that).
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline TheDarkLordTopic starter

  • Jr. Member
  • **
  • Posts: 46
  • Helpful? 0
Re: Question about alarm clocks and transistors
« Reply #27 on: November 19, 2011, 02:21:11 PM »
A non-retriggerable monoflop (MF), which is not able to start a new timing period, until the signal have been returned to the resting value.
A retriggerable MF, when timing out, will either retrigger or just keep its output active, if the input signal is still active.


I see, that's interesting - what exactly is a monoflop? A component or a type of circuit?

I have attached a new schematic with 2 alternative output stages. The first uses the transistor you already have and can be used with a CMOS-555, which will ensure that it works all the way down, while the battery discharges.


When you say "while the battery dicharges" does that mean it will discharge if I leave it plugged in overnight? Or just "discharge" like in the long run?

That circuit does seem very interesting - I think I will build that one instead.

The regular 555 will cease to work before the battery is run down all the way (to 0.9V/cell or 3.6V total), but since there are differences between the versions from different manufacturers, I cannot say at what voltage your specimen will give up.
The CMOS version will work all the way down (it's guaranteed to work at a voltage as low as 1.5V).
However, the CMOS version is expensive compared to the regular version, so you decide - no harm in trying the cheap one, as long as you test
how far down it will work.


What type of 555 timer is this? It doesn't actually specify.. but judging from the price, could it be the regular version?

If you haven't got a 1M trimmer resistor or don't want it to be adjustable, use a single resistor of around 820k for VR1 and R6 to get roughly 42s - that said, you might need to experiment a bit, as electrolytic caps have wide tolerances. The formula for the timing is:

 1.1*Ct*Rt = 1.1*C3*(R6+VR1)

Plugging in some numbers:
  1.1*47µ*820k = 1.1*0.000,047*820,000 = 42.394s


I like the thought of being able to adjust the time. If I'm not mistaken, a 1M trimmer resistor is a potentiometer of the same value? Which of the first three products on this page would be the most suitable?
The equation is really helpful - it's always nice to see equations applying to real life! :)

Wishing you a nice cold morning shower   :P ;D
(Later you could add a circuit to automatically call in sick to your school, if you sleep through all 40s of that).


Haha! Why thank you :D
I like the thought of that idea.. Very innovative ;D

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #28 on: November 19, 2011, 08:14:51 PM »
Hi,

what exactly is a monoflop? A component or a type of circuit?

It can be either. The established name is monostable multivibrator and it's related to the astable and bistable multivibrators.
Astable, meaning "not stable" changes its output constantly (it is an oscillator type).
Bistable, meaning "stable in two/both conditions" changes its output once for each activation (pulse) and is then stable until next activation - in daily terms it's called a flip-flop.
Monostable, meaning stable in one condition rests in one condition (either high or low, depending on exact circuit) and when activated, changes it's output in a pulse, going back to the stable condition after time out.
(Google "multivibrator tutorial" or similar to get a more in-depth look at them).
 

When you say "while the battery dicharges" does that mean it will discharge if I leave it plugged in overnight? Or just "discharge" like in the long run?

I was referring to the general use, but glad you mentioned it, because it does have an idle consumption that will drain some on the batteries - I'll  change it a bit to only power up when needed.
Oops, just realized I didn't give the 555 any supply (that's what happen when you get interrupted by mundane stuff like dinner ;D) - Kicking myself and will change that as well.
Won't be any major changes though - I'll update and repost, but its 3 a.m. so it'll be after nap-time :)


What type of 555 timer is this? It doesn't actually specify.. but judging from the price, could it be the regular version?

Yes, it's a regular (bipolar) 555. The CMOS variants usually have a C in the name.
LM555, NE555 and UA555 are the regular bipolar type 555, while LMC555, UA555C as well as ICL7555 are the CMOS variants.


If I'm not mistaken, a 1M trimmer resistor is a potentiometer of the same value?

Almost :)
A potentiometer is a "potential divider" (metering out the potentials/voltages).

A trimmer can be either a trimming potentiometer, with "trimming" meaning something to adjust once or rarely/from time to time, while a normal potentiometer is something you adjust often (like eg. the volume on a radio/HiFi).

A trimmer can also mean a variable resistor, i.e. a resistor which is not used as a potential divider.
Some trimmers are made as just variable resistors (to save a bit in large scale production), but they're mostly found in consumer electronics like clock radios, cheap radios and such. Usually, you just use trimmer potentiometers as variable resistors by shorting the wiper to one end (as in this schematic).


Which of the first three products on this page would be the most suitable?

Potentiometer #1 and #3 are regular carbon track potentiometers, so is #2, but with an added on/off switch (all the way counter clockwise the switch is off).
Neither of those are trimmers.

The next two (#4 and #5) are trimmers. Not a very good quality, but since you won't be adjusting too often, they should do fine, but don't use too much physical force on them. The difference between them is that #4 is mounted parallel to the board, while #5 is standing upright. #4 gives the most stable mounting, but takes up a bit more board space - I'd go for #4.

If you want it to be continually adjustable, you could use something like #1 of course - in that case, use as short leads as possible and connecting the metal encapsulation to either ground or V+ may be a good idea, as it will keep it from picking up noise. If the leads should act as antennas, a small cap (~10..100pF) can be used  at the board connection points.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: Question about alarm clocks and transistors
« Reply #29 on: November 20, 2011, 10:52:14 AM »
And here it is :)
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

 


Get Your Ad Here