Anyone Experienced with the Dialor 1.2?

[MIPS]

I got no problems taking extra photos if you need a better idea on something here.

[unbeldi]

Here is the current state of my diagram, after gleaning the meaning of R6 from Luns' posting.  This was a missing piece of info, as the transistor Q4 was missing a source of emitter potential to drive the base of the Darlington pair.

Please examine whether this matches the actual board now.

The off-normal (ON) contacts of the dial have to connect to D (pin 13 of the µP) as a switch to turn the power circuit on for the ICs.

@MIPS:  thanks for that picture. Indeed your black/white pair is the DP switch, connected to the P input. So we are all in agreement on that.

[unbeldi]

On the diagram, I labeled  the Zener diode with a 5.1 V value.  This may not be correct, as I remember reading elsewhere that this was pegged at 2.6 V, I can't remember the post right now.  Was that one of you?
Is it possible to read the fine print on that diode?

I would also be interested in the value of R3.

[MIPS]

The zener is a 1N4733A. R3 is 370 ohm.

This may not be correct, as I remember reading elsewhere that this was pegged at 2.6 V, I can't remember the post right now.  Was that one of you?

That's me and it's kind of weird. I know that the DC voltage of a telephone system is 48v. If any device in the house goes off hooks that will drop down to 5.6-2v, depending exactly on what device it is. (cordless, modem, rotary phone in the other room or this adapter) Regardless of the line voltage the voltage at the microcontroller is always between 2.6-2.1v. IMHO, kind of low.
No idea why it varies in a static test.

On the bright side, that schematic looks to be correct.

EDITED: It's a power issue. We are not delivering enough power to the microcontroller.
I repeated the experiment seen in the video above but with a multimeter on the G terminal and VDD. The moment you ground out the line the voltage goes up to +5 and everything works. Remove ground...and it drops down to 2.1 which is too low. Thing is I can't see an easy way to isolate and externally power it.

[unbeldi]

What is your source of loop current and dial tone?  Is this connected to a central office line?

On a good line, with battery at –48 V on Ring, and Tip on ground (or just below ground), a standard Western Electric 500 set should take the line down to ca. 10 to 12 V from Ring to Tip. I measured this on a central office simulator (Sage 930A) that can put out as much as 100 mA between Tip and Ring, and over 200 mA between Ground and Ring.  A 1960 Princess Telephone took it down to ca. 9 V when going off-hook.   On a Grandstream HT502 analog telphone adapter, which can provide up to ca. 30 mA of loop current, this dropped down to ca. 5 V.

A polarity guard, such as that in the Dialor (the diode bridge D1...D4), should only cause a minor voltage drop of no more than 1.4 V with silicon diodes. That's also what I found on a unit constructed from four 1N4005 diodes.

So the question is, why are you only reporting such low values?   The fact that your circuit starts working when grounding, tells me that something is wrong.  Ordinarily, when using a polarity guard, the circuit powered by the bridge must not be grounded on its common reference level.

Is your primary line grounded in any way or is it floating?
You have a modem involved, from I understand, does that introduce a ground through its power supply?  If it does, you may have to reverse polarity of the line to create a proper configuration.


Ok, that is the type of Zener I allocated too.
If you want to read all remaining resistor values, I'll add them to the diagram for posterity and future generations.

[MIPS]

What is your source of loop current and dial tone?  Is this connected to a central office line?

Loop comes from the wall and out to the pole. We don't use VoIP or similar systems here. Local exchange is about a kilometer away.

On a good line, with battery at –48 V on Ring, and Tip on ground (or just below ground), a standard Western Electric 500 set should take the line down to ca. 10 to 12 V from Ring to Tip. I measured this on a central office simulator (Sage 930A) that can put out as much as 100 mA between Tip and Ring, and over 200 mA between Ground and Ring.  A 1960 Princess Telephone took it down to ca. 9 V when going off-hook.   On a Grandstream HT502 analog telphone adapter, which can provide up to ca. 30 mA of loop current, this dropped down to ca. 5 V.

I was noticing that. There was a voltage drop and it varied between the type of phone but it was always rather average in where it sat off hook.

A polarity guard, such as that in the Dialor (the diode bridge D1...D4), should only cause a minor voltage drop of no more than 1.5 V.  At least that's what I found on a unit constructed from four 1N4005 diodes.

When the line is at 3.6v and I'm seeing around 2.4v-2.1v at VDD that sounds about right as well.

Is your primary line grounded in any way or is it floating?

Absolutely no idea.

You have a modem involved, from I understand, does that introduce a ground through its power supply?

