Rick's b.log - entry 2011/07/25

mailto: blog -at- heyrick -dot- eu

FYI! Last read at 18:55 on 2024/11/21.

The old phone's ringer

Well, a short test (jacking the phone line directly into the bell!) proved that the Livebox is capable of supplying enough current to ring a mechanical bell.
But something strange is going on here...

As you can see from the annotations, the ringer works by more or less passing the line current into the ringer. I've tried all sorts of permutations, and I get one of three results:

The other phone rings, this one doesn't.
None of the phones ring.
Telephony doesn't work (as if the line has been shorted).

Hmmm... More investigation needed.

How a rotary dial works

The way rotary dial works is by pulsing the line by placing a condition akin to having the phone hung up, and picked up. It is a well known fact that people with good timing can 'dial' numbers by tapping the switch that detects when the phone is hung up. Indeed, in some cases this could be used to bypass crappy payment mechanisms.

The question is, however, how do we implement this in hardware? It is, actually, remarkably simple when you see it in action.

The first part of the puzzle is the finger dial itself. This is a spring-loaded device that, as you insert your finger and rotate the dial, "charges up" the spring. Letting go of the dial will then cause the dial plate to return to its rest position.
Beneath the dial plate is a simple ratchet mechanism. As you turn the dial, nothing happens, the ratchet simply clicks over the teeth. However, when you release the dial, the ratchet engages and pulls the dialling mechanism under the power of the spring.

With reference to the picture above, the 'power' for this device is supplied through the lug in the middle. This rotates, under the power of the spring, as required by the position of the dial plate (i.e. we only need to run through a single tick to dial '1', but we need ten of them to dial '0').
The power from this is transmitted to the curly-gear (the diagonal stick). The bell end of this stick is a friction brake. As the stick spins, the copper bits you can see inside the bell move outward (centrifugal force) and rub against the inside of the bell. This provides a measured speed for the return, so the pulses can be timed. It also stops the dial from snapping back, which isn't good for fingers!
The top of the central cog has nobbles for controlling the switches. We'll see how this works in just a moment.
Finally, the diagonal curly-gear turns the white cog at the bottom of the picture. This is what performs the dial pulsing.

The picture above is the dial at rest. The phone current passes through the phone (the switch on the left is closed) and the speaker is active (the switch on the right is open).

Telephone dialler mechanism, during dial.

Now we see the mechanism during the act of dialling. This is when no pulse is being sent, the default position for any time the dial plate is not at its rest position. The notable things are that when the dial plate is released, the mechanism will engage, and the switch on the right is now closed. The switch being closed means the earpiece has been shorted, thus silencing it from the clicks that could otherwise occur during dialling. There are three connections here, as this phone has the option to also short the microphone.

Telephone dialler mechanism, pulsing during dial.

And here is the final stage of dialling - the sending of the pulse. As you can see, the white cog has pushed the contacts on the right to be open. This "looks" to the exchange as if the phone has been hung up. In this state, no current is flowing through the phone, so the sequence of current and no current can be used to signal numbers to the exchange.
The overall timing of the pulses is controlled by the friction brake, and the relationship of make/break is under the control of the white cog.

Repeat the final two photos as many times as necessary to signal a number. Most of the world, with the exception of New Zealand and a part of Norway, used 1=1, 2=2, ... 9=9, 0=10.

Mini-exchange

An idea that has been kicking around in the back of my head is to make a small telephone exchange, maybe capable of eight extensions, under the control of a small microcontroller. It would be an interesting thing to do with old dial phones, which you can pick up fairly cheaply in boot sales and the like. I got myself another Socotel S63 yesterday for a couple of euros. It's beaten up and manky, but appears to work. Just needs some TLC.

