Universal Telephone Interface with C8PDF

[pbirkel]

Yes, TLS = Telephone Line Simulator. Low-cost Teltone TLS-3 and TLS-5 are currently listed on eBay. Or you can get more aggressive and set up an analog-supporting PBX such as a Panasonic KX-TA308 hybrid (3 outside line; 8 extension lines). Or you can "hot-wire" a very simplified TLS substitute for yourself; e.g.: https://www.jagshouse.com/modem.html

VOIP, like the FIOS service, is iffy for using a modem or fax, although at only 300 baud you might not have any problems with a modem ... once you find someone at that speed at the other end :-}.

WRT Verizon "should" =/= "does". It's unlikely that they are actually verifying that at your local CO-equivalent; instead someone is just reciting from an old marketing brochure ...

[lowrybt1]

Before purchasing a TLS, I'm going to try using a CELL2JACK device to bridge from the CBT to a cellphone. It looks as if the CELL2JACK has configurable voltages and loading for ring strength, hook detection and off-hook voltage. If it doesn't work, I'll use the cell2jack for regular home phone service and get a TLS to explore the 567 capabilities.

[pbirkel]

Interesting device. Presumably you'll encounter the usual VOIP caveats if you try to establish a modem/fax connection across it.

[lowrybt1]

Success, mostly. I have the OSI 567 Board (without Votrax module) working and interfaced to a VOIP line. The board successfully completes its modem self-test, it is able to go off-hook and on-hook (and trigger a dial tone), and it was able to detect and answer an inbound call and attempt a modem connection to that inbound call. When I try to dial an outbound number, it goes off-hook but does not succeed in completing the call. (It detected a ring on the other end once out of 5 tries.) There are a few things that might explain the issue:

1) There's some issue with the components on the OSI 567 that drive the DTMF signals (in which case, I maybe stuck)

2) There's a problem with the Data Access Arrangement module's circuit that converts signals from the 567 into DTMF (in which case I could try another DAA.)

3) I have crossed some wires in making the connections between the 567 and the DAA (easy enough to check)

4) The interface circuit (Cell2Jack) that links the DAA to the VOIP line isn't "cutting it" because it works with line voltages and loads different from old school POTS connections. The interface does work with an old rotary phone -- hook power, ringing, audio -- so I'm not certain what the issue could be. It might be that the DAA is powering 4 contacts on its RJ11 but the Cell2Jack uses only contacts 2 and 3. A professional telephone line simulator would really help diagnose this -- as pbirkel suggested. TLS units are kind of expensive.

In any case, I'm feeling really good about getting this far. Now what do I do?

[pbirkel]

Regarding #4, when you say "The interface does work with an old rotary phone -- hook power, ringing, audio" you mean that pulse dialing lets you "detect and answer an inbound call and attempt a modem connection to that inbound call" and also an outbound connection of the same sort? So the only problem is inconsistency in initiating outbound connections using DTMF and the Cell2Jack?

Certainly (re)check #3.

Then I'd be patient about acquiring a TLS as ... yikes on current eBay TLS pricing. I acquired my TLS-3 in 2018 for $50 plus shipping. That was my limit then and I wouldn't go much higher today.

I'm unclear exactly what modem service you expect to connect to using Cell2Jack, or even your Verizon FIOS? A TLS should give you a solid local modem connection with full setup/control signal sequences, but if that's all that you want then you can tweak a simple connectivity circuit to a modem in another hobbyist computer of the same era -- lots of choices there! Several circuits/approaches are documented online.

High-brow guidance:

https://www.epanorama.net/circuits/teleinterface.html
https://www.epanorama.net/circuits/tele ... inger.html

Or the low-brow method:

https://www.jagshouse.com/modem.html

[bxdanny]

Congratulations. That UTI board was notoriously complex and difficult to program. The need for that DAA made it quite expensive too. I don't think a whole lot of them were sold, it was much easier and cheaper to use a Hayes-compatible modem.

I didn't think that using a VOIP line would be a problem. Anyway, traditional analog phone lines aren't even available anymore in many places. My mother was told by Verizon a few years ago that she needed to switch over to FiOS, as the regular copper phone lines in her building were going to be deactivated. When the Verizon technician arrived, he said that he needed to remove her living room air conditioner temporarily in order to bring the FiOS line in. She said nothing doing, and switched to Optimum Voice (from the cable TV company) instead, as there was already a cable-TV line in the apartment. She now has two phone lines coming from a cable modem, one for voice and one for fax. The fax machine works fine with it.

