2N3055 heat sink

[AnalogKid]

A 555 pin 3 output never is a sawtooth waveform. A pseudo-sawtooth waveform appears at the top of the timing capacitor at pin 2, but it actually is a portion of the standard exponential charge/discharge waveform of an R-C network.

There is a more simple 555 astable circuit that gives nearly 50/50 duty cycle with fewer parts.



One way to clean up your circuit is to use this circuit for both U1 and U2. To make U2 change frequencies, have U1 switch in and out an additional timing capacitor rather than an additional timing resistor. This will take one diode.

ak

 

[audioguru]

You have the pin 7 output shorted to +12v so the transistor inside at pin 7 is probably blown up.

 

[audioguru]

I forgot to say that your diodes have no part number. They should not be high current very slow 60Hz 1N400x rectifiers, instead use 1N4148.

 

[gary350]

My parts arrived in the mail today so I finished the circuit and tested it. It is built like the circuit in TTL book. There seems to be no way to upload video here so I uploaded it to Facebook and put link to video here. Listen to it. It does not sound like a siren should sound it needs some work. It needs to sound better than this.

https://www.facebook.com/video.php?v=620167881815825

 

[audioguru]

The circuit makes a European two frequencies bee, boo, bee, boo sound instead of the smoothly sweeping frequencies slowly up and down from a North American siren. There is no link to your video. Another siren video on YouTube sounds great but its schematic is completely wrong.

Today your circuit has a 555 directly driving an 8 ohm speaker. With a 12V supply the output will try to go to +9.5V but its poor little output transistor has a max current rating of 0.2A but then 9.5V/8 ohms makes a current of 1.2A and blows it up.

 

[rjenkinsgb]

Why not connect the output of the first 555 to pin 5 of the second, via a suitable resistor?

That pin overrides the comparator threshold voltages [on pins 2 and 6], allowing for frequency modulation.

Or, buffer the ramp from the first one and use that for smooth up-down modulation.

 

[gary350]

The circuit makes a European two frequencies bee, boo, bee, boo sound instead of the smoothly sweeping frequencies slowly up and down from a North American siren. There is no link to your video. Another siren video on YouTube sounds great but its schematic is completely wrong.

Today your circuit has a 555 directly driving an 8 ohm speaker. With a 12V supply the output will try to go to +9.5V but its poor little output transistor has a max current rating of 0.2A but then 9.5V/8 ohms makes a current of 1.2A and blows it up.

I added an LM386 that will take the load off the last 555. It is tiny bit louder not much. Siren does not sound right I might play around with adjusting the Hz today before connecting the 2N3055. Before my new parts arrived I tried a 10K resistor in place of the 47K the circuit did not work. I don't understand why pin 3 of the first 555 is connected to pin 7 of the other 555. Once I connected 2 frequency generators to the same speaker the 2 frequencies combine and make a siren sound seems to me the 2 555s should connected directly to the same speaker? I wonder if connecting a large value capacitor from pin 3 of the first 555 to ground will round out the square wave a little bit and stop the sound from making a sudden jump from high to low?

 

[AnalogKid]

I wonder if connecting a large value capacitor from pin 3 of the first 555 to ground will round out the square wave a little bit and stop the sound from making a sudden jump from high to low?

No.

The 555 output tries to be a zero-ohm output impedance source. Adding a capacitor to GND will make it work harder and get hotter, but it still will try to make a perfect square wave. A large enough cap will round off the corners of the waveform, but also shorten the life of the chip.

ak

 

[AnalogKid]

I wonder if connecting a large value capacitor from pin 3 of the first 555 to ground will round out the square wave a little bit and stop the sound from making a sudden jump from high to low?

No.

The 555 output tries to be a zero-ohm output impedance source. Adding a capacitor to GND will make it work harder and get hotter, but it still will try to make a perfect square wave. A large enough cap will round off the corners of the waveform, but also shorten the life of the chip.

