Relaxation oscillator with neon or...

[Externet]

How do I make a once or twice a minute beeper/clicker/tone/chime using a neon bulb ?
Voltage supply is 120VAC. No transformers desired. Piezo buzzer or tiny speaker is OK.

120VACphase-------------|>|-------------/\/\/\/\/-------------neon--------------neutral

(With the DC smoothing capacitor after the 1N4007 rectifier and the timing capacitor across the bulb?)
-Is to warn/remind an elderly that an appliance is on or been left on- Yes, a disposable camera flash once a minute could be good too.

Or, alternative, a capacitive coupler supply from 120VAC to drive the beeper + flasher once a minute ?

 

[danadak]

This might help, attached.



Regards, Dana.

 

[crutschow]

Wouldn't you want the lamp/buzzer to be active for some period of time after the 1 minute interval?

 

[Externet]

Wouldn't you want the lamp/buzzer to be active for some period of time after the 1 minute interval?

Beeping/chiming/flashing once every minute while appliance is left on should suffice. Would be heard at some moment and get the appliance turned off.

Question... is it a new thing/trend/mode that the links as Dana posted cannot be open and seen but instead go to saving when clicked ?

 

[danadak]

All depended on how web page designer set up the links.


Regards, Dana.

 

[crutschow]

I assume you are looking for minimum parts count/cost(?).

 

[Pommie]

Question... is it a new thing/trend/mode that the links as Dana posted cannot be open and seen but instead go to saving when clicked ?

The links are pdf files so it's your computer doing it. For me they just open as pdf files.

Mike.

 

[Externet]

I assume you are looking for minimum parts count/cost(?).

Mostly size, to fit in small spaces.

The links are pdf files so it's your computer doing it. For me they just open as pdf files.

Interesting. Will try to find a setting in my compfuser; other .pdf open fine.

 

[AnalogKid]

I don't think a simple neon bulb oscillator will work. If the voltage swing across the bulb were equal to 1 time constant (it isn't, but this is a first-order guess), t = 60 seconds. That's a 10 M resistor and a 60 uF cap, with a peak charging current of around 13 uA. High voltage electrolytics have more leakage current, so you might get away with a smaller cap, but it feels like this approach is dancing on the edge or reliable performance.

The good news is that it is easy to build and test. I'd start with 47 uF / 200 V.

That still is a large cap. Another issue might be the effect high current spikes have on the lifetime of the neon bulb. Hmmm . . .

ak

 

[crutschow]

That still is a large cap. Another issue might be the effect high current spikes have on the lifetime of the neon bulb.

A resistor in series with the bulb will control the peak discharge current.

 

[AnalogKid]

I was thinking that might mess with the firing voltage (vague memory of something about this in one of the neon books floating around), and I don't see a neon bulb in LTS. But it also might work as a current limiter for a speaker or piezo beeper or whatever in series with the bulb. Maybe something with a zener across it to limit the peak voltage it sees. Still chewing on that idea.

ak

 

[ChrisP58]

There is a neon bulb in LT Spice. It's in the Misc folder. Right-click the symbol to adjust it's parameters.

Attached is a possible start for your project.

 

[AnalogKid]

"There is a neon bulb in LT Spice."

I would like to say 'I don't know how I missed that.", but actually I do. Ah well, life is choice.

Here is a re-work of #12. I decreased the breakdown voltage to 90 V based on several NE-2 datasheets. They don't list the extinguish voltage so I left it at 50 V. R2 represents a small 8 ohm speaker to make a "tick" sound.

The LTS model has has a parameter called "Zon", which I assume is the effective resistance during conduction. It was set to 2 K, and I decreased it to 1 K just to see how it affected the output.

Disconnecting C1 lowered the oscillator freq, making for a smaller C2.

ak

 

[ChrisP58]

"There is a neon bulb in LT Spice."

I would like to say 'I don't know how I missed that.", but actually I do. Ah well, life is choice.

Here is a re-work of #12. I decreased the breakdown voltage to 90 V based on several NE-2 datasheets. They don't list the extinguish voltage so I left it at 50 V. R2 represents a small 8 ohm speaker to make a "tick" sound.

The LTS model has has a parameter called "Zon", which I assume is the effective resistance during conduction. It was set to 2 K, and I decreased it to 1 K just to see how it affected the output.

Disconnecting C1 lowered the oscillator freq, making for a smaller C2.

ak

View attachment 137170

Thanks for the refinement AnalogKid. I had an idea that I thought might work, but it was late and I just threw the concept together without spending much time on refinement.

Externet. Be careful when you build this. The entire circuit, including the speaker, is at mains potential and represents a shock hazard, so make sure you have an adequate enclosure to protect it.

​

 

[AnalogKid]

it was late and

That's how I missed the neon bulb in the library.

ak

 

[AnalogKid]

Here is a real schematic.

Built as shown, with an NE2H bulb and 8 ohm speaker, periods after the first one are approx. 48 seconds. Pretty close to the sim, considering that the H bulb has different voltages.

BUT - the tick in the speaker wasn't very loud. A 4" general purpose spkr was louder than a 2" laptop spkr, but neither would get my attention if I weren't waiting for it.

Next I tried an assortment of piezo integrated beepers, the kind with a 1-transistor oscillator built-in. Most were better than the speaker, primarily because a short beep is much more noticeable than a tick. One is pretty good, but it is about the size of a golf ball.

The problem is total energy. If a 10 uF cap discharges from 90 V to 50 V, that is a total energy of 0.028 Ws. If the discharge lasts, say, 100 ms, that is a brief 1/4 W pulse, some (much?) of which is dissipated in the bulb. With an 8 ohm speaker the peak power at the start of the tick was around 7 mW; not very much. Not surprising since LTS thinks the internal impedance of a conducting neon bulb is around 2 K. This suggests that a high impedance sounder like some kind of buzzer or beeper will get a larger percentage of the available power.

Notes:

For a more accurate energy calculation, I tried to measure the voltage ramp across C1. The 10 M scope probe forms a voltage divider with R1, reducing the peak voltage to a value below the bulb's breakdown voltage, preventing oscillation. Grrr . . .

There is a large variation in circuit timing with different neon bulbs.

ak

 

[AnalogKid]

I also tried a piezo disk. With the cap charged, yuo get one blink and a very nice tick when you connect the disk. After that, the disk impedance is so high that there is not enough voltage across the lamp to reach conduction.

ak

 

[ChrisP58]

I also tried a piezo disk. With the cap charged, yuo get one blink and a very nice tick when you connect the disk. After that, the disk impedance is so high that there is not enough voltage across the lamp to reach conduction.

ak

Would a resistor across the piezo disk fix that?

 

[KubaSO]

You could replace the "speaker/piezo" with a 12V or 24V reed relay coil. The input resistance will typically be 1kohm/2kohm, respectively. It should turn on for quite a bit longer than an 8 ohm resistor would. The reed contact can drive a piezo buzzer using a separate capacitive voltage dropper. If you can't get a reed relay affordably, get in touch via a private conversation. I can get them to you just for asking. Make sure you reference the particular project you need it for.