Bell 103 compatible modem questions

[neazoi]

Hi, I investigate some of these old bell 103 300 baud modems schematics and I would like to ask a few questions.

1. The penny whistle modem on page 6 says "unit to unit communications" and it says that the high/low band switch must be used for the purpose. So this unit is meant to be used full duplex when calling a "server" and not actually for full duplex communications between two such modems, am I right?

2. The LDS modem uses the transformers primaries and secondaries (with parallel capacitors) as audio filters for distinguishing between the two tones?

3. The LDS modem is only capable of communicating with a server in full duplex. There is no provision for two such modems to communicate between them, either full or half duplex?

Please let me know, so that I know if I am thinking on the right path according to meeting the bell 103 specs that I am reading about.

 

[betwixt]

1. They can work full duplex but you have to manually configure them so one uses the low frequency pair and the other uses the high frequency pair. At 300 Bauds it is possible to fit TX and RX pairs 'one above the other' within telephone bandwidth. There is no automatic configuration method that ensures they work properly and obviously if they are not set to opposite band their signals collide.

2. Yes, but note that acoustic modems basically became redundant around the late 1970s. Having used them I can confirm they were horrible, even someone sneezing nearby could break the data link!

3. They can talk directly to each other if correctly configured. A server isn't required. Where it becomes difficult is the telephone line itself, acoustic modems use the telephone handset microphone and earphone so the 'hybrid' in the phone electronics is responsible for splitting the transmit and receive paths. If you just join modems together with a cable, they won't work.

As there is only one pair of wires carrying DC, ringing voltage, transmit audio and receive audio to a telephone they have a method (using transformers and resistors in old style phones) to route incoming audio to the earphone and microphone signals outgoing to the line. Basically the line contains a mix of both but as you know what you are sending, if you subtract it from the line signal, anything left must come from the other end.

Brian.

 

[std_match]

1. Yes, the Pennywhistle modem can only receive on a fixed frequency pair (2025 and 2225 Hz), so full duplex is impossible with such simple hardwre.
The transmit frequency pairs are switchable, so half duplex is possible.

3. The LDS modem has both the receive and transmit frequencies fixed, so two such modems can't communicate at all with each other.

 

[neazoi]

1. Yes, the Pennywhistle modem can only receive on a fixed frequency pair (2025 and 2225 Hz), so full duplex is impossible with such simple hardwre.
The transmit frequency pairs are switchable, so half duplex is possible.

3. The LDS modem has both the receive and transmit frequencies fixed, so two such modems can't communicate at all with each other.

Thanks, that is what I wanted to make sure. I do not see any switchable upper/lower mark/space frequencies in the circuits, or they are partially switched.

- - - Updated - - -

1. They can work full duplex but you have to manually configure them so one uses the low frequency pair and the other uses the high frequency pair. At 300 Bauds it is possible to fit TX and RX pairs 'one above the other' within telephone bandwidth. There is no automatic configuration method that ensures they work properly and obviously if they are not set to opposite band their signals collide.

2. Yes, but note that acoustic modems basically became redundant around the late 1970s. Having used them I can confirm they were horrible, even someone sneezing nearby could break the data link!

3. They can talk directly to each other if correctly configured. A server isn't required. Where it becomes difficult is the telephone line itself, acoustic modems use the telephone handset microphone and earphone so the 'hybrid' in the phone electronics is responsible for splitting the transmit and receive paths. If you just join modems together with a cable, they won't work.

As there is only one pair of wires carrying DC, ringing voltage, transmit audio and receive audio to a telephone they have a method (using transformers and resistors in old style phones) to route incoming audio to the earphone and microphone signals outgoing to the line. Basically the line contains a mix of both but as you know what you are sending, if you subtract it from the line signal, anything left must come from the other end.

Brian.

Brian,
about 1. It can't be configured on RX, only TX can be switched between the two bands. Please confirm that from the schematic, I might be missing something. So two such identical modem models cannot communicate directly through a telco line because the receiver band is not switchable.

About the 3. again this modem RX but also TX bands cannot be switched, so two identical modems of this model should not be able to communicate directly through them. Please check this out. I have a suspicion taet these modems were good only for communicating with a server side modem (another model modem in a school for example), which of course could not make calls to peoples houses, so band switching was not needed.

About your last paragraph, I think that the telco dies not do something with the voice information. At least in analog lines, it dedicates the channel to you and it does not know information about information direction and it is in the headset/modem job to distinguish them.

 

[betwixt]

As long as the two ends use different frequencies it will work and full duplex can be used - I used to do it all the time. That particular design may not be as configurable as it could be but the theory still holds true.

I think that the telco dies not do something with the voice information. At least in analog lines, it dedicates the channel to you and it does not know information about information direction and it is in the headset/modem job to distinguish them.

It isn't the telco, it's the telephone itself that does it. Remember these modems do not have a line interface at all, they couple the audio tones to/from a normal telephone handset siting in a holder with the modem loudspeaker and microphone acting like your ears in a voice call. The circuit to split the two paths is inside the telephone, not at the telco switch center.

Brian.

 

[neazoi]

As long as the two ends use different frequencies it will work and full duplex can be used - I used to do it all the time. That particular design may not be as configurable as it could be but the theory still holds true.

No I do not refer to the full duplex. I refer to the fact that two such modems are able to communicate with each other in full duplex. From my understanding, modem A, the client, needs to transmit at low tones pair and receive at high. On the other side, modem B, the server, needs to to transmit at high tones pair and receive at low. This way full duplex comms are feasible.
BUT, if modem B now calls modem A, the tones must then be reversed. It is this last feature that I refer to when I asked about the pennywhistle and the other modem. I just needed to verify this from the circuits point of view, that such a feature is not supported in these modems.

It isn't the telco, it's the telephone itself that does it. Remember these modems do not have a line interface at all, they couple the audio tones to/from a normal telephone handset siting in a holder with the modem loudspeaker and microphone acting like your ears in a voice call. The circuit to split the two paths is inside the telephone, not at the telco switch center.

I see. But does it really matter? In fact it may be for advantage. The local phone speaker receives only the other party voice and the local phone microphone, transmits your voice (or the data) only to the other end, without looping back to your local speaker. This is an advantage, isn't it?

I design a small 300 baud bidirectional modem using discrete components, that is why I am asking these things. To somplify it and to make it compatible for every country in the world, I wanted to do it with acoustic coupling. But this raises other interference problems like you said. So my idea was to attach it to an existing phone headset cable. Most phones I have come across nowadays have a detachable head piece (it's cable can be unplugged) so I could connect the modem in series with it. So this would be acousting coupling but, coupled with wire to the phone device instead of the telco line. How do you find this idea?

 

[betwixt]

For reasons of isolation and safety you would have to be very careful. Normal telephones are designed so it is impossible to touch any conductive part because it is quite possible that the line has high voltage on it. The two telephone line wires are both above ground potential and can have as much as 50V DC + 75V ringing voltage across them (~125V peak) so there would be a clear risk of electrocution. It is quite likely that the microphone and earphone do not share a common connection and it almost certainly isn't isolated from the incoming lines. I do not know of any telco that permits direct connection to their lines, direct or indirectly, unless the equipment is certified by them to be safe.

Consider also that some of the microphone signal is deliberately sent to the earphone (sidetone) to make the phone more user friendly. Without it the user tends to shout because they can't hear their own voice!

I'm not sure what you are trying to do with this project, these days 300 Bauds (30 characters per second) is considered extremely slow and modem ICs are available that do all the hard work for you.

Brian.