I don't understand why pin 3 of the first 555 is connected to pin 7 of the other 555.

R4 and R5 form a voltage divider. The intent of the circuit is that R3 is effectively in parallel with R4 when the U1 output is high, and in parallel with R5 when the U1 output is low. This will create two different sounding tones that share some characteristics, but not what you want which is (I assume) two completely different frequencies.

For U1, 1.0 M and 0.22 uF do not combine for a 1.2 second period. Check your math.

I've got company coming today so no time to sketch a circuit. But if I did ... My guess is that what you really want is a two-tone (European-style) siren; both tones are 50% duty cycle square waves; each tone is on for the same amount of time, approx. 1 second. Is this correct?

If so, AND IF the two tones are fairly close together in pitch, you can do this through U2 pin 5. But the sirens I've heard have two very different tones. For that, you need to switch the U2 timing capacitor, not one of its resistors. Do you have a 2N7000 small MOSFET?

ak

 

[audioguru]

Since R4 in series with R5 charge the timing capacitor then when pin 3 of 555 #1 goes high then R3 is parallel with R4 and charge the capacitor a little faster producing a slightly higher frequency from the 555 #2.
When pin 3 on the 555 #1 goes low and R3 is 10k then the capacitor is not able to charge high enough so the oscillation stops. Poor design.

Now you are feeding the 9.5V output high of the 555 #2 into the input of an LM386 that has a maximum allowed input of only ±0.4V so you probably have destroyed the LM386.

You do not filter the rectangular waveform at the output of a 555 by shorting its output with a capacitor to ground. That would destroy the output transistors of the 555. A filter is a series resistor feeding the filtering capacitor.

The datasheet of a 555 shows what other people here are talking about. It shows "pulse position modulation" that uses a signal feeding pin 5 to vary the frequency. Unfortunately the signal also varies the duty-cycle that varies the output loudness that a real siren does not do.

You need a Voltage Controlled Oscillator to replace 555 #2 and you need a triangle waveform to replace 555 #1.
An LM358 can make one and an LM324 can make both of them.

 

[gary350]

I did a few experiments, learned a few things and changed the circuit drawing. I was going to try what you said connect pin 3 of 555 #1 to pin 5 of 555 #2 when I removed the capacitor on pin 5 of 555 #2 wow the circuit started working much different and better. Then I removed capacitor from pin 5 of 555 #1 wow again this makes a big difference. TTL book does not show capacitor on pin 5 I put one their anyway. Pin 5 caps are both gone. Next I connected pin 3 from 555 #1 to pin 5 of 555 #2 and it did like you said but volume is 70% lower and siren sound has changed it sounds worse. I even tried 10K as show on an online circuit drawing still low volume and sounds bad. I changed it back the way it was 47K resistor pin 3 to 7 R3. I had changed C1 a few times and not changed the drawing now it is correct .2uf works much better than .1uf. I noticed 555 # 2 getting hot after searching for a low value resistor I found 22 ohms and 39 ohms both 5w. 22 ohms is too low 555 still gets warm. 39 ohms works good. I have some 1/4w 10 ohm resistors I can solder several in series to get 30 or 40 ohms which ever works best. Next I want to increase C3 from 470uf to 1000uf to see if I notice a difference in the sound. Things are looking good siren sounds more like a European siren but still not 100% perfect. I am powering this with 8 AA batteries 1.65v each = 13.2v no load. Meter shows batteries are 2 amps. Attached is the circuit with all the changes.

Question about 2N3055 v to EB. If the transistor has 5v 1ma signal on the base then it is changed to 5v 2ma does the output of the transistor change? Seems like I recall only the base voltage determine the output watts at C not the base current?

It appears to me rotating both 555s 180 degrees is better for soldering. Pin 3 is on the correct side for the speaker. Pin 4 is correct for +12v at the top. Pin 7 is better located to solder to pin 3. Pin 1 is in a better place to solder to ground. No advantage or disadvantage for pin 2 & 6. Everything seems to be in a better location except pin 8.

