Restoring a Lorenz Lo15

For service and support with i-Telex, its components and teleprinters in general

Moderator: M1ECY


Topic author
MattisLind
Rank 2
Rank 2
Beiträge: 18
Registriert: So 3. Mär 2019, 22:35
Hauptanschluß:

Re: Restoring a Lorenz Lo15

#21

Beitrag von MattisLind » So 16. Jun 2019, 22:09

The PWM frequency for the step-up is fixed to 40 kHz. Set by the micro controller. The current source is using hysteresis switching. There is a high speed comparator comparing the voltage over a measurement resistor with a reference voltage representing the current set from the micro controller ( a lowpass filtered PWM signal ). There is high speed MOSFET reacting on the output of the comparator, switching on and off the HV generated by the step-up feeding it through a 100 mH coil. The current in the coil will rise and when it reaches 42 mA the comparator will switch off and theis will cause the coil to slowly unload its energy into the magnet and when the current has gone down to 38 mA it will turn on again. Thus the timing will depend on the input voltage and the resistance of the magnet.

An update to the carriage feed problem. It only happens if I type into the region where the printing carriage cannot go further. I.e I f I try to print 80 characters the last 10 or so will not be printed. The carriage is in the rightmost position. And there will be nor more motion to the right. However it seems that the same condition persists when the carriage return has been executed. So then it does nothing in the leftmost position.

If I only type 60 characters per line, sending the CRLF before reaching the rightmost position it will go on printing perfectly. So there is some mechanism that stops the carriage from advancing that is not reset when there is a carriage return.


Topic author
MattisLind
Rank 2
Rank 2
Beiträge: 18
Registriert: So 3. Mär 2019, 22:35
Hauptanschluß:

Re: Restoring a Lorenz Lo15

#22

Beitrag von MattisLind » Mi 19. Jun 2019, 22:09

The problem with carriage spacing after entering into the end zone is till not resolved. But I am convinced that it has to do with the spacing stop sleeve and the associated clutch. I am not sure if the clutch slips somehow or f the spring force is not enough. Need more research.

But I got the ribbons that I ordered and hooked up the current source and looped the transmitter and receiver so I could type and se the result directly. As far as I can see it types everything just fine so it decodes everything ok. But when typing letters the ribbon wanders slowly upwards so that it eventually jumps off!

https://i.imgur.com/BydczK8.mp4

Anyone has an idea why this happen. It works fine in figures mode. But not when typing letters.

Well. Took a second look now. It was quite obvious. I just threaded the ribbon wrong...


Topic author
MattisLind
Rank 2
Rank 2
Beiträge: 18
Registriert: So 3. Mär 2019, 22:35
Hauptanschluß:

Re: Restoring a Lorenz Lo15

#23

Beitrag von MattisLind » Do 20. Jun 2019, 11:14

Researched the spacing problem a bit more.

Here is when the spacing stop lever is not yet engaged and it is spacing normally.

Bild

And here is when it is engaged and spacing is stopped:

Bild

When often when pressing CR (but not always) it will remain in the latter position and is not going back into the first position. I tried putting some oil on the lever but it didn't fix it.
However I recognised that it seems that a seconds CR always clears up the situation. Not sure why that would clear it, but it does.

Manually pushing it with a screwdriver also resolves the situation.

Then I found another oddity: If it receives a BELL once, it will keep hammering on the BELL regardless of character received. This is cleared by a CR as well.

So for currents and voltages:

This is marking state measured over the 51 ohm measurement resistor:

Bild

2V is equal to around 40mA. Switching frequency at 90V is around 15 kHz as it seems.

This is on the keying transistor, loop returning from the magnet, right now the line voltage is set to approximately 90V:

Bild

The selector magnet creates really nasty spikes. Killed the first transistor I tried. A 300 V SOT-23 PNP is used now together with a MOV. Considering use of a TVS instead.


Topic author
MattisLind
Rank 2
Rank 2
Beiträge: 18
Registriert: So 3. Mär 2019, 22:35
Hauptanschluß:

Re: Restoring a Lorenz Lo15

#24

Beitrag von MattisLind » Do 20. Jun 2019, 17:45

One more thing. Is it normal that the Lo15 is not able to make the CR before typing next character? Now it types a character on its way back to the home position. Is it necessary to insert a NUL character in the stream to make it work properly or is it just my Lo15 that is slow in doing the CR?