The principal problems to understand are that microcontrollers run around 5V, and telephones use much freakier voltages (in the order of 24V in use, 90V AC for the ringer, and seeing no problem with switching things on and off in a way that needed a capacitor to alleviate sparking). I think the best way to handle this would be to use opto-isolators to monitor the line, and relays to control it. In this way, we can keep the phone part and the controller part electrically separate.
Furthermore, we only need two paths, for with a maximum of four phones, it would require two to make a conversation, thus meaning a maximum of two conversations at one time.
Also required would be tone generators for dial tone, ring tone, and busy tone. These would be a common source, switched into the line as required.
The ringer would have a cadence supplied by a relay. As this could be annoying if left on all the time, not to mention wearing out the relay in no time, the ringer cadence relay is only activated when ringing is required. Additionally, as now two relays are required for ringing, it would act as a safeguard against unwanted ringing current through the system, which being a roughly 90VAC tapped off the mains supply, would not be a particularly pleasant thing to meet by accident.
Here is a rough outline of what I have in mind, and notice how many relays it's going to need!

Honestly, I doubt I'll have the time or inclination to get this built, but if anybody fancies a chat about it, mail me!

Vide grenier

Yeah, it was a Sunday, another vide grenier. ☺
I did okay, got another rotary-dial phone for a couple of euros, and a Belkin USB WiFi adaptor for 50 centimes (!). I might give this a whirl in my OSD, see if it works.

Last year, this vide grenier offered rides in a balloon as an option. This year they upped the epic level...

The pilot was a bit of a show-off, I wonder how many people's lunch got recycled? I wasn't able to go on a ride, too expensive and I think mom would have completely freaked. Never mind, I did it once back in 1992, though I think flying over the French countryside might be somewhat prettier than Bridgwater (though that's infinitely better than somewhere like Bracknell...).

Sorry Rob, I kept my eyes open for a cheap Minitel. Well, for all the assorted rubbish on offer, there wasn't a single Minitel to be found.

McDonald's stingy on the fries... again.

I guess I should be ashamed eating at McDonald's, but there aren't many choices on a Sunday. Either way, my shame is nothing compared to the shame McDonald's should feel offering up the following as a large fries.

Mom's new box o' bits

In a "back to school" special, a French tech company were offloading older machines. I picked up an eeePC 1001HA for a mite under €150. Supplied with Windows XP (yay!), it is basically the same spec as my 901.

At first glance, here are my observations:

French keyboard, yuck - but this isn't a fault of the computer, more where it is. ☺
The display is a double-edged sword. It is better, sure, but the main thing is that it is bright. I mean really bright, my 901 seems dim in comparison. However, it is a shiny display so during daytime use, reflections are a consideration.
Fitted with a harddisc (not an SSD), it feels more capable because Windows can swap out to let more stuff fit into memory. Trust me, when you have Firefox running with a few dozen active tabs, Thunderbird, and then you want to play a video... well, 1Gb is a lot but with the way modern programs fritter away memory, it can seem small at times. But a machine with a genuine harddisc, it can swap. Which means it can seem "bigger" at the expense of a little bit of speed.
The machine itself is better, a 10" display instead of the 8ish of mine. My eyesight can cope, but I think mom might appreciate the larger size. Physical resolution is the same - 1024x600. It isn't a bad size, but I like my smaller profile.
I've not run any hardcore tests on data throughtput. I may try playing an HD H.264 video to see how it copes. With the loop filter disabled, my eeePC can just cope at average bitrates provided there isn't too much motion; though I have a video that is encoded at around 2kbit and a lot of motion, my eeePC just falls over on this. I think the chipsets are the same, but they may not be. Worth seeing if this works on the 1001.
One of the things I really like about my eeePC is the WiFi adaptor adapts to current conditions. Right now in the living room, in sight of the Livebox, I'm chugging along at 54mbps. In my room, the other side of a thick stone wall, the speed can vary between 22mbps (good) to 5.5mbps (not so good). Sometimes it drops lower, but this isn't so helpful. The usual is 18mbps.
Mom's 1000HA and my Android phone don't do this. They appear to negotiate 54mbit and stick to that. Consequently, the area of coverage is less than my 901 due to the inability to step down the speed.

At any rate, it is nice to have my machine to myself, let mom do whatever on her own. Plus I have shoved the remnants of her profile onto the NAS. From my point of view, that's €150-ish well spent - no more "but I'm debugging" vs "but I want to look up <meh> on Amazon", which usually ends up with the inarguable "but I'm your mother!"...

Your comments:

No comments yet...

See the rest of HeyRick :-)

	Rick's b.log - 2011/07/25
It is the 21st of November 2024 You are 18.189.185.63, pleased to meet you!