BELL 202 Modem - Working issue

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Hello, I have made BELL202 modem using TCM3105 IC.
The circuit output part is as per IC datasheet using Simple telephone interface. My problem is the two boards connected directly RJ11 to RJ11 of other board it does communicates.
Single board when checked it communicates when 600Ω is put at secondary of transformer.
Is this right way of testing?
Kindly help :confused:
 

Attachments

KrisBlueNZ

Joined Oct 17, 2012
111
OMG is the TCM3105 still available?!
I'm not sure about the design of your line interface, but it does at least have a 600 ohm impedance. To test the transmitter without another board, run it into a 600 ohm resistor. Otherwise, connect two boards together directly. Obviously one board's transmit and receive frequencies must be set the opposite of the other board.

You said "it does communicates". Do you mean "it doesn't communicate"? If so, start by looking at the line signal with an oscilloscope to see whether the right frequencies are present and at the right amplitude.

If you tidy up your schematic and mark all the component values clearly, I may be able to tell you whether your line interface is properly designed.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Yes, when two boards connected one to another it doesn't communicates.
they works independently when 600ohm connected at secondary of transformer.
Can u just explain how the output part works?
its frequency n waveform how will it look like?

I have simplified the circuit. Please look into.
Thanks. :)
 

Attachments

Last edited:

KrisBlueNZ

Joined Oct 17, 2012
111
Thanks for tidying up the circuit. It's a lot easier to read now.

The op-amps you've marked U2, U3 and U4 have their inverting (-) and non-inverting (+) inputs exchanged.

I would test each board independently first. Connect a 600 ohm resistor across the secondary of T1 and connect an oscilloscope across that resistor. You should see a clean sinewave, whose frequency changes as you switch the TXD line from low to high and back.

These modems use two sets of frequencies. Each modem is configured as either "answer" or "originate". When you connect two of them together, one must be set to "answer" and the other must be set to "originate". This ensures that the second modem receives using the frequencies that the first modem is transmitting, and vice versa. If both modems are set the same, they will both transmit on the same frequencies, and they won't receive anything.

The line interface drives both sides of the transformer with complementary transmit audio signals. The receiver is a differential amplifier that responds to the voltage across the transformer, and includes a subtractor that is supposed to subtract the transmitted signal (which comes in via R8). When T1 is driving into the correct impedance (600 ohms), ideally none of the transmitted signal should appear at the received audio input to the modem - the only signal that should be present at RXA is signal that is coming from the other board. In practice the transmit signal rejection will never be perfect but it should be at least 15 dB if the loading is right.

If this doesn't help you solve your problem, first tell me the result of your tests on a single board running into a 600 ohm resistor, looking at the signal with an oscilloscope.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
The opamp U2, U3 and U4 are connected as per the typical diagram shown in datasheet.

Testing individually - At transformer's secondary connecting oscilloscope with 600ohms doesn't shows pure sine wave.

I'm using hyperterminal to transmit & receive using 1200 baud, 8 bits, parity none, stop bit 1, flow-control none , ANSIW. are they proper?
i have read delays is required RTS/CTS delay ? DCD ? pin connection from PC?do we really need handshaking signals?

I have attached some images of output across T1.
Second image is jst expanded view.
third is hyperterminal which recieves the data i transmit (loopback)
and one while in trans-receiving with output at TXA pin of IC and across 600ohms T1 secondary.
but the two boards connected doesn't works.

Thanks.
 

Attachments

Last edited:

KrisBlueNZ

Joined Oct 17, 2012
111
The opamp U2, U3 and U4 are connected as per the typical diagram shown in datasheet.
So they are! The data sheet is wrong too. In Figure 3, "Telephone line interface circuit", the op-amps marked U1A, U1B and U1C all have their inverting and non-inverting inputs exchanged. U1D is right.
Testing individually - At transformer's secondary connecting oscilloscope with 600ohms doesn't shows pure sine wave.
Probably because of those errors. Actually I'm surprised you see anything at all. But you need to fix those errors first.
I'm using hyperterminal to transmit & receive using 1200 baud, 8 bits, parity none, stop bit 1, flow-control none , ANSIW. are they proper?
Yes that's OK. Although Hyperterminal is a horrible piece of software and I would use ANYTHING else in preference to it.
i have read delays is required RTS/CTS delay ? DCD ? pin connection from PC?do we really need handshaking signals?
No, not in a dumb modem application where you're transmitting and receiving constantly.

