1974's dumb terminal CRT problems.

[ChristofferB]

Hey!

I bought an old dumb terminal today. pretty good condition, although, upon powerup: nothing on the screen.
I didn't have it hooked up to anything like a computer, but it should probably still show something on the monitor; a prompt maybe.

It's a British Cifer PS502 terminal.
I think there's an issue in the monitor section of it; there's a snapping sound from the monitor board, maybe the HV area.
Has anyone encountered this issue with CRT's before? Is it fixable?
Thanks in advance,
--Christoffer

[SeanB]

EHT tripping, probably from all the dust around the EHT connection on the CRT itself. Easy to clean with a DRY ( and I mean dry, toast it in the over at 120C/240F for a half hour before using it so it is really dry) cloth, used to wipe off the glass after discharging the CRT. You take a length of stranded wire, put a crocodile clip on the one end, and strip the other end bare for 2 in ches. Then clip the croc clip on the metal band around the CRT ( there is a metal band, probably with a silver braided copper strap going from each screw to the others, with a black wire leading to the CRT base panel and another to the main board) and wind the bare wire around the metal stem of a plastic handled flat blade screwdrive. Then slide the end under the silicone cap on the red wire attached to the side of the CRT. There will be a crack and a slight flash, and the tube will be discharged.

Then use the dry cloth to wipe around the connector, the connector itself and the red wire.

[ChristofferB]

Thanks! I'll try that! But could that cause the screen to not display anything? - Could also be because of lack of serial connection.

--Christoffer

[mazurov]

Thanks! I'll try that! But could that cause the screen to not display anything? - Could also be because of lack of serial connection.

--Christoffer

You should be able  see the [blinking] cursor with just power applied. Shut off the room lights to check  that CRT brightness is sufficient.

[Towger]

Should show a flashing cursor.

[ChristofferB]

Huh. So; there's definately no flashing cursor, but that's not so suprising, as the voltage rails on the TTL logic cards measure 1,3V. Probably should have a word with my power supply. The snapping of HV seems to have ceased, so i guess that just needed a bit acclimation.

--Christoffer

[Towger]

From memory, last time I used a real terminal in the 80s.

[tautech]

EHT tripping, probably from all the dust around the EHT connection on the CRT itself. Easy to clean with a DRY ( and I mean dry, toast it in the over at 120C/240F for a half hour before using it so it is really dry) cloth, used to wipe off the glass after discharging the CRT. You take a length of stranded wire, put a crocodile clip on the one end, and strip the other end bare for 2 in ches. Then clip the croc clip on the metal band around the CRT ( there is a metal band, probably with a silver braided copper strap going from each screw to the others, with a black wire leading to the CRT base panel and another to the main board) and wind the bare wire around the metal stem of a plastic handled flat blade screwdrive. Then slide the end under the silicone cap on the red wire attached to the side of the CRT. There will be a crack and a slight flash, and the tube will be discharged.

Then use the dry cloth to wipe around the connector, the connector itself and the red wire.

+1
Be carefull these things bite.
If theres a clicking sound EHT is tracking somewhere.

[TerraHertz]

There will be a crack and a slight flash, and the tube will be discharged.

Slight additional to that. The first discharge will happen before the screwdriver end actually touches the EHT terminal. You should make sure the tip actually touches the metal, and hold there for a while, to be CERTAIN the EHT capacitances are down to zero volts. Personally I like to leave the shorting strap in place for at least a minute, as I've encountered CRTs that 'rebounded' voltage somewhat after being briefly shorted to ground. Presumably leakage inside the tube from uncoated glass surfaces that retained some charge.

Then use the dry cloth to wipe around the connector, the connector itself and the red wire.

Or, if there's visible dust and grime, take off the EHT cap, and clean everything with a rag soaked in turps and/or household cleaners. Once it's actually clean, then worry about drying off any moisture.  Grime absorbs moisture from the air, and that intermittent HV snapping sound is symptomatic of dirty surfaces intermittently forming conductive paths. Each current pulse blows that path clear, and it takes a while for a new one to form. The snapping may stop after a while, as the gear warms up and the balance between air humidity and surface absorption shifts.

When you are ready to replace the now clean EHT cap on the clean CRT glass, coat the glass and cap contact surfaces with some silicone oil. Or Vaseline/white petroleum jelly works too. Stops electrostatic surface creepage.

