Build your own Pulse to tone converter?

[joey67]

Does anyone have plans for building your own pulse to tone converter? I know they can be purchased online, but I would like to build my own for the fun of the project.

[twocvbloke]

There's probably plans out there, but unless you have the hardware for programming PIC chips, then it could become a costly project to build (acquiring the chips, the programmers, the software, the chip programming, etc.) compared to the price of a pre-made Rotatone, and may not be as reliable either...

[George Knighton]

Some of the in-line converters I've seen are so cheap, I wouldn't even think about building my own for that price.  :-)

[Bill]

Does anyone have plans ...

Not to be condescending, but it kinda depends on what you mean by "plans". If you are looking for a Heathkit sort of approach, where you simply assemble some parts according to a detailed instruction sheet, without really knowing what they do or why, then the answer is probably "no". If you know enough about electronics to design some circuits of your own, then the answers range from "maybe" to "you bet". And it would be a fun project.

In order to avoid the microcontroller-and-software problems mentioned by twocvbloke, I would go back to the way we would have done it in the 1970's and 1980's, when special-purpose ICs started to become available, but before computers and microcontrollers began to dominate electronics.

I would expect that the circuit would have three sections. First is a pulse counter, with either a 1-of-10 or BCD output. Both are available as ICs. The second section is a DTMF encoder,, driven by the pulse counter. Again available as an IC. Third, and perhaps trickiest, is a bunch of timing circuitry, to ensure that all the parts perform when needed, and for as long as needed but no more. You have to reset the counter between dial rotations, for instance. You have to ensure that the encoder does not generate outputs while the counter is counting, but that when the time comes, it generates the tones for as long as needed but no longer. And so forth.

A Google search will get you started in the right direction on all of the ICs you would need, as well as some circuit approaches. But  it will not design the thing for you. This would be a fun project, but as George says, not particularly cheap. Google will uncover such stuff as this, a BCD-to-DTMF encoder board
http://www.monroe-electronics.com/CATV_prodpdf/3171b_ds.pdf
Note that the none of the timing mentioned above is included in this board.

Let us know if you decide to do it.

Bill

[Mr. Bones]

Fascinating thread. I only need the tone for limited things, since my POTS works fine with pulse, thankfully. Saves me from explaining a house full of non-useable paperweights, at least

Posted by: George Knighton    Posted on: Yesterday at 09:39:57

Some of the in-line converters I've seen are so cheap, I wouldn't even think about building my own for that price.  :-)

I utilize a freebie Uniden cordless 900 mHz as my touchtone keypad; even has memory, speed-dial, re-dial, caller ID, and many other features. When needed, (rarely)I simply press the button(s), while enjoying the call on one of my rotaries. It's not a  replacement for the pleasant dialing of my rotaries, just a tool to work with when having to enter # or *, mostly.

It fits well within my budget, leaving me more to spend on desirable telephone stuff.

[joey67]

Bill,
   We appear to be on the same page in so far as the approach we would take for building the circuit; avoiding microprocessors and programming. I am not a circuit designer by any stretch of the imagination though. When reading your circuit description it would make sense to me if you included a digital latch (memory) between the pulse counter and the DTMF encoder. This would give the encoded BCD a place to stay while the timing section of the circuit detects what to do. I would very much like to collaborate with you on this project; and develop a circuit that other members could build into a project box. I have an experimenter's board and experience at this kind of thing, so we could do all of our designing and testing there.

[paul]

This page: http://www.radar58.com/RotoTone/RotoReport/RotoReport.html , seems to have a converter design in it, among its discussion of a very interesting construction of a telephone.

[joey67]

Thanks Paul, very big help with the link you sent !

[cadman]

Talk about your small world! I'm the guy that built that Rototone in the link above. Back in 2002 I built this phone for a college course (actually, I built it for me...the report was for the course..and it reads like it). What makes it complicated is the digital display (Sperry Panaplex) and the shift registers. You could multiplex that portion and reduce the IC count, or eliminate it entirely and really clean things up. This being 2013, I think I'd take a stab at doing it with a uC. -Cory

[Phonesrfun]

Hey, we have a real celebrety amoung us now!  Welcome to the forum, Cory.