 

[audioguru]

Instead of severely overloading the output pin 3 of 555 #2 with DC into the resistor R6 and speaker, why not try capacitor-coupling AC to the speaker with a 470uF capacitor in series with R6? Then the value of R6 can be reduced to 10 ohms for no overloading and a louder output.

Increasing the value of C3 worked well because your puny AA batteries (with poor contacts in the holder) are not powerful enough to keep the voltage anywhere near 13.2V when they are overloaded.

Guess what? When you increase the base-emitter voltage of a transistor then the base current increases. And if you increase the base current of a transistor then the base-emitter voltage increases. This is shown on a graph in the datasheet of every transistor.

 

[gary350]

I will test all these changes in a few minutes. I was planning to replace R6 with a 200 uf cap anyway.

Here is a circuit I found online Ambulance Siren if this actually works I want to build it to see how it sounds. Looking at the circuit can anyone tell it is works? Why is R1 & R4 different values? Why is R7 & R10 different values? I know that will make 1/2 of each osc is different but why, larger pulse? I tried to look up C458 & C547 all I get is advertisements, Buy your parts here!!! "Click here to down load. " I'm not downloading anything computer has been screwed up too many times doing that. I would like to know if 2N3904 or 2N2222 can be substituted for C458 & C547 I have a bunch of them? I need to run this on 12V also. C5 & C6 are a problem I guess 4 caps .01uf in parallel will work.

 

[Nigel Goodwin]

R1 and R4 are different either to give an asymmetrical output, or to compensate for the double transistors..

For 0.04uF, use the preferred value of 0.039uF.

Any small NPN's should be fine.

However, I fail to see how it can work with R6 at 10K and R5 at 1.2M?

 

[audioguru]

The website eleccircuit.com has comments in it that the writer is a Noob (does not know electronics or English).
A C458 might be a BC458 or a 2SC458. Who knows? I agree that any little transistor will work if protection diodes are added to fix its next problem:

With a supply exceeding 6V then the transistors will have avalanche breakdown of their emitter-base junctions which will slowly destroy them.

 

[gary350]

I had already been thinking about building 2 osc. but not sure how to combine output into 1 speaker. This is only osc I know to build, their could be a better way.

 

[audioguru]

This circuit might work:

 

[gary350]

This circuit might work:

That circuit gave me an idea. 2 555 osc running at different Hz connected to a relay that runs by a 555 relay timer. The 2N3055 gets alternating Hz for the speaker. I find lots of circuit board relays on PC boards. Solid state relay might work better than mechanical but only ones I know about are high voltage solid state I worked with in industry. Tomorrow I look at some of the specs on relays I was playing with 1 a few weeks ago rated 12v.

 

[audioguru]

A relay wears out. Also its coil produces a high flyback voltage which would destroy the 555 driving it.
Simply mix the outputs of the sound producing 555s with 2 resistors feeding the 2N3055 then use the switching 555 to reset (pin 4) one of them and a transistor inverter to alternately reset the other one.

Your 2N3055 is a DC amplifier and will not work properly with an input coupling capacitor. Also you are feeding its base with such a high voltage that its base-emitter reverse maximum allowed voltage of -7V is exceeded which damages it.

How can you filter the squarewave buzzing that sounds like an acid rock overdriven electric guitar?

 

[alec_t]

If you want an up-down swept frequency instead of the nee-naw siren sound you could try this circuit :-

U1a and U1b create a large amplitude slow up-down ramp which modulates the frequency of a sawtooth tone generated by U1c. (Without D1 the duty cycle, rather than the frequency, would be modulated).
Inputs of the three unused gates of the CD40106 package should be grounded.

I've attached a siren.wav file of the sound it generates. (Delete the .txt extension). Also attached is the LTspice simulation file.