It shouldn't. The Hayes Smartmodem 300 needs 13v AC and should remain isolated from the phone system or else things would misbehave.

If you want to read all remaining resistor values, I'll add them to the diagram for posterity and future generations.

I hate metal film resistors. I don't have any reliable cheat sheets on hand to decode the color bands. Best I can do is give you the bands on each one.

R1: brown-red-black-black-brown
R2: brown-red-black-black-brown
R3: orange-blue-black-black-brown
R5: brown-red-black-black-brown
R6: brown-brown-black-black-brown
R9: brown-brown-black-black-brown

I'm still just mindlessly experimenting here so pardon the following.
I'm still curious about the DC voltage from the line. As another test I hung a DMM off the line and took just the dialer off hook. 1.3v. No good. Tried to dial. The moment the D contacts closed the line went up to 5v. Dialed my cell. It worked.

Tried the above again but with another device off initially hook. This time we had 3.2v on the line. Took the dialer off hook and tried dialing. Voltage never increased. Put the other device back on hook and tried with the dialer still off hook. Voltage went up to 5v and dialed out.

Tried the opposite. Took the dialer off hook and dialed one number to bring the line voltage up, then took another phone off hook. Voltage dropped down to 3v. Couldn't dial out. Put other phone back on hook and continued dialing number, voltage went back up and it dialed out. It's almost like the loop is limiting my voltage when it detects another device on the line....or something.
If you want a visual demonstration of this I can make another video if you would like.

[unbeldi]

I hate metal film resistors. I don't have any reliable cheat sheets on hand to decode the color bands. Best I can do is give you the bands on each one.

Thanks, that is fine.   The 5-band resistors are pretty much the same, simply insert a third digit after the first two rings, because they needed the extra precision.  Below is a chart from Digikey for 4 and 5 bands.

R1: brown-red-black-black-brown  = 120 Ω 1%
R2: brown-red-black-black-brown  = 120 Ω 1%
R3: orange-blue-black-black-brown = 360 Ω 1%
R5: brown-red-black-black-brown = 120 Ω 1%
R6: brown-brown-black-black-brown = 110 1%
R9: brown-brown-black-black-brown = 110 1%

I'm still just mindlessly experimenting here so pardon the following.
I'm still curious about the DC voltage from the line. As another test I hung a DMM off the line and took just the dialer off hook. 1.3v. No good. Tried to dial. The moment the D contacts closed the line went up to 5v. Dialed my cell. It worked.

Tried the above again but with another device off initially hook. This time we had 3.2v on the line. Took the dialer off hook and tried dialing. Voltage never increased. Put the other device back on hook and tried with the dialer still off hook. Voltage went up to 5v and dialed out.

Tried the opposite. Took the dialer off hook and dialed one number to bring the line voltage up, then took another phone off hook. Voltage dropped down to 3v. Couldn't dial out. Put other phone back on hook and continued dialing number, voltage went back up and it dialed out. It's almost like the loop is limiting my voltage when it detects another device on the line....or something.
If you want a visual demonstration of this I can make another video if you would like.

I think when using the telco loop, or any loop for that matter, it is good practice to characterize it without any load before experimenting further.   I always measure:

a) loop voltages
- between T and R
- between ground and T
- between ground and R

For a good ground you can use the ground pin on any of your electrical outlets, or the ground on your oscilloscope, which should be the same.

b) loop current.
Simply set your DMM to DC current range of 200 mA or so and use the leads to short the telephone line. The DMM has very low resistance for current measurements, so the loop is essentially shorted this way.   The reading gives you the maximum current you can get from the telephone company.

From here on you can start experimenting.  How much does the line voltage drop when you connect a 500-set?  When the voltage drops to only 3.x V, I'd say you are overloading it with something, it should not be that low.

[unbeldi]

ok, final circuit for the DIALOR V 1.2.

[luns]

EDITED: It's a power issue. We are not delivering enough power to the microcontroller.
I repeated the experiment seen in the video above but with a multimeter on the G terminal and VDD. The moment you ground out the line the voltage goes up to +5 and everything works. Remove ground...and it drops down to 2.1 which is too low. Thing is I can't see an easy way to isolate and externally power it.

This is what I suspected over in the other thread. If you're okay with battery power, you could try disconnecting R3 from the diode bridge, and connect it to the + terminal of a 9V battery, and put the - terminal to VSS. This would drain the battery pretty fast though, so you'd need to disconnect it when you're not using it (another relay with your hook relay would do it).

But I'd still rather see you just try the dial directly. If you missed it in the other thread, just put the DP contacts in series with the modem.

