M24SP DB25 to DB9 adapter

[Valerio]

Such adapters exist and there is no need to build something. They are called "RGB to VGA converter".

I don't think I've ever seen a ready-made one that would work in this case? I've seen simple ones that leave the RGB lines untouched and just attempt to reconstruct the VGA hsync/vsync, or complex ones that take digital RGB and do a full-frame conversion - neither are needed here as the 6300/M24 sync signals are fine with the VGA monitor I have (NEC 1990FXp, for the record).

 

[Valerio]

I don't work well with verbal descriptions--how about a schematic? You might also look to see what the IBM 5153 CGA monitor does, since it has to deal with the same issue.

Here it is:



"C" stands for any of the colour lines output by the 6300/M24 (R, G, or B). If we call V_H the voltage at C_OUT when intensity is high (ie, on) and V_L the voltage at C_OUT when intensity is low (ie, off), then with the circuit above we have V_H / V_L = 1 + R1/R2. Let's say we want V_H to be twice V_L, so this would mean R1=R2 (let's call this common value just "R"). In this simplified case, V_H = V_ref * R3 / (0.5*R + R3), where V_ref is the voltage corresponding to logic 1 (let's say 5V).

Then let's say we want V_H = V_ref / n (ie we want V_H n times smaller than V_ref) then this means R3 must be equal to R / (2*(n - 1)).

We have one more free variable - R itself - which I guess we could set by trying to maximise the 'HCT08 switch speed (there's something in the datasheet about optimal load) or by aiming for the VGA-standard output impedance (? 75 Ohm - not sure this makes sense here).

All of this if I did not make any mistake in my math!

 

[Chuck(G)]

Maybe, but I wonder about the combination of resistance+output capacitance of the HCT device.

Maybe the more "brain dead" approach might be to use a cheap video DAC, such as the ADV7120, which would allow "tuning" of each color. You might even be able to work out a good scheme for "brown" (low intensity yellow).

 

[Valerio]

Maybe, but I wonder about the combination of resistance+output capacitance of the HCT device.

Maybe the more "brain dead" approach might be to use a cheap video DAC, such as the ADV7120, which would allow "tuning" of each color. You might even be able to work out a good scheme for "brown" (low intensity yellow).

Thanks - I had a look, looks like the ADV7120 needs a pixel clock input, which is not available on the external connector (unsurprisingly). Unfortunately it's not even present on the expansion board connector on the video card

Anyway HCT08s are cheap - I might give it a try next time I place a Mouser order...

 

[Chuck(G)]

You might also consider using JFETs as switches--they're routinely used at RF frequencies. GaAs varieties even faster--GaAs MESFETs have t[sub]pd[/sub] in the picosecond range (if you want to get ridiculous).

Actually, you shouldn't have to worry too much about prop delay--IIRC, in the early days, an opto was used to couple the video signal to hot-chassis CRTs.

But, have a look at this page, which dedicates itself to the "brown" color of IBM 5153 CGA display. Note that the input to the circuit is TTL RGBI, the same as the M24--and I suspect that full-scale voltage to the RGB transistor inputs is likely 0.7V.

https://www.aceinnova.com/en/electronics/cga-and-the-brown-color-in-ibm-5153-color-display/

 

[Valerio]

You might also consider using JFETs as switches--they're routinely used at RF frequencies. GaAs varieties even faster--GaAs MESFETs have t[sub]pd[/sub] in the picosecond range (if you want to get ridiculous).

Actually, you shouldn't have to worry too much about prop delay--IIRC, in the early days, an opto was used to couple the video signal to hot-chassis CRTs.

But, have a look at this page, which dedicates itself to the "brown" color of IBM 5153 CGA display. Note that the input to the circuit is TTL RGBI, the same as the M24--and I suspect that full-scale voltage to the RGB transistor inputs is likely 0.7V.

https://www.aceinnova.com/en/electronics/cga-and-the-brown-color-in-ibm-5153-color-display/

Thanks - very interesting article! I'll try and make sense of the circuit. I do note however that it uses 74S05's, same logic family as the M24 output stage. Incidentally in the mean time I found that a 74AS808BN (6$ each) is virtually perfect for this - hex AND gate with 6ns propagation delay and high (48mA) current drive. May need more than one to implement the true CGA palette though.

