Interfacing old PC to a modern PC monitor

[valvekits]

My TV theory is stuck in the seventies with negative line sync pulses, back porches and swinging burst and I’m afraid all PC monitors and video formats have simply passed over my head in the intervening years!. However, a friend of mine would like to replace a defunct monochrome CRT monitor connected to a 1980’s vintage (US) industrial PC. No schematics available but from scoping signals, the PC provides Horizontal Sync, Vertical Sync and Video and they all look to be at TTL type levels. Line frequency seems to be around 15kHz. I’m left scratching my head as to if and how these signals can be displayed on a modern PC monitor. I very much doubt it would be useful just to combine these signals together at TTL levels to create a composite waveform, so any suggestions are more than welcome.

Eddie

[paulsherwin]

If it is PC compatible hardware and has an ISA (AT) bus, you should be able to fit a VGA card and plug in a standard VGA monitor.

Paul

[Michael Maurice]

I once had this, the problem was that the VGA socket on the computer was missing a hole so that a modern VGA lead wouldn't fit.

My solution was to carefully drill a hole in the socket using a 0.8mm PCB drill bit.

[julie_m]

If it's a 9-pin connector, then it will be CGA. There are 4 signals: red, green, blue and "bright", and separate syncs, and the levels are TTL. The line sync is close enough to TV rate that you can get away with feeding it into a TV set via the SCART socket.

You will have to AND together red and bright, blue and bright and green and bright, and resistively mix these with the straight red, green and blue lines respectively to get the analogue RGB signals a TV requires (you want 0.7V into 75Ω for full brightness), and you will have to combine the syncs. When a SCART connector is used for RGB, you need an open collector transistor to present the same negative-going sync on the picture input as a real picture signal would produce. (You can resistively combine the R', G' and B' to give you a mono picture signal, for the benefit of TVs with only a partially-wired SCART socket.)

74HC TTL ICs should be well up to the job. Good luck with it!

[cmjones01]

If it's monochrome and TTL, there's every chance it could be an MDA or Hercules video signal:

http://en.wikipedia.org/wiki/IBM_Mon...isplay_Adapter

These run with a horizontal sync frequency of 18.432kHz, which is generally too high for TV-type monitors to be tweaked up to and too low for modern PC monitors.

It's worth checking more carefully with the scope what the horizontal frequency is. If it is 18.432kHz then it might be easiest just to repair the original monitor! Otherwise it's a case of getting a modern equivalent, which is likely to be expensive since they're only used in industrial applications now, or finding a second-hand one somewhere. I'm sure there's a surplus dealer somewhere with a mountain of them. It's just a question of finding them!

Chris

[valvekits]

Thanks for the info so far chaps, here is a picture of the monitor chassis, the socket for the scan coils has melted, a few components blown and the reverse side of the board is badly charred. The CRT has lots of burnt in graphics so it seems like a good opportunity to see what modern technology has to offer. The computer might just pre-date the IBM PC as we know it. The edge connector on the lower left is where the signals come in along with the power. I used a digital scope to measure the line sync pulses and whilst doing this I recall the scope display was helpfully informing me that the frequency was 15.xx kHz, whilst the sync pulses measured around 6 to 7uS so this rules out the EGA possibility.
So far, it looks like I could use the existing signals to drive a CGA monitor by matrixing the video onto RGB and tying the Intensity line high.
I read that CGA is not compatible with VGA so it looks like I can only upgrade to CGA unless anyone has any other ideas.

Eddie

[McMurdo]

Is this off a machine tool? They tend to use their own standards, particularly regarding the polarity of the TTL syncs.

This is known in the trade as a slow-scan monitor, a colour PC monitor line frequency wont go that low so rule it out; they're also analogue whereas this is digital. You need a TTL video with seperate syncs configured with corect polarity (though im sure you could make a little adapter board if the syncs are wrong). Trying to get it to work on a TV wont work, the phosphors on a colour domestic CRT arent high-enough res for TTL text and will be blurred...and burnt-in within weeks!

