I have a 1953 Seeburg HF100G, that is running beautifully. Wanting to 'play' with it more I’m looking for some help creating what I would call a ‘Digital Jukebox Wallbox.’ I’m fairly adept with components and circuits, but programming is not my forte. Here is a link to the manual for which I would like to recreate.

http://www.kjq.us.com/images/3w1.pdf

And for those that had their interest roused this involves an old valve (2050 tube).

Any collaboration or direction would be greatly appreciated.

 

no bites...

 

I had a look, but got confused by all the mechanical parts. If you could identify the sequence and timing of pulses that need to be sent, that might help.
It would seem to be an ideal job for a microcontroller, but it's probably possible with a handful of logic ICs.

 

That is a work of art .... what do you want to make it do that it inst doing now?

However, if you must play with it ...

Before programming comes planning and design, by far the most difficult and time consuming part.

Identify all the potential inputs, including internal mechanism feedback.
Identify all the required output functionality, motors, solenoids lights etc.

Write a high level description of the required operation and then break that into lots of smaller routines and subsystems.

You want to end up with a backbone that orchestrates lots of small simple routines.
(All written in English as bullet points that refer to specific inputs outputs or events. (Pseudo code))

Once that is done you can then decide on a platform to implement it, be that timers relays and cam sequencers or a microprocessor. It matters not as all the detail is now broken into its most basic logical form.

Programming generally isn't hard you need to important things, a system you can manipulate, which you will have to learn how to control and more importantly a very clear and detailed description of all aspects of the task to be programmed.

As a general rule the more capable the programming environment the harder it is to learn, so if all you need is ON,OFF and a few slow timers then pick a simple one with a high level interface (IDE) that hides all that complexity you don't need to know about, Essentially unless you really want or need to know what the uP is doing at a low level don't get into it.

If you want specifics I will be happy to help if I can.

 

Why in earth do you want to take a working piece of art. And turn into something that is worth nothing. That is to dumb if you ask me.

 

Why in earth do you want to take a working piece of art. And turn into something that is worth nothing. That is to dumb if you ask me.

I don't think he has the wall unit, just wants to emulate it.

 

Rather than trying to emulate the wall unit, I would try to determine what you need to do to select a song in the HF100G. That may be no more than just closing a relay contact across the push-button selection push-buttons. That can be readily done from a remote keyboard and a circuit or microprocessor to convert the key to BCD code (such as this) which is sent to the jukebox. The BCD code is then decoded to close the contact across the desired pushbutton.

 

I had a little look around but couldn't find the sequences, but you might be able to find something here:
http://dir.groups.yahoo.com/group/seeburgjukeboxinformationclub/
You have to sign up to see images, but it seems quite active.

 

A simple way to convert the signal from 10 pushbuttons into a BCD signal is a CD40147 10-line to 4-line priority encoder. The signal can be converted back to 10 individual signals by a CD4028 BCD to Decimal decoder, which can then control relays across the jukebox pushbuttons.

You could use one encoder-decoder pair for the numbers and one for the letters. A CAT-5 shielded cable has eight conductors which could carry the two sets of four BCD signals with the shield being the common (ground).

 

I had a little look around but couldn't find the sequences, but you might be able to find something here:
http://dir.groups.yahoo.com/group/seeburgjukeboxinformationclub/
You have to sign up to see images, but it seems quite active.

Yep, I'm already a member there, and have asked the same question and have received little fanfair over there.

 

A simple way to convert the signal from 10 pushbuttons into a BCD signal is a CD40147 10-line to 4-line priority encoder. The signal can be converted back to 10 individual signals by a CD4028 BCD to Decimal decoder, which can then control relays across the jukebox pushbuttons.

You could use one encoder-decoder pair for the numbers and one for the letters. A CAT-5 shielded cable has eight conductors which could carry the two sets of four BCD signals with the shield being the common (ground).

I would like for this to work through the Wired Selection Receiver on the jukbox, I don't want to 'molest' the inner-workings of the jukebox that are already in place.

http://www.verntisdale.com/schem/wsr5-l6.jpg

 

I don't think he has the wall unit, just wants to emulate it.

That is CORRECT!