 

[lightstar]

Guys, don't forget to connect PIN2 to ground to avoid the underlined attribute for blue colors in color mode.
ftp://ftp.oldskool.org/pub/drivers/ATT/6300/Documentation/Olivetti M24-240 FAQ.html

Let's look at the monitor. If you have a monochrome monitor can be connected to ground on pin 2 on DB25 jack Olivetti video card, so you believe to be attached sheet for a color screen and remove the underscore attribute on some color combinations from place to strange horizontal stripes.

 

[Valerio]

Guys, don't forget to connect PIN2 to ground to avoid the underlined attribute for blue colors in color mode.
ftp://ftp.oldskool.org/pub/drivers/ATT/6300/Documentation/Olivetti M24-240 FAQ.html

Yes that's a good point - pin 2 and pin 10 indicate to the system what type of monitor is connected, and their default status is logic 1 (they are internally pulled up to Vcc with 2.2k resistors), which means monochrome. The only other valid config, according to the Theory of Operations manual, is with pin 2 (MONID0) at 0, indicating a colour monitor.

 

[Valerio]

But, have a look at this page, which dedicates itself to the "brown" color of IBM 5153 CGA display. Note that the input to the circuit is TTL RGBI, the same as the M24--and I suspect that full-scale voltage to the RGB transistor inputs is likely 0.7V.

https://www.aceinnova.com/en/electronics/cga-and-the-brown-color-in-ibm-5153-color-display/

So it seems the circuit uses open-collector inverters and resistor dividers to manage the voltage for each channel. The open collector outputs also make it easy to implement a wired-OR configuration that selects the special treatment for the R=1,G=1,B=0,I=0 ("brown") case. Trixter's website has a nice explanation: http://www.oldskool.org/pc/cgacal

OK, so why the odd name for color 0 1 1 0? Well, we now turn to the schematic of the CGA monitor. The 4 bits come in, and are buffered by 74S05 chips, Q201 and Q253 on the schematic. Now for the interesting bit: There are 4 of the open-collector inverters -- Q201d, Q201f, Q253e and Q253f -- whose outputs are connected together. This common signal will only be high if the input signals are... wait for it... I=0, R=1, G=1, B=0. That is, for this particular color 6. For that color, Q206 (a normal NPN transistor) is therefore turned on, and this will reduce the level of green in the final display. In other words this color is a little redder than you might expect.

Now the interesting question is - does the original M24 colour monitor do the same, i.e. display dark yellow as brown? Anyone has a colour monitor to check this? (although I guess we could check the schematics, if we have them?)

 

[Trixter]

Incidentally Olivetti calls he DEB board DOB or DCOB. Lord nows when l'll get to see a photo of it

I'm an M24 collector. I have several 6300s and variants, and a Xerox 6060. Trying to get my hands on a real M24.

I have the DEB and have scanned the docs and archived the software and put them on ftp.oldskool.org. I have not yet taken pictures of it or put it into a system, but I'm definitely planning on doing so sometime.

Now the interesting question is - does the original M24 colour monitor do the same, i.e. display dark yellow as brown? Anyone has a colour monitor to check this? (although I guess we could check the schematics, if we have them?)

I have a few color 6300 monitors, and yes, color #6 is displayed correctly.

 

[Trixter]

For pure 15.7 KHz TTL, I've used the gglabs CGA2RGB unit which converts TTL to RGB with the color #6 fix to show brown. From there, you have 15.7 KHz horizontal / 59.9227 Hz vertical RGB analog signal you can figure out what to do with. You can use a line doubler or scan converter to make it 31.4 KHz horizontal to show on a VGA monitor, or you can connect it to a capture device. I do the latter, connecting it to a Epiphan capture card that lets me specify the exact frequencies/front porch/etc. to capture CGA.

