News:

"The phone is a remarkably complex, simple device,
and very rarely ever needs repairs, once you fix them." - Dan/Panther

Main Menu

3-wire Phone on a 4-wire network

Started by Craterranch, September 16, 2015, 07:39:24 PM

Previous topic - Next topic

twocvbloke

Quote from: Phonesrfun on September 18, 2015, 09:50:15 PMHow long it takes to charge, I don't know.

Unless they're the size of a pepsi can, then they should charge up within a second, so not sure how they're supposed to take longer... ???

19and41

I remember reading some of the reviews of the Rotatone converter and it was said there needed to be a time period for the unit to accept a charge.  I don't know why they can't adapt such a circuit to accept a backup battery or other power source instead of having to try a series of adapters to see how each responds to environments that most phones should be able to operate within? 
"Any sufficiently advanced technology is indistinguishable from magic."
— Arthur C. Clarke

unbeldi

#32
Not knowing what the internal resistance is of these units, and the capacitance of the storage cap, it is hard to be precise about charging time.
However, we know that the maximum current delivered on a loop is rarely ever larger than 50 mA these days. On ATAs it may only be half that.

So, just as an illustration, let's say the storage capacitor is 1 farad, at 50 mA charging current it would take ca. 100 seconds to charge. But we cannot let the unit draw that much current, because this would be detected as an off-hook condition by the exchange. So, let's drop that down to 10 mA, perhaps still too much, and now the charging time goes up to 500 s!  Time constant is  t ~ R x C, where R is the internal resistance and C the capacitance.

If they use a real supercap instead of a very large capacitor, it can have a capacitance much large than 1F. But one would need at least two of them in series, because supercaps are only good for about 2.5 V, and we should run the microcontroller at least at 3.3 V.

Given the ubiquity and low cost of cell phone chargers these days, it surely would seem opportune to simply provide a connector on these gizmos to accept external power, when needed.   Convenience is great, but all too often, the line power is barely enough to run an old phone, and for the exorbitant cost of these units, even those Chinese versions,  the design choice seems baffling.

19and41

Well said.  If someone could work up a modification for at least one of these, I think there would be a market among us who do not get the proper operation of these devices.  I've not seen one yet that did not have a problem in some applications.
"Any sufficiently advanced technology is indistinguishable from magic."
— Arthur C. Clarke

Craterranch

This issue has been resolved. I purchased the Dialgizmo and everything works correctly. The problem must have been the adapter I got from China...either it didn't work on our office system, or it was just defective. I'm guessing the latter since it didn't work at a friend's house either. Thank you everyone for your suggestions.

twocvbloke

Given the misplaced components on the board, I'd say defective due to a manufacturing defect...

19and41

There are a few types of these devices that have too critical a leeway in their operating voltages.  The Dialgizmo seems to be most tolerant of the lower voltages.  Others may or may not work depending on the line voltage given them.
"Any sufficiently advanced technology is indistinguishable from magic."
— Arthur C. Clarke

unbeldi

#37
Quote from: 19and41 on September 29, 2015, 12:19:19 PM
There are a few types of these devices that have too critical a leeway in their operating voltages.  The Dialgizmo seems to be most tolerant of the lower voltages.  Others may or may not work depending on the line voltage given them.

The line voltage is actually not the problem.  Except for key systems, almost all subscriber lines provide ca. 48 V line level. This applies also to the vast majority of analog telephone adapters (ATAs) and embedded multimedia adapters (EMTAs) in cable modems, because the international DOCSIS specifications require it for a device to be labeled compliant.

The actual problem is the available current on the line, many line interface circuits only provide 25 mA anymore, which is just barely above the operating limit of historic telephone carbon transmitters.  Think of a telephone line as a constant current source, not a voltage source.   The voltage collapses when going off-hook to below 10 V, likely closer to 5 V.  5 volts is still plenty to run any of these gizmos, most will probably work as low as 3.3 V.   The typical DTMF generator chips these days, work still at around 2 V.

