RS232/V.10 Interface card

Fachforen für Entwickler und Bastler
Benutzeravatar

Topic author
M1ECY
Rank 7
Rank 7
Beiträge: 686
Registriert: Fr 11. Aug 2017, 13:13
Wohnort: Hitchin Hertfordshire, UK
Hauptanschluß: 669089

Re: RS232/V.10 Interface card

#21

Beitrag: # 36559Beitrag M1ECY »

Somewhat delayed, just getting back to this.

I have some MAX232 modules here now -If I understand Fred correctly, the MAX 232 can effectively replace the Opto for the TX and RX path?

Regards
Sean
669089 Siemen G - T100S Online 24H
299709 Antosh G - Creed 444 - Double Current R + D (0800 - 2100) and a bit tempremental
459724 NC
Benutzeravatar

Topic author
M1ECY
Rank 7
Rank 7
Beiträge: 686
Registriert: Fr 11. Aug 2017, 13:13
Wohnort: Hitchin Hertfordshire, UK
Hauptanschluß: 669089

Re: RS232/V.10 Interface card

#22

Beitrag: # 36579Beitrag M1ECY »

Bit of an update from this weekend.

THe MAX232 Module idea ended up with some magic smoke :) I do not know if this is just poor modules from China, or an overenthusiastic R and D engineer.

I changed plans and decided to try to make what I have work.

So, based on an earlier Double Current Card (V1.4) I started some modifications earlier today - the Siemens TGN4 which started this idea off is based on +/-6v signalling, which I finally realise is actually v.28, not v.10.

The I-Telex system for Double Current works using the VF 0, 30 and 60 lines - with the 30 becoming the common connection, and the 60 being +ve, the 0 becoming -ve

Because the power supply works by voltage multiplication the off load voltages are in fact somewhat higher, more like +/-45V

I opted to try to drop these voltages down to suit the machine - this required some modification to the original board, to lift the original supply pins from the backplane connector out of the board, and to install Zener diodes and current limiting resistors to suit - my initial parts holding only had 5.1v diodes in stock, so I tried these, with 1.5k limiting resistors

Bench tests looked promising - with the card out of the I-Telex system and running on bench power supplies, it is possible to see the emitters of the switching transistors operate with an oscilloscope. Helpfully, the Atmel sems to run in some kind of loop when not connected to the I2C bus, so there are cycles of changed polarity.

While the card was on the bench, I also needed to modify the Keyboard connections - with only 6V at a few milliamps to play with, I removed all the 3K3 resistors that drive the sender optocoupler and substituted with just one 430R component - this should then mean about 10ma is available to drive the opto.


Card suitably "Bodged Up", time for a test.

Powered up in I-Telex - No Smoke..... Encouraging....

Connected a multimeter to check if there would beany polarity changes on the receiver output - Yes, switching is happening, good.

Connected the TGN4 (Siemens T1000).....

Yellow Lamp permanently on.... Okay, ignore that for now.... pop machine to online, and type a few keys - Green Light flashes - Cool, so I-Telex and machine are half talking.

Problem is, Yelo Lamp on ,means I-Telex thinks we are already in use.....

Sneak up on I-Telex - Power up with machine disconnected, press button on card,then plug machine in... Aha, Configuration page prints!

Configure for Keyboard dialling - Okay, this works.....

Manage an outgoing call to Henning, and that's where I have got to.

Things to do:

Understand why I-Telex thinks the machine is online, even when it isn't

Massively reduce dissipation - 1ma of current is enough - I had guessed 10ma

More to do - suggestions welcomed.

Cheers
Sean
Folgende Benutzer bedankten sich beim Autor M1ECY für den Beitrag:
deBart
669089 Siemen G - T100S Online 24H
299709 Antosh G - Creed 444 - Double Current R + D (0800 - 2100) and a bit tempremental
459724 NC
Benutzeravatar

FredSonnenrein
Founder
Founder
Beiträge: 2317
Registriert: Fr 3. Jun 2016, 13:49
Wohnort: Braunschweig
Hauptanschluß: 8579924 hawe d

Re: RS232/V.10 Interface card

#23

Beitrag: # 36581Beitrag FredSonnenrein »

Hi Sean,
M1ECY hat geschrieben: So 26. Mär 2023, 21:54 Yellow Lamp permanently on.... Okay, ignore that for now.... pop machine to online, and type a few keys - Green Light flashes - Cool, so I-Telex and machine are half talking.

Problem is, Yelo Lamp on ,means I-Telex thinks we are already in use.....
The Interface of the i-Telex knows (regarding the interface line from the teleprinter to the i-telex boards) three "states":

Standby Idle: permanent space polarity (yellow LED off)

Connected Idle: permanent mark polarity (yellow LED on)

Connected "writing" (input on teleprinter): mark and space pulses (yellow LED on, green LED flashing)

Best regards,

Fred
Grüße,
Fred Sonnenrein, Braunschweig
i-Telex 952741 (Lo133), 8579924 (T100s), 781272 (T100), 792911 (T68d) oder 531072 (T.typ.72)
Bei besetzt oder gestört bitte 531002 versuchen.
Benutzeravatar

Topic author
M1ECY
Rank 7
Rank 7
Beiträge: 686
Registriert: Fr 11. Aug 2017, 13:13
Wohnort: Hitchin Hertfordshire, UK
Hauptanschluß: 669089

Re: RS232/V.10 Interface card

#24

Beitrag: # 36586Beitrag M1ECY »

Thanks Fred,

This helps greatly, So, I will check the DIP swich settings of the LAT card - I think this can be changed at the machine (hoping)

Regards
Sean
669089 Siemen G - T100S Online 24H
299709 Antosh G - Creed 444 - Double Current R + D (0800 - 2100) and a bit tempremental
459724 NC
Benutzeravatar

Topic author
M1ECY
Rank 7
Rank 7
Beiträge: 686
Registriert: Fr 11. Aug 2017, 13:13
Wohnort: Hitchin Hertfordshire, UK
Hauptanschluß: 669089

Re: RS232/V.10 Interface card

#25

Beitrag: # 36588Beitrag M1ECY »

Some more update news.

So, the resting state polarity cannot be changed. Which is a bit frustrating. I obviously have the machine wired correctly,it prints perfect text on receive, and transmit.

So, I am guessing that for some reason, the resting state for the TGN4 is reversed from the normal situation.

Of course, reversing the connections to the card do not work - it cant be that simple.....
669089 Siemen G - T100S Online 24H
299709 Antosh G - Creed 444 - Double Current R + D (0800 - 2100) and a bit tempremental
459724 NC
Antworten

Zurück zu „Entwickler-Ecke“