Author Topic: How do I simulate an electric bell in QUCS?  (Read 1265 times)

0 Members and 1 Guest are viewing this topic.

Offline cogburnd02Topic starter

  • Newbie
  • Posts: 3
  • Country: us
How do I simulate an electric bell in QUCS?
« on: October 16, 2019, 02:56:52 pm »
I have an electric bell, which I would like to trigger from a couple of transistors (a smallish npn transistor switching on a larger pnp transistor switching the bell directly, with an arduino switching on the smaller first transistor. (I can post the part numbers of the transistors I was thinking about using if that'll help.))

How do I simulate this in QUCS? As a relay in series with an inductor and then in series with itself? If I am supposed to use the 'relay' part in QUCS for this, how do I measure/figure out the hysteresis voltage of the part I have? (I have a good multimeter, LCR meter, and scope.)

Here's some pictures I found on the web that look like the device I have: enclosed, (https://images-na.ssl-images-amazon.com/images/I/61Bu9utZmtL._AC_SL1500_.jpg) open (http://catalog.miniscience.com/catalog/electricity/Buzzer/Electric_Bell_Doorbell_m.jpg). And here's the wikipedia diagram (https://upload.wikimedia.org/wikipedia/commons/9/95/Electric_Bell_animation.gif) explaining the theory.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: How do I simulate an electric bell in QUCS?
« Reply #1 on: October 16, 2019, 03:55:11 pm »
Just one question: why?

I woudn't bother, just model the bell as a resistor which draws the amount of current specified on the bell's datasheet or measure the resistance, with a multimeter, if that's not available.

Add a diode in reverse parallel with the bell, to protect the transistor from the inductive back-EMF generated when it's turned off.
 

Offline cogburnd02Topic starter

  • Newbie
  • Posts: 3
  • Country: us
Re: How do I simulate an electric bell in QUCS?
« Reply #2 on: October 17, 2019, 04:24:58 am »
I'm fairly sure that the bell introduces a 'ringing' in the line (https://en.wikipedia.org/wiki/Ringing_(signal)) (or perhaps just oscillation) and I just wanted to make sure that I wasn't--at any point--driving (too much current or voltage) either of the two transistors or the arduino output pin over/out of spec.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: How do I simulate an electric bell in QUCS?
« Reply #3 on: October 17, 2019, 11:00:25 am »
I'm fairly sure that the bell introduces a 'ringing' in the line (https://en.wikipedia.org/wiki/Ringing_(signal)) (or perhaps just oscillation) and I just wanted to make sure that I wasn't--at any point--driving (too much current or voltage) either of the two transistors or the arduino output pin over/out of spec.
Have you had any problems with the circuit? Simulation can't tell you everything. There are plenty of circuits which will work in simulation, but not real life and vice versa.

To answer your question. There will be ringing, but not at the audio frequency produced by the bell. Look at how the bell works. When the bell is first turned on, the clapper will be away from the bell and the contacts in series with the electromagnet will be closed. The electromagnet is an inductor, with a certain internal resistance and capacitance, forming an RLC circuit. The inductance will probably dominate first, causing the current to initially be quite low, gradually rising, until the clapper hits the bell and the contacts open. An arc will form across the contacts and the RLC circuit will ring, probably at a frequency between a few 10s to 100s of kHz. The arc will also produce much higher frequencies, up to the microwave range and beyond, which can cause more trouble.

It's possible the spurious RF emissions could interfere with other the microcontroller, but it isn't something a simulator will be able to predict. To start with, just connect a reverse parallel diode in parallel with the bell, to suppress the high voltage produced, if the transistor interrupts the current, whilst the electromagnet in the bell is turned on. If you have problems with interference, they can be cleaned up with a snubber network (an RC circuit in parallel with the bell) and ferrite beads.
 

Offline cogburnd02Topic starter

  • Newbie
  • Posts: 3
  • Country: us
Re: How do I simulate an electric bell in QUCS?
« Reply #4 on: October 17, 2019, 09:45:50 pm »
> Have you had any problems with the circuit?

I haven't built the circuit yet. I want to make sure I'm not going to zap my transistors and arduino before building the circuit.

> not at the audio frequency produced by the bell

Yeah I figured it would be at the same frequency that the bell goes clack-clack-clack-clack-clack-clack-clack at when I hold the ringing part of the bell, but it's powered on (& 'ringing')

Thanks for the other stuff.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13217
Re: How do I simulate an electric bell in QUCS?
« Reply #5 on: October 17, 2019, 10:34:02 pm »
No. it wont even be that. It will electrically 'ring' with a decaying exponential amplitude envelope sinewave, initial amplitude set by the stored energy in the coil, its inductance and the effective stray capacitance across it (including a contribution from inter-layer capacitances across parts of the coil), at a frequency set by the inductance and the stray capacitance, at a repetition rate determined by the clapper mechanics, e.g. its inertia and the restoring spring force. To further complicate the issue the inductance will vary with how close the clapper arm is to the pole piece.  Before the damped sinewave you can expect various truncated pulses to the same amplitude caused by contact bounce.  You'll also get a burst of pulses when the clapper contact makes on the return stroke.

If you are tempted to scope it to see what's *really* happening, I strongly recommend using a scope with >150V max input rating at its BNC connector, + a BNC T connector on it and a NE-2 neon fitted to a BNC plug connected to the spare arm of the T to clamp the back EMF if your probe flashes over internally.  Then use a x10 (or better a x100) probe on the other arm of the T.  If the scope's a low voltage one e.g. +/-30V or USB connected, I really wouldn't care to probe the bell coil without a named brand HV probe with at least a 1000V rating.

Your best option is probably to design pessimistically from first principles: snub the coil with a diode + a series resistor equal to the DC coil resistance across the coil to limit the peak back-EMF to double the supply voltage without slowing down the magnetic flux decay significantly,  and choose a driver transistor with a Vceo rating (Or Vdss if you use a MOSFET) of at least *THREE* times your max supply voltage, and a current and power rating such that it can carry the max coil current (calculated from the DC resistance and the max supply voltage) continuously at the max ambient temperature your ciruit is ever likely to be in.

As I've got an old Dynamco scope with valve inputs, that can withstand high voltage transients that would probably blow a modern scope and an old Morse practice buzzer I inherited,  I took a photo of the waveform across the coil, with the buzzer running on 13.8V, and the contacts positive.   The scope input attenuator + x10 probe were set for 50 V per vertical division, DC coupled with 0V fractionally over 1 div down from the top of the screen, and the timebase was set to approx 0.25 ms per horizontal division (uncalibrated).  Its difficult to see them because of the poor contrast between most of the trace and the negative transient peaks, as I had to turn the brightness up to resolve them, but if you open the picture in a new tab, you can see the transients going off the bottom of the screen, so they must be peaking at over 250V.  You can clearly see the contact bounce on make.  Sorry, no damped sinusoid as promised - I couldn't get the timebase to trigger repeatably at a sweep speed fast enough to resolve them.

Buzzer coil resistance 4.4 ohms, inductance 10.4 mH with armature up, and 11.8 mH with it down, in firm contact with the pole piece.
« Last Edit: October 17, 2019, 11:46:05 pm by Ian.M »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf