atcomsystems.ca/forum Forums Other TDM Phone Systems 1A2 and Comkey Telephones & Systems Chaining 66 blocks with AMP connectors

look at three way amphenol bridging adapters. cpc or pcp.

Greg - that is too cool! Looking forward to seeing more of this project.

Jeff grow up it is all bullshit and wild imagination !

Thanks Jeff! I'll try to follow up with schematics and whatnot. Glad someone's interested.

Here's the circuit layout of rev B of the board. I included as much info as I could about the components on the silk screen mask so it could give someone a hint of what's going on sans schematics.



This is revision B of the board.. one thing that's wrong is the direction of the diodes for the "BELL CALL" section on the silk mask. When I got the boards back from the printer and soldered it up, phones wouldn't ring unless I desoldered the diodes and flipped them all around. Will fix that in rev C. Was surprised it mattered, I thought the bells would ring on either side of the AC cycle.. didn't catch that in R&D.

Schematics as they stand, 3 pages so far.

It needs a 12v/1.5A wall wart supply to properly run up to 4 phones.

Empirical tests with my VOM monitoring the 12v power consumption (+/-10mA) with 4 phones and 2 lines, it shows the card uses:

• 20mA when idle (no matter how many extensions)
• 800mA when in full use with Lines 1+2+5 off hook w/4 extensions
• Someone buzzing an extension adds 60mA

In the 800mA case, this means 12 lights in total are lit (3 per extension), 4 relays energized (2 per line), plus power consumed by the hybrid networks for intercom's and the active DTMF circuit.

It uses about +60mA when buzzing an extension.

It uses quite a bit more if a PowerDSINE is using the 12v to actively ring all 4 extensions; I couldn't measure it well because the 1Hz flashing line lamp was throwing off the reading, but the highest reading I saw seemed to add +500mA for ringing.

I don't have one of the Cambridge "black magic" ring gens to test, but I imagine they'd be similar in power consumption. I'll try to get a few for testing.

Of interest, during development, I used the AC ring generator from my old AT&T 551a, which I think puts out 30Hz (I never checked), but when my +12v PowerDSINE modules arrived, which put out 20Hz, gotta say, I prefer the powerDSINE's slower 20hz; a less strident ring. I should probably post a wav file comparing the two rings.

When I make a rev C of the board, I'll include a connector that makes it easier to attach the PowerDSINE ring gen, so that it can tap switched 12v from the board, so the ring gen is "off" when no ringing is present.

Anyway, on to the schematics...

* * *

Schematic - Page 1/3: Covers the relay logic used for the lamps, the A lead handling from the extensions, and the ring and line detect optocouplers, along with the hold logic path through the relays for the hold resistor. (A 160 ohm 1W resistor at the lower right). Small DPDT relays are used for the A lead handling (relay "A") and the line detect (relay "L").

Schematic - Page 2/3: Covers the oscillators for the lamp flash (2 Hz hold "wink") and ring flash (1 Hz during ringing), and the circuit that stretches out the ring detect to keep the ring lamp flashing in between the rings, and finally the ring lamp circuit that mixes the ring stretch and ring oscillator to flash the lamp.

Schematic - Page 3/3: Covers incoming call ring programming for bells, and I added a separate section that allows the buzzers be the audible for incoming calls, allowing one to hear calls without a ring generator. And finally, at the bottom, the DTMF detect circuit used to handle extension buzzing for the intercom.

* * *

Technical questions/feedback welcome.

Quite the job!


Sam

Here's a pic of the history of the project..

From top to bottom, the R&D prototyping boards which I started around the time I joined this forum, so exactly 10 years. Spent years on and off (mostly off) noodling the circuit and reading up on analog/phone circuit designs, studying patents, BSPs, product schematics. Then I started designing the board's circuit layout. At some point last year the project's inertia finally picked up to where I felt ready to print some boards, which became Rev A. Made some fixes, and a month later printed Rev B which I'm happy with:



I've done a bit of industrial circuits before, and this was no different, albeit more challenging. In particular, I found patent #4,093,830 the most inspiring; here's my version of it as a sketchbook drawing I made early on in the project:



It's a very simple line card circuit, and put me on the road to using a similar pair of optocouplers/pair of relays. I also liked the circuit description. It's clever use of transistors and the rarely used pin 6 on the optos impressed me too, but I didn't use that in my design, preferring my own techniques.

To layout the board, I used ExpressPCB's free software, which I highly recommend; very easy to learn and use. And their board printing service is quite good. Using that software locks you into using their service though; you can't take a file made with that software to other board printers. But this was fine with me; their printing service seemed fine, and I don't usually switch printers once I find one I like.