What I'm wondering about with regard to that Cell2Jack device is how it handles the one big difference between dialing a cell phone and a landline phone: the need to press "send" or "call" on a cell phone after entering the number. Does it initiate the call a certain length of time after the last digit is dialed? Or do you have to press a key, like "#", to actually make the call?

Anyway, couldn't you connect a regular phone (via a splitter) to the jack and listen in, to hear whether the DTMF tones are being sent properly?

BTW, what does the DAA look like? I don't think I ever saw one, although I did once see a 567 UTI, with Votrax. I think someone had gotten it for the Votrax speech synthesis, and wasn't actually using the phone interface at all, but I'm not certain.

[Mark]

I used a totally unofficial homebrew DAA to send and receive OSI KCS audio data through the phone before I had a modem. I recall it consisted of a 600ohm matching transformer and a couple of capacitors. Worked pretty well overall. The basic procedure was to call your buddy, set up the transfer, jack the OSI audio onto the phoneline & listen quietly for the frantic "Stop! There was an error!" coming from the receiving end. Heh.

I've used the 9v battery circuit to connect two modems locally to simulate phone lines. The only thing I had to do was to disable dial tone detection before the sender tried to "dial", and of course manually answer due to no ringer voltage.

[bxdanny]

Somehow, the telcos were able to charge $200 for official, legal DAAs. At least I seem to remember reading that. I'm sure plenty of people used "unofficial" ones, but I guess OSI couldn't come right out and tell people to do that, in violation of FCC rules. They had to say in their instructions to do it the legal way.

I have also received KCS audio over the phone. One time Mike Bassman played his cassette tape of the Extended Monitor into the phone for me, and I recorded it on my tape recorder using (I think) one of those suction-cup pickup coils that you would stick onto the phone receiver. The tape worked fine. (I initially thought there was corruption near the end of the tape because the "Z" command didn't work, but then I found out that's how it really was on the tape version that OSI sold.)

[lowrybt1]

Very cool hearing about KCS audio being delivered over phone lines. Given the bandlimiting on VOIP, that wouldn't today, I suppose. Of course, nowadays we can share mp3 files....

Danny and Mark, I've attached some photos of the Honeywell DAA I'm using. Got it on eBay for $40.

PBirkel, thank you for the links to resources about POTS circuits and operations. When I mentioned the rotary phone, it wasn't in the context of trying to interface with the 567. Rather, I used it with the Cell2Jack to see if the Cell2Jack wasn't/couldn't generate the voltages/cycles to make a rotary phone ring -- thinking that a failure to ring the rotary phone might indicate the C2J device won't cut if with the DAA/567. (Admittedly, making this up as I go in terms of troubleshooting.) The C2J was able to ring the rotary phone.

Danny asks a good question about how the C2J prompts the bluetooth-connected cellphone to send the dialed number to the carrier. It isn't necessary to press pound to make that happen. So, I'm guessing the C2J device waits a moment after receiving the DTMF tones and then sends the dialed number to the cellphone with a prompt/signal to initiate a call.

As for the Votrax, I actually tracked down the person Danny mentioned. Can't recall his name. He obtained the 567 with Votrax many years ago, removed the Votrax module (VSK version) and then unpotted the circuits so he could see how the Votrax VSK module works (that's the module that interfaced with most 567's). Then he threw out the module. Sigh. I've not been able to find another device that used the VSK. The TRS-80's voice synthesizer used a modified VSL module. In theory the VSL can interface with the 567 board (and OSI used the VSL briefly) but the implementation of the VSL with the TRS-80 bears no resemblance to the setup on the 567.

[bxdanny]

Nice pictures, thanks.

But what makes you think that you "tracked down" the particular person that I mentioned? Actually, the more I think about it, the more I think that the unit I saw was actually a sample unit that was sent to the store where I was working by someone at the OSI factory, or rather someone associated with it. Specifically, I think it was Rick Whitesel (whose name is on most of the OS-65U utilities) who sent it to my boss, Hal Pollenz. Hal and Rick remained friends after Hal moved to New York and Rick moved to Massachusetts, and would visit each other from time to time. So unless Rick is the person you were talking about, I don't think you dealt with the person who supplied the UTI card I was thinking of.