Replacement depends on CRT size and if it has to fit inside a console or special cabinet. You can still get colour slowscan CRT TTL monitors in certain sizes (14 and 15 being popular) but for monochrome your best bet is an LCD panel with multisync adapter.

These are available from Trident displays amongst others.

http://www.trident-uk.co.uk/jkcm/Products/Monitors

If you dont mind dealing with a compnay in Europe then these may be cheaper, we've used them and they are great quality. They're called Iris Display.

http://www.industrial-monitors.com/m...eplacement.htm

[valvekits]

I used the info from Kevin's post to trawl the net and it seems that there is a converter scaler that converts "Legacy" sources such as older RGB to VGA. Any comments on these type of converters?
http://www.ambery.com/rgbcgatovgac.html

Eddie

[McMurdo]

Does it say how much they are? The ones I've seen are all dearer than a new monitor.

That one in your link is CGA RGB which is 4-bit colour. (red/green/blue/intensity). I guess you could modify it to run in mono, I dont know how the box would recognise non-standard TTL syncs. Some do automatically, but others dont. Are the HV syncs on your machine CGA compatible? Just being TTL isnt enough, they have to be the correct polarity; all combinations are used!
What is it off, exactly? I deal with alot of industrial CNC equipment and might recognise the control system.

[Dave Moll]

Originally Posted by McMurdo

Does it say how much they are?

Yes it does:

$92.00

[llama]

Originally Posted by valvekits

...the scope display was helpfully informing me that the frequency was 15.xx kHz, whilst the sync pulses measured around 6 to 7uS so this rules out the EGA possibility.

Eddie

I'm pretty sure that EGA monitors are "multi-sync" and so include CGA in their capability. Widens your choice.
Graham

[valvekits]

Kevin, I will try to get some more info this week; the computer controls some kind of a vacuum chamber and the monitor displays recipes (I think). I am sure it's possible to repair the existing scenario but when I looked at the burnt in monitor I got the idea that it might be "easier" to upgrade it! - From what I have been reading on the CGA requirements for the sync pulses it looks as though timing and polarity might be okay, so that might be the first course of action.

[cmjones01]

Originally Posted by llama

I'm pretty sure that EGA monitors are "multi-sync" and so include CGA in their capability. Widens your choice.

Yes, EGA monitors also work at CGA line frequency. I'm still using an EGA monitor on my elderly Tektronix 3002 logic analyzer. I think the line frequency changeover is controlled by the line sync pulse polarity, but I might have misremembered that.

EGA monitors are, however, much rarer than CGA-compatible ones in my experience. A handful of resistors and diodes can usually persuade even a modern SCART-equipped flat panel TV to display a CGA-like TTL video signal. Also suitable are the monitors popular with home computers in the late 80s: the Philips CM8833, Commodore 1084S, and even the indestructible Microvitec Cub will do the job nicely. These monitors seem common enough to find in decent condition.

All of these solutions require some electronic hacking, though, which may or may not be acceptable in this application.

If valvekits could provide a sketch of the observed sync waveforms (especially their polarities), I'd be happy to propose a simple interface circuit which should allow a SCART-equipped TV to display the signal.

Chris

[valvekits]

Here is an update: My friend managed to repair the monitor by replacing the burnt out components and tin copper wire for the tracks. It turns out that replacement monitors are still available as open frame monitors with various options regarding sync polarity and video etc so there doesn't seem to be much point in trying to re-invent the wheel. However, he is planning to bring the signals out on a separate connector so we can experiment in the future - you never know we might end up with a separate thread in success stories. Thanks for all your contributions to this thread.
Cheers
Eddie

[cmjones01]

That's good news. I'm glad the old monitor was repairable. It's also useful to know that open-frame monochrome monitors still exist: they must be a dying breed in themselves.