R1: brown-red-black-black-brown  = 120 Ω 1%
R2: brown-red-black-black-brown  = 120 Ω 1%
R3: orange-blue-black-black-brown = 360 Ω 1%
R5: brown-red-black-black-brown = 120 Ω 1%
R6: brown-brown-black-black-brown = 110 1%
R9: brown-brown-black-black-brown = 110 1%

I think you're misreading the multipliers. Taking R1/2/5 for example, reading the stripes out straight, they're 12001. The Final 1 means to insert one 0 there, so the resistance is 12000 or 12k. R3 would be 36k, and R6/9 11k.

ok, final circuit for the DIALOR V 1.2.

One more revision to the figure; the Darlington pair's collectors should go to V1 (together with Q1) instead of VDD.

U1 smells like an Atmel 2051 to me. I think it's spec'd for VDD only down to 2.7V.

The Q2/Q3 darlington is somewhat puzzling. I suspect it's there to try and mimic how the tip/ring voltage would increase during dialing with early touch-tone pads (there's a varistor that goes in series when a key is pressed that gets shunted off otherwise), but does so by pulling down the voltage the rest of the time. The timing of when it does so doesn't really make sense - it ought to do so only when emitting a tone, and not necessarily when the dial is turned.
I don't know if a CO would care about this anyway, and I think it's bringing the voltage into a brown-out sort of range that it might not recover enough from even if the voltage comes back up afterwards (without a proper reset on pin 1).

I would try disabling the darlington. There's a few ways to do this, but the easiest is probably to just remove R6.

[unbeldi]

R1: brown-red-black-black-brown  = 120 Ω 1%
R2: brown-red-black-black-brown  = 120 Ω 1%
R3: orange-blue-black-black-brown = 360 Ω 1%
R5: brown-red-black-black-brown = 120 Ω 1%
R6: brown-brown-black-black-brown = 110 1%
R9: brown-brown-black-black-brown = 110 1%

I think you're misreading the multipliers. Taking R1/2/5 for example, reading the stripes out straight, they're 12001. The Final 1 means to insert one 0 there, so the resistance is 12000 or 12k. R3 would be 36k, and R6/9 11k.

The final 1 means that the tolerance is 1 %, the first three bands are the digits, the fourth is the multiplier, which is 1 Ω (black) in every case, so here we need only read the first three bands, in fact.

[unbeldi]

One more revision to the figure; the Darlington pair's collectors should go to V1 (together with Q1) instead of VDD.

Oh, you're saying I have the resistor in the wrong place.  Yes, looks that way.

Like this:

[luns]

The final 1 means that the tolerance is 1 %, the first three bands are the digits, the fourth is the multiplier, which is 1 Ω (black) in every case, so here we need only read the first three bands, in fact.

You're absolutely right, my bad. When I read resistor values, I automatically ignore the precision, and read out the remaining bands. It should have occurred to me there wouldn't be four significant digits.

Oh, you're saying I have the resistor in the wrong place.  Yes, looks that way.

Like this:

It looks correct now, though I would put R3 vertical instead of horizontally - the way it's drawn, it looks like the darlington is somehow feeding the R3/DZ regulator, but they're supposed to be more independent than that. At the same time, V2 as you have is not the best node to split since both R6 and Q4 have a say in what the actual voltage is, but they're in different places. Pin 11 would be a better place to split at; you can treat it as a voltage source - even if it has a nonzero output impedance, that can be thought of as being part of R6 instead.

[unbeldi]

U1 smells like an Atmel 2051 to me. I think it's spec'd for VDD only down to 2.7V.

Is there a variety of them that doesn't require any external clock components?
I also suspected this at at first, and the grounding resistor on pin 1 would be right to keep the reset logic inactive, but our board doesn't seem to have any clock input to the chip.

[dsk]

Thank you for this thread, I ordered this: http://tinyurl.com/o66q2av
(first picture) So I did not need to solder.
What I got, second picture, and how it looks in my payphone.
last picture.

I got the one pictured in the start of this thread.

The 2 in wires was put in to where the pulse contacts should be, and the  G D and P terminals was used for G=common P=pulse contact  (N.C.) and D was the shunt contact usually shunting the receiver.

Works well on 24 and 48 V systems.  I have not figured out how to find * and # but holding 1 to I hear a beep, and release =  redial.

dsk

[dsk]

Now I found, if you need a * after dialed some digits, just dial 1, and hold to you hear a tone. # is 2 and hold.
But
If I want to dial e.g. *21# then I have to dial 1 and hold to I have heard 2 beeps, and release, then 21 and 2 and hold to I hear the beep, and release.  I'm sure it is other functions, but I have not figured it out. So a manual in English or Norwegian cod be convenient.

dsk