Circuit for replicating bell codes - any suggestions

[rogs]

Edit: The counter and latch could be replaced by DIP switches or binary-coded rotary switches, for setting predetermined codes rather than random ones.

Already been there ! (see my post #25)....... easy to miss - this is now a pretty long thread!

 

[alec_t]

Here's my offering:

U2a provides a ~0.5Hz clock signal for shift register U1. Pressing the 'Play' button disables the clock for a time set by R2/C1 and loads an 8-bit code into U1. On button release the code is clocked serially out of U1, and ANDed with the clock pulses to drive level-shifter transistors Q1/Q2. Q2 provides gate-drive pulses of ~0.25s duration to whichever N-FET is selected by Sw1. The N-FET switches the appropriate bell. The 'Attention' button turns on Q2 directly.
This example is just a proof-of-concept 8-bit design, hard-wired to send bell code 10111011. The code is determined by the connections of the D inputs of U1 (a 74HC165 parallel-in-serial-out shift register), and could be varied using 2 hex-coded selector switches such as https://uk.farnell.com/te-connectivity-alcoswitch/drd16e04/switch-rotary-20v/dp/1330343
Adding another 74HC165 and 2 more selector switches would enable 16-bit codes to be sent.
The sim waveform shows the pulses at bell-selector switch Sw1. The first 2 pulses are the 'Attention' signal and the reply. Then follow the 8-bit sequence and its reply.

BTW, where is this signal box you're restoring?

 

[dr pepper]

Is this the kind of thing your looking for, you'll have to imagine the red led is a bell, i havent got any electromagnetic bells knocking about, the keypad is a standard 12 key pad, 16 keys would be easily implemented, button 2 is a urgent police required signal, cant remember the others:

 

[KeepItSimpleStupid]

Dr. Pepper:

Did you just do that?

The OP mentioned that the bell is a plunger like thing that goes Ding! which would be really close to an old electromagnetic doorbell. The OP mentioned that someone said that most bells worked on 8-12 VDC and he used a car battery to test. So, we don't know the current requirements of the bell. He also mentioned that the distance between the two bells were on the order of 10-15 feet or so. There were 16 codes the OP wanted to create.

So, I got the impression that there is a tour guide doing a demo and just has too much to do to worry about "keying" (from morse code) the required code that they wish to demonstrate. So the "tour guide" pushes a button and someone in the basement responds ringing the second bell with the required response. The "second person" in the basement interprets the code and responds.

I also got the feeling that the OP wanted to select either bell, so then the "tour guide" could do both operations. Later in the thread, there was a hint to put an "electronic interpreter" to respond to the bells.

The real specs of the "ultimate" or the simple system really have not been formally specified.

The OP apparently has rather extensive programming experience and limited, but some "programming in solder". Circuit Cellar's founder refers to "solder as his favorite programming language".

All I really did, was to suggest the PICAXE because of the simple requirements that really doesn't require a development system. The only development system is the USB to miniature phone plug cable. My suggestion is a bit unconventional, but should be very easy to construct and program. The additional development tool required is a crimping tool for the connectors. Probably one of these: https://www.google.com/url?sa=t&rct...=7bF5dgp7b3-XOmfYnroygw&bvm=bv.59378465,d.cWc The price is out to lunch, but good deals can usually be found on ebay

Build the 118 kit with a few add/subtracts and 3 resistors to create the I2C bus and the interrupt bus. A few connectors get added to the octopus cable basically create an I2C/Interrupt cable, a power cable and a "bell" cable". Then build a I2C daughtercard for the KB encoder with an I2C/Interrupt connector and power cable. The keypad suggested has both a row/columd and single closure interface. The I2C bus chip that I selected is only available in surface mount, but you can buy the chip with the I2C chip mounted to a DIP carrier.

I think it would make the programming extremely simple. It could be as simple as wait for interrupt, read the keys, (prioritize mult-presses) and do the appropriate output. Using some sort of data structure for the bell encoding could be the only complication if one waned to add it.

 

[dr pepper]

Yes, I allready had a keyboard handler routine.
And I did a sequencer a while back which easily adapted.
Theres only 3 buttons programmed as of now just as a test, seems to work.
The software runs on a pic16f628a, however its intended for a lowly 16f54.
Personally I'd fire the bell with a relay, then you can just parallel it up to the existing system, the relay contacts will probably outlive the bell's mechanics.

 

[KeepItSimpleStupid]

I don't think there is an existing system because of the mentioning of a car battery to test. We don't have an estimate of the current.

 

[dr pepper]

I looked up 'block bell' on youtube and found a few examples, they look like a 6" bell, and we know they work off 12v, and can have 2 or 3 miles of cable run, so it's be a guess they'd pull an amp or 2, an 8a relay ought to do the trick.
Heres a model layout with what looks like recovered block instruments, looks like 2 bells are paralleled.

Selecting a bell isnt a problem and doesnt need software, just a c/o switch.
Replying to a signal as far as I know is just repeating it back, that shouldnt be so difficult, jst play the same sequence twice with a delay inbewteen.
The only thing I'm not so sure about is the first ding, this is to attract attention, if you were doing more than one bell code I suspect this attention ding isnt req on any other than the first code, and that might need special attention in the software.
We are now waiting for the o/p to tell us more.