M24 is different, obviously. If we've determined it's TTL, then the output is 400 lines, 25 KHz horizontal, and about 58 Hz vertical (I don't have the timings in front of me right now but it's not identical to CGA). a TTL to RGB conversion will either require a VGA monitor that can scan down to those rates (like an early multisync, or an Amiga VGA monitor), or a scan converter or capture card compatible with those rates. The vertical refresh is not the issue; it's the horizontal to worry about.

 

[Valerio]

I have a few color 6300 monitors, and yes, color #6 is displayed correctly.

Thanks for confirming it! Do you happen to have the schematics as well? It would be interesting to see if the #6 color adjustment is made in the same way as in the IBM monitor (as well as what kind of line termination the monitor has in place for colour/sync lines).

 

[Trixter]

Sadly I've been unable to determine the exact model of the monitor to find the specs; I suspect I'll have to open it up and look for something. It's an AT&T CRT318H as marked, but that doesn't tell you anything. In previous discussions on this forum, it was guessed it was either Hitachi or Hantarex. (The M24 monochrome monitor is most likely a Hantarex MDU 1231)

 

[1ST1]

How about a simple solution like that?

 

[Chuck(G)]

Got to be a little more complicated. VGA is analog 0->0.7 volt full-intensity; M24 video out is 5V TTL.

 

[1ST1]

Ok, then different resistor values.

 

[Chuck(G)]

You're going to have to 'splain to me how that works, then.

 

[1ST1]

Right, I forgot. But with little modification it should work. Increase just the upper resistor value and add another 58 ohm to ground. The upper resistor and the new resistor behave as a voltage divisor. 5V should be 0,7V at the output. So you need a voltage divisor with the factor 7.14 (5V/0,7V=7,14). 58*7,14=430 ohm (not really, exactly it is 414 ohm, but 430 is closest availabe)

Intensity is at high level: R, G, B is not influenced by the diode, so full TTL level goes through 1:7,14 voltage divisor to about 0,7V VGA. Red is red. Green is green, Blue is blue.

Intensity is at low level: The diode pulls R, G, B towards 0V over the 58 ohm resistor. This 58 ohm is now (as intensity is on low (ground) level parallel to the other 58 ohm which is permanently on ground. two parallel resistors: 58²/(58+58 )=29 ohm, in total half the value of one 58 ohm resistor. That means the voltage divisor is now 1:14,28, that means (without the diode) R, G, B output is maximumm 0,35V on 5V TT. But as there is the silizium diode which has a breakthrough voltage of 0,6V, the R, G, B voltage is somewhere at 0,45...0,6V (I would need spice simulator to see exact behaviour). That is lower than 0,7V and the monitor should display a darker color. As this should be the vice versa of the intensity signal meaning, we need an inverter for the intensity signal. Or double the 430 ohm and reverse the direction of the diode.

 

[Chuck(G)]

That's beginning to get at what I meant by "more complicated". The output of RGBI is LSTTL, so it's not nearly rail-to-rail. which is why I suggested some CMOS or at least a JFET switch.

For me, the simplest would be a switchable high rail between 0.7 and 0.4 (for example) V, driven by a MOSFET switch off the I pin. The RGB outputs would probably drive either CMOS or JFET switches from that. Doubtless there's a video mixer/selector IC out there that could pretty much do the job.

(I like JFET in this application as they conduct at 0V--you turn them "off" by applying a suitable "pinchoff" voltage. That's another reason you see them used in video work).

 

[Trixter]

Intensity is at high level: R, G, B is not influenced by the diode, so full TTL level goes through 1:7,14 voltage divisor to about 0,7V VGA. Red is red. Green is green, Blue is blue.

This wouldn't get yellow correct, however (see previous discussion on how the monitors intentionally display yellow as dark brown).