It's probably a good idea for your receiver to monitor the DCD output from the TCM3105 so that if the transmitter or the signal path fails, you will see an indication of this, rather than just no data or garbage data.

morethegr8 and I have talked via Skype. His application is a receive-only circuit for use with a legacy device (an "axle counter") that transmits Bell 202. I've sent him a partial diagram for simplified receiver that doesn't use any op-amps and can't transmit, which he can use with the axle counter. He can use the diagram he currently has, once he's fixed the op-amp connections, as a transmitter to test with.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Hey, i'm extremely sorry.... i m busy with some other project. i will get bck to BELL202 modem in some time .. !! :)

Regards,
Amit More.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Hey Kris,
I have changed the circuit as per your instructions, it works well now.

But only one query i have... when single modem is tested with no secondary 600 ohms termination, it reflects backs whatever is transmitted.
this should not happen right?

The transformer we are using is OEP 1200.
Datasheet attached. Is this the right one we are using?
 

Attachments

KrisBlueNZ

Joined Oct 17, 2012
111
I'm not sure what you're asking.

That transformer looks good.

Are you talking about the transmitter circuit that comes from the modem data sheet? You need to explain your question more clearly.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Testing modem independently. Not connecting 600 ohms at secondary. my modem is receiving what is transmitted by itself. This should not happen right?

Else, when two boards connected they are working fine.
 

KrisBlueNZ

Joined Oct 17, 2012
111
Which modem, the transmitting one or the receiving one?

The receiving modem should only be receiving what the transmitting modem is transmitting. The receiving modem's TXA output is not used.

The transmitting modem will receive whatever it transmits, no matter what is connected to its line output - nothing, a 600 ohm load, or the receiver. That shouldn't matter, because you don't care what the transmitting modem receives, right? The data flow is just from the transmitting modem to the receiving one; you don't need data to flow the other way as well, right?

It might be better to Skype me. I'm not available all the time so keep trying.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Ya,
Then these boards are working fine.

For all i post the working schematics with Waveforms.

Thanks Kris for all your support.:)
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Hello kris,

These modems are very sensitive to adjustment of pot R9.
As i need to adjust them to get proper data at hyperterminal.
Bcoz sometime i find garbage value at hyperterminal.

and if i change it to V.23 mode by setting TRS, TXR1 and TXR2 to low. do i need anything else to change?

what should i do to eliminate that garbage value sometimes being received?
 

Attachments

KrisBlueNZ

Joined Oct 17, 2012
111
The circuit you posted is the circuit you're using for the transmitter, right?
You're using my very simple circuit for the receiver, right?
R9 affects the transmitted audio level of that circuit. You're not using that circuit as a receiver.
The TCM3105 is not fussy about its received audio level. There is a nominal level specified, but you can go at least 6 dB higher or lower and the modem will still receive reliably.
The receive bias input, pin 7 on the receiving modem, can make a lot of difference to the received data accuracy. I suggest transmitting a 1-0-1-0-... stream and adjusting the voltage on pin 7 of the receiver to get exactly half the supply voltage on the RXD output; this means that the duty cycle of the received data is exactly 50%. You may want to vary the signal level at the receiver to make sure that the RXD pin stays at 50%. The adjustment may be more critical at lower received signal levels.
 

Thread Starter

morethegr8

Joined Feb 14, 2011
31
Hello Kris,

Modems are connected in V.23 mode.
I have connected 4 modems. The configuration details is in attached doc.

I have problem that this config gives junk value sometime at output while 2 modems connected works fine.

If i lower the baud to 300 it works with no garbage value at output.

Is this baud related issue? At 1200 it works but sometimes garbage value comes in between.
As per IC specs it has to work at 1200baud.

Kindly help.
 

Attachments

KrisBlueNZ

Joined Oct 17, 2012
111
I can't tell anything from the information you've given me.

What is the significance of having two sets of modems? What is the "cross cable" at the bottom of the diagram? What are you doing with the data into and out of the modems?

I need a complete description of your test system, and the problem you have.
 
Top