I think that for older phones, the display would not be needed.  The handy feature that some of the newer commercially available pulse to tone converters have is the ability to somehow get * and # out of a rotary dial sequence, and to get some limited speed dialing also from some rotary dial sequence.  I still have the ability to dial with rotary phones so I don't have any of the rototones that are sold.

The following might be a topic for further discussion, but I will bring it up here.

Many phone collectors that are also into switching have built functioning Step-by-step (SXS) switches.  I have a small set that I have put together.  Those, by definition of how they are made, only work with rotary dial phones.

Back in the day when dial central offices were converting to touch tone service, it was necessary to place a tone to pulse converter in between the line finder and the first selector of a SXS office so that tones could be decoded and outpulsed from there.

I am under the impression that small versions of these are both scarce and costly.  How difficult would it be to create one of these that could decode, store, and send a series of numbers from a TT phone pad?  The output would need to be capable of pulsing 48 volts at what ever the current requirement of the switch would be.  I don't know what that is, but 50 mA sounds about right.  Probably needs to be pulsed through a relay.

[cadman]

Hey Bill!

Yes, an interesting problem. There's a couple ways to tackle the TT->Pulse problem. Just thinking out loud now, one could start with a DTMF decoding chip, like the MT8870. This provides a BCD output whenever a tone is recognized. One could take a cheap little microcontroller and program it to read that BCD value and pulse an output line x-many times to cycle a relay to simulate dialing. Of course you'd have some software and programming involved.

An old school method that might work would be to use that DTMF chip to preset a value in a counter. The 74192 IC is an old up/down decade counter chip with individual output lines. A little oscillator running at 10PPS could be running in the background and if the value of that counter is anything other than zero, enable the oscillator to simultaneously pulse a relay and decrement the counter. When the counter hits zero the relay pulsing is disabled. The "carry" line on the counter could correspond to 0/Operator (10 pulses).

I haven't played with this stuff in years but it shouldn't be too tough to get going. Now on the SxS stuff, I don't know a thing but would love to learn!

[cadman]

Thinking about this a little more, the WE units must have had some sort of memory to compensate for the speed of pushbutton "dialing" while the mechanical switch was busy. How fast can a crossbar/SxS run? 20PPS?

'Back in the day' could you pushbutton dial faster than the exchange could handle? I'm sure nobody in this group would have tried that 

Using the micro, or a Basic Stamp (do they still make those?), you could buffer DTMF as quickly as the phone is 'dialed' and simultaneously step the switch.

[Phonesrfun]

They had to have a memory of some sort, because one can definitely bang out numbers on a keypad faster than some mechanical switches can handle.

sxs switches are the slowest, and they are usually rated at no more than 10 pps, although some can get as much as 20 out of them.  There is just the inertia of the switch that has to be dealt with.  So, on a sXs the standard is 10.

Crossbar and panel switches can go much faster because the pulses are stored in a register of relays, but how much faster I don't know.

In the day, a subscriber who was connected to a SXS exchange got no real pick-up in speed from having a touch tone phone.  Touch tone dialing pretty much killed off any remaining step offices.

[Mr. Bones]

'Back in the day' could you pushbutton dial faster than the exchange could handle? I'm sure nobody in this group would have tried that  

    Of course not, that would just be wrong! Especially when trying to win free Steppenwolf tickets, etc., off of KY102.(Won!) Or, so I've heard... it was well before my time, you understand!

Best regards!

[dsk]

I just located this page:
http://www.youtube.com/redirect?q=http%3A%2F%2Fboris0.blogspot.com%2F2013%2F09%2Frotary-dial-for-digital-age.html&session_token=y1NyEsT_GAVg4SGBgWUPURD-6lR8MTM4MDE3OTQ4MkAxMzgwMDkzMDgy

http://tinyurl.com/mj79j75

I had some problems opening that page but I managed to make pictures of the page:
dsk