Aside from manufacturing defects, as may be the case in the discussed unit, the reliability of these converters depends on the design decisions and  the installation environment. Telephone lines in combination with telephones presented a vast spectrum of differing electrical conditions, not just today, but throughout history, and this diversity was a major engineering challenge for  the telephone companies and equipment makers, resulting in very exacting specifications and prohibition of foreign equipment on lines to control quality.
So it's not surprising that these devices don't always work.

luns

#38
Quote from: G-Man on September 17, 2015, 12:33:26 AM
Here is another photo captioned with very explicit instructions.




For the benefit of those who can't read said instructions;

The text at the top says to insert the phone line crystal head (mod connector).
The text at the lower right says to connect the telephone equipment to the two indicated lines (P1 and L2).

Presumably L1 and P1 indicate Line and Phone respectively, while L2 is common to both. Connecting the phone to L1/L2 would be bypassing the dialor and equivalent to just plugging the phone into another outlet on the same line (the green/red wires that go under the board just go straight to the socket up top).

luns

#39
This may be moot given Craterranch is already up and running with the Dialgizmo, but I have a hunch what's going on, and would like to suggest a few things to try if you're curious. Feel free to ignore them if you're not.

I don't know what sort of isolation the dialor has between its supply and the off-hook phone load. The way the dial is connected, the transmitter is shunted during dialing, so the phone pulses between being a dead short, and an open circuit. Not all phones behave this way; some don't have this shunt and pulse between being simply off hook and open circuit (on hook).

I'd be curious to see how the dialor works with the 302 with this shunt disconnected. Disconnect the wire going to the R terminal on the dial and see if that gets you any further than before.

The shunt normally bypasses the induction coil, so with it disconnected, the dial pulsing will be going through that and the transmitter. I don't know whether the inductance of the induction coil might have other odd effects on the dialing, and it may be hard on the dial pulse contacts if you don't have the 61a filter on the dial. There's a more involved modification to the wiring that we could try that would get around this. Let me know if you're interested and I can break it down into steps for you.


Oh, I should also add... I think the seemingly dislocated components may actually be a deliberate revision to the circuit after the circuit boards had already been fabricated. There are footprints for a diode, an LED and a resistor, but from the looks of it, there's two resistors, the LED and a transistor on there. This wouldn't happen by accident.

Craterranch

Quote from: luns on September 29, 2015, 08:22:44 PM
This may be moot given Craterranch is already up and running with the Dialgizmo, but I have a hunch what's going on, and would like to suggest a few things to try if you're curious. Feel free to ignore them if you're not.

I don't know what sort of isolation the dialor has between its supply and the off-hook phone load. The way the dial is connected, the transmitter is shunted during dialing, so the phone pulses between being a dead short, and an open circuit. Not all phones behave this way; some don't have this shunt and pulse between being simply off hook and open circuit (on hook).

I'd be curious to see how the dialor works with the 302 with this shunt disconnected. Disconnect the wire going to the R terminal on the dial and see if that gets you any further than before.

The shunt normally bypasses the induction coil, so with it disconnected, the dial pulsing will be going through that and the transmitter. I don't know whether the inductance of the induction coil might have other odd effects on the dialing, and it may be hard on the dial pulse contacts if you don't have the 61a filter on the dial. There's a more involved modification to the wiring that we could try that would get around this. Let me know if you're interested and I can break it down into steps for you.


Oh, I should also add... I think the seemingly dislocated components may actually be a deliberate revision to the circuit after the circuit boards had already been fabricated. There are footprints for a diode, an LED and a resistor, but from the looks of it, there's two resistors, the LED and a transistor on there. This wouldn't happen by accident.

Thanks for the detailed information, however I have already sent the Chinese unit back for a refund (I hope).

EdPDX

I have had one of these for 3 or so years just sitting on a shelf as I had no PDF or instructions for the device when I bought it new. Just this morning I hooked it up to a 3 wire ARMY phone (TA236/FT).  Here is how I did it:

L2= Green and Yellow
the terminal directly above it- joined by the legend <--PHONE--> = Red wire
Plug a standard reverse line into the DIALOR2.0 and you should get a dial tone. Call your phone and you should hear it ring. Answer and hear yourself.

I have not figured out how to get a dial tone as advertised. I'll post hear if I figure it out.