Benutzeravatar

DF3OE
Founder
Founder
Beiträge: 1085
Registriert: Di 7. Jun 2016, 09:45
Wohnort: Edemissen- Blumenhagen
Hauptanschluß: 925302 treu d
Kontaktdaten:

Re: Restoring a Lorenz Lo15

#25

Beitrag von DF3OE » Di 25. Jun 2019, 10:40

Mattis, fortunately I could get hold of an English manual for the Lo15 (a), the version that you have.
We will be scanning it as soon as possible and put it online here.
mfg
henning +++

925302 treu d - T1000Z (Hauptanschluss)
55571 fvler a - T100S
218308 test d - T1000S/LS (Werkstatt)
733760 egon dk - GNT5 (Werkstatt)
92612a treu d - T1200SD (Listenabruf KLKL)

Benutzeravatar

FredSonnenrein
Developer
Developer
Beiträge: 989
Registriert: Fr 3. Jun 2016, 13:49
Wohnort: Braunschweig
Hauptanschluß: 8579924 hawe d

Re: Restoring a Lorenz Lo15

#26

Beitrag von FredSonnenrein » Sa 29. Jun 2019, 10:19

MattisLind hat geschrieben:
Do 20. Jun 2019, 17:45
One more thing. Is it normal that the Lo15 is not able to make the CR before typing next character? Now it types a character on its way back to the home position. Is it necessary to insert a NUL character in the stream to make it work properly or is it just my Lo15 that is slow in doing the CR?
This is a regular effect.
Because the CR is usually followed by a LF the first actual character will be printed at the first position.
Folgende Benutzer bedankten sich beim Autor FredSonnenrein für den Beitrag:
sr6346
Grüße,
Fred Sonnenrein, Braunschweig
i-Telex 8579924 (T100s) oder 781272 (T100) oder 792911 (T68d) oder 531075 (Creed75)


Topic author
MattisLind
Rank 2
Rank 2
Beiträge: 18
Registriert: So 3. Mär 2019, 22:35
Hauptanschluß:

Re: Restoring a Lorenz Lo15

#27

Beitrag von MattisLind » Sa 6. Jul 2019, 18:28

FredSonnenrein hat geschrieben:
Sa 29. Jun 2019, 10:19
This is a regular effect.
Because the CR is usually followed by a LF the first actual character will be printed at the first position.
Yes. But I have a CR followed by a LF. I got some replies on the GreenKeys list that it seems like the standard operating procedure was to do a CR CR LF. But I don't know really.
I have been thinking about how to tighten the return spring so the return would be quicker. Possibly overcoming the need for the second CR.
DF3OE hat geschrieben:
Di 25. Jun 2019, 10:40
Mattis, fortunately I could get hold of an English manual for the Lo15 (a), the version that you have.
We will be scanning it as soon as possible and put it online here.
That sounds great! BTW. Did you have a spare reperferator on/off knob available that I could purchase? It would be so nice to have it complete. The other parts (springs and brushes), I took from the other Lo15 that I have. It is now considered as a spare parts machine.
DF3OE hat geschrieben:
Do 25. Apr 2019, 10:58
You will need 13mm black Nylon ribbons (Gruppe 28 in Germany) with lugs at the end to activate reverse switch.
Something like this:
I ordered precisely those do get the revering effect. Unfortunately they delivered the variant without reversing... Hopefully they send me new ones.

In the mean time I found out that Pelikan 571935 should fit very well. And it has the reversing switch. And it is black. I do prefer black, really.

Still haven't found the root cause for the BEL that keeps belling until CR and also not found the root cause for carriage advancing not reset after going into the stop zone. On the other hand the double CR solves both problems!

The governor switch is causing arching which creates ozone. When it does it seems like the speed is increasing a bit. What is the proper way to prevent the arching in the switch. I have cleaned it. Or is it normal behavior ?

I have done some work on the software and now it is able to communicate with my Linux laptop. The STM32 uC is doing baudot to ASCII conversion.

Here is a link to a to short film clip when doing an ls command on my Linux laptop.

Lorenz Lo15 login to Linux laptop

Antworten

Zurück zu „Service and Support“