News:

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

Main Menu

Dial tuning, is it good enough or not?

Started by dsk, April 20, 2014, 03:27:04 AM

Previous topic - Next topic

dsk

Hi
I had to fix a non working W.E. 5H dial, and by using eye and ear, I got a result pretty well working.
The spec are probably not OK but its working, and by PC monitoring in Audacity I got this result.
What do you think?

dsk

dsk

#1
Now it is tuned. (its fun to play with this)
The speed varies a little around 10 pulses pr sec. That's better than to have higher tension on the spring.
The break ratio are no approx 65 %  (UK standard are set to 2/3 or 66.7% and US and Norway uses 60%)
The exchanges use to accept both standards so this seems reasonable to me. (my own standard  ;) )

dsk

Mr. Bones

Dag,

The vast majority of members here know infinitely more than I do about dials. That being said, 10.07 p.p.s. certainly ought to be more than close enough. Looks like you got the make / break ratio down pretty good, as well.

I've had Audacity for years; will have to figure out how to utilize it in this fashion.

Best regards!
Sláinte!
   Mr. Bones
      Rubricollis Ferus

Scotophor

#3
It looks reasonably good to me. Info I have from various sources gives slightly different details regarding U.S. dial specs. The one I consider most authoritative has the following:
Pulse rate = 10 pulses per second nominal
phones being refurbished should be adjusted to within 9.5 to 10.5 p.p.s.
phones in use are considered OK within 9.0 to 11.0 p.p.s.

Break-to-make ratio = 62% break nominal
phones being refurbished should be adjusted to within 62 ±2% (60% - 64%)
phones in use are considered OK at 61.5 ±4% (57.5% - 65.5%).

If your phone dials out reliably it may be just fine on your system (especially if it is meant to handle both U.S. and U.K. dials), but according to this, 65% break is very close to the limit for U.S. dials.
Name: A.J.   Location: LAPNCAXG, EDgewood 6

dsk

I've been asked to share how I do this.
Assuming you know Audacity (free) you plug in a suitable cord. to the microphone jack.
(I have used one with male plug in both ends. )
Connected tip and sleeve to the pulse-contacts (or line cord) with some test cables with clips.
To get a better reading adding a sound may help, I just connected my signal generator (picked from a container) to the same 2 points.
I guess if you use a radio with some hissing sound, you should protect the radio by using a resistor in series with the output from the radio.  You have toe experiment withthe record level, I used some 30%.   Start recording, dial zero, stop recording.
Zooming in to start and end of the first and last break and makes between pulse 1-2 and 9-10 will give you pretty good timing data.

dsk

dsk

The last thing is the math, i have made a simple spreadsheet, using the 6 points marked in the first picture and calculates what I believe are reasonable good data, based on visual observation of a smooth and nearly constant speed on the dial return.

I have not found any way to share the spreadsheet but a pm to me with your email address and I'll mail it to you.

dsk

dsk

Always learning!
Today my Norwegian payphone went to strike! The dial would not work. I picked it out, and measured up and down in every manner I could, its perfect! nearly 0 ohms at closed contacts, breaking properly, perfect rate 11pps.

At last after testing with other dials, no it has to be that dial, a little drop of greasy oilish gel was in the pulse contacts, so cleaning again, and yes its working. This kind of errors are extremely frustrating, and hard to find.

dsk

dsk

Audacity is just great.

dsk

I got the question about why  measuring 9 pulses, here is my answer:
The pulse registered by the exchange is break.  The exchange still need a pause between the breaks, this has been deemed to be with a preferred length. Here we use 60% break, 40% make, UK use 2/3 to 1/3.  If the make is considerably shorter, it may not be recognized, and the 2 breaks looks like one long break, or a flash.   After the last break the circuit remains closed, so it will be hard to tell the length.  That makes it hard to measure the last make, so I measure from the first break opens, to the tenth opens. Only 9 pulses  (brake   make) should last for 9/10 of a second. (at 10 p.p.s.)
Then I calculate those 9 divided on the time consumed (displayed in audacity) and get out the dial speed pretty accurate.

dsk

andre_janew

Of course, the real test comes when you try to dial out with the phone.  This is the most important test of all!

dsk

You are absolutely right, and many exchanges accept great inaccuracy, this are more a tool during tuning. No need for repair when it works.

dsk

dsk

#11
Here is an example of a slow dial:
(approx 7 pps and it should be 10.)
dsk

TelePlay

#12
And this is a WE #6 dial as removed from a phone showing a bit of a problem. Audacity not only helps determine PPS but is also a great tool to identify why or where the pulse problems are happening.

This excel spreadsheet I devised lets me calculate the PPS over any number of digits dialed, the break/make ratio of any digit, the individual break/make ratio of all digits and the average break/make ratio of all 10 digits dialed. Overkill, yes, but  it can be used to find minor flaws or dirt in the gear train to be corrected.

This spread sheet is for a #2 dial that is a bit slow, not the #6A dial with problems shown in this reply.

TelePlay

This is the improvement from one ultrasonic cleaning and no oil yet applied. Same WE #6A dialed in the last post but working better.

dsk

This should be good enough, a little oil may do a difference, and at least make it live longer. You should probably not do any adjustments between 9 and 12 pps.
dsk