One of the problems I caught in the Rev A board: I took what I thought was a clever shortcut, using negative logic to run the lamps. This involved controlling the lamps via the ground wires, and tying the lamp signals common. This seemed OK because in the 2565 phone I used for R&D, the pairs of lamp wires were isolated. But turns out that /didn't/ work for the 2564 phones, which buss the lamp ground inside the phone. So fixing that involved switching to positive logic; NPN transistors became PNP.. which changed the gate logic a bit, causing more components. So Rev B works for all the 1A2 phones I had handy; 2564,2565, and 2830. Changed a few other things too, including powering the DTMF circuit only when the intercom is in use.

I'm pretty happy with the "Rev B", but will probably do a "Rev C" just to add a feature for the PowerDSINE ring generator, so it can be added in such a way that it's only powered when there's ringing, and completely off when the system is idle, so idle power consumption is low.

My goal with this project was to use old chips that were available in the 80's. No one-board computers or software. I could have easily used a PIC chip, Arduino, or RaspberryPi to do a lot of the timing and oscillators.. but I wanted to learn more about analog electronics, so working within that design goal allowed me to immerse myself in analog a bit more than usual. Consulted with an EE friend from time to time for general guidance.

I standardized on LM339 ICs as the main component for the logic and oscillators; these are "comparitors", a flexible gate-like device that are like opamps, but tailored to this type of analog work. I used power transistors that are overrated; they don't get hot or even warm during full use. Wanted to avoid having too many relays. I used relays where it made sense, such as redirecting tip/ring paths and ring voltages. Originally I'd used solid state for ringing, but decided against it for cost and flexibility.

My goal of avoiding computers was so that one could just assemble this board and it would work; no USB programming or software. Just wire it up and go. All the "programming" is in the circuit traces, and choice of components.

Another design goal was to have the board be easy to setup phones without 66 blocks, and use for the most common use; up to two lines, up to 4 phones, with programmable ringing and intercom with simple TouchTone intercom signaling. Basically all the features one would want out of a 1a2 system for a typical small office with front office receptionist and other employees. Intercom was important I felt as a default. Basically tried to get as close to a ComKey setup, but using 1A2 phones.

Will try to follow up with a short video showing 4 phones and 2 active lines, demoing ringing + buzzing + lights. 1A2 is so simple, there's not much to show really.. but it's nice to see it all working.

Greg:

Some random thoughts:

I'm not sure why that other fellow felt the need to curse out Jeff for being enthusiastic about your project. That said, I am intrigued by it. I don't find any part that's bovine excrement.

Have you, by any chance, taken a look at the "KSU-less" version of 1A2 that is used in the ComKey 416 system?
There is a 5th line add-on box that contains the electronics to provide "one more" CO line to a system that already is using its maximum of 4 lines. I have a spare if you'd like to hold one in person and see what makes it go.

You are correct in your discovery that 20Hz sounds much nicer than 30Hz. Thirty is used in subscriber key equipment power units because it's easier (cheaper) to derive than the "better" 20 Hz, used in PBX's and CO's. Cord board buzzers in models 550, 551 and 555, especially, sound so much better when powered by 20Hz. It produces a nice purr.

The Black Magic® 20Hz supplies are powered by a 5-volt input. I may have seen some whose inputs were (ISTR) 12, 24 and/or 48 Vdc.

It's too bad that you're not closer to Lancaster, PA, where, in June of every year, a hundred or so telephone collectors, and especially, "switchers" gather to display and play with their home-brew PBX and key system equipment. We are all members of C*NET. Check out www.CKTS.info.

I'd be delighted to speak to you privately about your project and a few of mine.

PSTN : 631-sevin for nyne-0100

C*NET: 1+ sevin for nyne-0100

EXchange numbers obscured to repel 'bots.

Jeff was not cursed out . It was my opinion he could take it or leave it . You should not Mother him he is a grown man and seems like a good person .

I think this is just fantastic Greg.
I was actually curious why you didn't have any processors or Raspberry Pi involvement or the like. It was interesting to read your design model above.
As I have thought about this type of thing over the years I had thought how cool it would be to have a phone module to plug each phone into that could communicate over wifi or ethernet to a central KSU or possibly Asterisk appliance so one wouldn't have to run 25 pair wire everywhere.

I actually built a 2 line phone system that was similar in function to 1a2 from scratch in the early 80s when I was in jr high school. I still have it somewhere, Ill have to post pics of it. You guys would laugh your asses off but it worked for my dads office!