 

I had a look, but got confused by all the mechanical parts. If you could identify the sequence and timing of pulses that need to be sent, that might help.
It would seem to be an ideal job for a microcontroller, but it's probably possible with a handful of logic ICs.

Therein lies one problem with this task, the exact pulse timing is what I don't know.

 

Rather than trying to emulate the wall unit, I would try to determine what you need to do to select a song in the HF100G. That may be no more than just closing a relay contact across the push-button selection push-buttons. That can be readily done from a remote keyboard and a circuit or microprocessor to convert the key to BCD code (such as this) which is sent to the jukebox. The BCD code is then decoded to close the contact across the desired pushbutton.

This would require changing/adding circuits and relays to the jukebox, which I won't/can't do.

 

Thank you all so much for at least taking a look at this for me. As 'crutschow' said I want to emulate the 3W1 Wallbox which will entail figuring out the pulse timing for which the WSR (Wired Selection Receiver) is listening.

 

Thank you all so much for at least taking a look at this for me. As 'crutschow' said I want to emulate the 3W1 Wallbox which will entail figuring out the pulse timing for which the WSR (Wired Selection Receiver) is listening.

Well, then good luck with that. The manual says it outputs a sequence of 2-21 pulses with 1/25 second interval which selects a "selector coil circuit" in the jukebox and then, after a pause of 1/5 second, then another 1-5 pulses at 1/25 second interval to select the "group solenoid". I think you will have to learn how to program a microprocessor to convert the 20 push button selections to generate that particular sequence of pulses. Doing it with discrete circuits could get complicated.

 

I think I may have figured it out, it's going to be hard to explain though. I now think a microprocessor is necessary, there doesn't seem to be any other way.
Just looking at the first diagram:
I'm going to call the contacts on the outside edge of the disk X1-X21 counting clockwise from where the spinning contact starts (the big arrow from the centre point). There is a bit of a gap and then Y1-Y5. There is also the shorting strip which goes from X11-X21.

Nothing happens until number and letter buttons are pressed

If button 10 is pressed there will be 10 pulses, X1-X10
...
If button 5 is pressed there will be 5 pulses, X6-X10
...
If button 1 is pressed there will be 1 pulse, X10

The letter buttons are trickier.
If A,C,E,G or J are pressed the shorting bar is grounded so X11-X21 is one long pulse
If B,D,F,H or K are pressed, that doesn't happen and there are 11 short pulses instead (X11-X21).
The Y1 pulse always happens.
If A or B is pressed there are no pulses after Y1
If C or D are pressed there is a Y2 pulse
If E or F are pressed there are Y2 and Y3 pulses
If G or H are pressed there are Y2, Y3 and Y4 pulses
If J or K are pressed there are Y2, Y3, Y4 and Y5 pulses

(There is no I).

I could be very wrong.

 

I think I may have figured it out, it's going to be hard to explain though. I now think a microprocessor is necessary, there doesn't seem to be any other way......

I could be very wrong.

This sounds plausible, and then the moment of humility... LOL

Do you think you could write that programming?

 

Well, then good luck with that. The manual says it outputs a sequence of 2-21 pulses with 1/25 second interval which selects a "selector coil circuit" in the jukebox and then, after a pause of 1/5 second, then another 1-5 pulses at 1/25 second interval to select the "group solenoid". I think you will have to learn how to program a microprocessor to convert the 20 push button selections to generate that particular sequence of pulses. Doing it with discrete circuits could get complicated.

How the jukebox works is relatively easy, the selector coil, and group coil circuits are understood, there are 20 selector coils within each of the 5 groups coils on the 'pinbank' or selector assembly. The electrical selector or 'keyboard' works in tandem with the selector assembly. Now the hard part to understand, you can play or select songs without the keyboard and pinbank connected to the WSR, the WSR as a whole is not needed to make selections for record play, it is only needed to 'decode' the pulses for the pinbank or selector assembly.

And Mark you are correct, there is no I, they skipped it so as not to be confused with the 1 key.

 

I don't know if I've posted this, but here is the scheme for the WSR.

http://verntisdale.com/schem/wsr5-l6.jpg
http://verntisdale.com/schem/wsr5-l6a.jpg