Edit: Another thing to check is whether you can see filament glow at the back of the electron gun.
Also whether the neck of the tube has some dark-silvery metalized areas inside the glass. If those have turned white, the tube is no longer under vacuum.

[ChristofferB]

The snapping may stop after a while, as the gear warms up and the balance between air humidity and surface absorption shifts.

Indeed, after leaving it on while checking other things (again, no power on the digital stuff, think the problem's there) the snapping has disappeared. There is glowing from the neck board area (the tube, not the board). I'm sure it'd be a good idea to clean the CRT regardless, but it might not be first priority now.

Thanks!

--Christoffer

[atferrari]

Huh. So; there's definately no flashing cursor, but that's not so suprising, as the voltage rails on the TTL logic cards measure 1,3V. Probably should have a word with my power supply. The snapping of HV seems to have ceased, so i guess that just needed a bit acclimation.

--Christoffer

Or maybe, it rendered its soul forever... 

[cosmicray]

there's definately no flashing cursor,

In addition to all the other remarks, some of those old dumb terminals had dip-switches to configure baud-rate, cursor flash, and other settings (because pre-NV ram era).

[ChristofferB]

there's definately no flashing cursor,

In addition to all the other remarks, some of those old dumb terminals had dip-switches to configure baud-rate, cursor flash, and other settings (because pre-NV ram era).

Well I found that strange as well. It doesn't. It has a small DIP switch, 3 poled, but nothing accessible without unscrewing the cabinet. Some settings, like scroll/static and standby mode are available on the front, but nothing related to cursors or baudrate.

I'm posting some pictures of the digital cards, they're rather nice.


This board, I think is the CRT driver and keyboard interface, keyboard input being one of the ribbon cables.


This is the main logic board. No UART to be found, so i guess it's all 74-series logic-
On a seperate note; deciphering 74-series logic datecoded 1974 is a bitch!
-The small DIP board top left seems to house PROM's, and some decoding logic, so that could be the main system ROM. Can't find any EPROM's on there.

[SeanB]

The 2 UARTs are there, in 28 pin packages near the DB25 sockets. Nice that all IC's are in sockets. good but bad for intermittent and crazy faults. With a board that age and with the 5V rail being 1V3, Look at the power supply and see if there is a LM350k getting hot on a heatsink, or a discrete power supply using a LM723 regulator with some pass transistors also getting hot. If so suspect those red dipped dry tantalum capacitors, they are the right series and the right age to be low resistance shorted. The faulty one will be very hot after 20 seconds.

[ChristofferB]

The power supply issue was a dead LM3055, that's fixed and the 5V rail reads as it should. Still nothing on the CRT.
A missing promt character or.. anything.. I guess could be both a CRT problem, or the logic. Should i for starters reseat all IC's?

The 2 UARTs are there, in 28 pin packages near the DB25 sockets

Nope, I thought that too; they're 74154's - Carefully having their number blacked out by black marker, of course .

[SeanB]

Time to reseat all IC's then, it should cure a lot of issues. Otherwise a crystal is not oscillating.

[ChristofferB]

So, after reseating all IC's on the logic cards (Some of them had pretty tarnished legs - a few missed some!), the previous fault resurfaced:
Low voltage on the 5V rail. Either some other problem in the PSU is killing the power transistor for that rail, or the logic boards are shorted.
They read about 130 ohm, but I'm not sure if that's low enough to consider it a short. It still would only draw about 0.2A at 5V...
Would it be worthwhile to just go ahead and change all electrolytes on the psu board (and tantalum's on the logic boards) ? Or is the fault probably elsewhere?

Also: The +-12V supply is fine, and that drives the buzzer/siren thing, so whenever the 5V rail is off: it howls like crazy, far too loud to concentrate. Would it damage anything to disconnect the 12V rail, even though some logic /proms might need it?

Update: Removing the logic boards, the 5V rail reads normal- so it's probably a short or atleast heavy load on the digital boards. Looking at the tantalums now, not sure wether regular ceramic decoupling caps fail like that too. - on the 5V rail, 16V tantalums are used, could one replace those safely with 10V's ? They're only decoupling so one might start to try the board without them.