CPC2VGA

[eto]

To easily connect a CPC to a VGA cable I'd love to have an adapter PCB which is much easier and more reliable to solder than an adapter cable. I made a quick design for a PCB but I am not sure if it should contain resistors or not. 

For the Atari ST, it also works without anything but it's reported, that with resistors in the RGB lines (45-100 Ohms are mentioned) the quality will be (much) better. Also someone discovered that some capacitors for hsync and vsync (10nF and 470pF) can help with the quality. 

Will those resistors be a benefit on the CPC too? And would a capacitor for Csync also improve the quality?

[McArti0]

For real VGA you must add separator Hsync and Vsync (PIN14)
Csync work only with VGA-RGB-IN. like TV or 15kHz Monitor.

[eto]

For real VGA you must add separator Hsync and Vsync (PIN14)

I couldn't find the original spec but according to some sources (e.g. https://en.wikipedia.org/wiki/VGA_connector) pin 13 accepts both H-sync and C-sync. It might be that some VGA monitors behave differently but I have never come across one. I own an adapter which splits C-Sync into H-Sync and V-sync but I never had a situation where this adapter works but a pure 1:1 connection from C-Sync to pin 13 doesn't.

[McArti0]

I have BW analog VGA monitor H35kHz. And I must add Lm1881 separator to use it.

[Bryce]

That's the problem with VGA monitors. What works on one, doesn't necessarily work on another. If you are going to make a really good CPC to VGA adapter, you should make it extremely flexible. Switchable variable resistors on the colour signal and the choice to split or combine the sync signals.

Bryce.

[robcfg]

I'd like to try the simple one because I'd like to use it with a VGA to HDMI converter.

If it works, it would make me very happy 

[eto]

As I don't know how to apply a sync splitter to it, it probably will be not a very good adapter, but maybe a good enough adapter for most cases.

I just ordered some PCBs for another project where I use the same VGA connector footprint. Once they are here I can check that the footprint fits. If that's the case, I will order the CPC2VGA PCBs and give it a try. If everything works, I will publish it. As I will have plenty of PCBs I can also send them to whoever is interested and is willing to pay for postage (worldwide 2€, Germany 1€, all untracked).

[Bryce]

As I don't know how to apply a sync splitter to it, it probably will be not a very good adapter, but maybe a good enough adapter for most cases.

Do a search for LM1881. That chip does all the splitting for you.

Bryce.

[eto]

Do a search for LM1881. That chip does all the splitting for you.

I did. But it seems it needs some additional resistors and capacitors and I am not sure if I can just copy/paste the schematics.

Regarding power supply: A few years ago I bough an adapter from Spain which doesn't use the LM1881 and requires no power. From the components it looks very similar to a 74LS123 based splitter. I don't know for sure as the number is removed from the IC.

 Is it possible that this gets the power from the LUM signal like the Scart hack does? If yes, could that also work for the LM1881 or LM1980?

[SerErris]

For a proper implementation it also needs a nand and an and gate as well.

See here:
https://www.epanorama.net/documents/vga2rgb/vgamonitor.html

In the section of "How can I build a circuit which converts composite sync signal to HSYNC and VSYNC ?"

There is the full layout. The problem is, that the LM1881 does not ouput HSYNC, but just VSYNC and passes a Composite Sync (so the composite signale without the actual video signal). 

It might work, but it also might not. 

The proper implementation would be with the two gates to ensure, that you get proper HSYNC signal.

[XeNoMoRPH]

Do a search for LM1881. That chip does all the splitting for you.

I did. But it seems it needs some additional resistors and capacitors and I am not sure if I can just copy/paste the schematics.

Regarding power supply: A few years ago I bough an adapter from Spain which doesn't use the LM1881 and requires no power. From the components it looks very similar to a 74LS123 based splitter. I don't know for sure as the number is removed from the IC.

 Is it possible that this gets the power from the LUM signal like the Scart hack does? If yes, could that also work for the LM1881 or LM1980?

I have this too hehe.

[SerErris]

@eto it looks like he uses a proper implementation with this chip EL4583, which is a better solution to the problem, as it directly outputs VSync and HSync and you only need some resistors and capacitors:

Where do you get the information that it uses a 74LS123 ?

https://www.digikey.de/de/products/detail/rochester-electronics-llc/EL4583AIS-T7/11523794
https://rocelec.widen.net/view/pdf/598jp2pzwr/INSL-S-A0001389934-1.pdf?t.download=true&u=5oefqw

[SerErris]

And here is a link to another picture, where you can read half of the chip marking.

Does not look like a 74LSxxx chip.

[eto]

Where do you get the information that it uses a 74LS123 ?

It was a guess based on this circuit which contains very similar elements: https://www.electroschematics.com/sync-signal-separator/

The problem is, that the LM1881 does not ouput HSYNC, but just VSYNC and passes a Composite Sync (so the composite signale without the actual video signal). 

I read that this is not a problem if the H-Sync is fed with C-Sync. 

t looks like he uses a proper implementation with this chip EL4583

The EL4583 is fed with Composite video which the CPC does not have. It has only Composite Sync. LUM is "kind of" a composite video signal.

I think I will check the adapter and see how the pins are connected. That might give an idea. Also how it gets power.

[SerErris]

Typically with VGA, there are really all kind of monitors. Some of them do accept CSync at HSync, Some of them do not accept CSync at HSync at all and some do not if VSync is available (HSync cannot be CSync).

Regarding Composite Video - That is true for both LM1881 and EL4583.

[eto]

I think I will check the adapter and see how the pins are connected. That might give an idea. Also how it gets power.

I have absolutely no clue what that thing is doing. CSync of the CPC is directly connected to pin 13 of the VGA connector. CSync also goes to a pin of the IC but it's neither the pin it should be for the EL4583 nor the LS123. I also do not understand where it could get the power from. 
Pin 9 of the VGA connector is connected to some if the passive components which doesn't make any sense at all as this should be a 5V signal to the monitor. 

Unfortunately I can't test any of the pins on one side of the IC as they are covered by the VGA connector. 

[SerErris]

It is definetely not an 74LS chip. That marking would be clear do read and longer then the red X in the picture above.

You could pair it with a 74XX00 Chip (4-Nand) to implement the clean H-Sync.

However the diagram posted here  https://www.electroschematics.com/sync-signal-separator/ must be faulty.

It connects with a R/C timing network the VCC line with CExt but that is supposed to get placed between RCExt and CExt. In the first vibrator that is correct, in the second it is wrong (to my understanding) RCExt is floating, which would result in unpretictable behaviour. (Pin7 and Pin6 should be connected, not Pin16 and Pin6).

There is another chip (10 pins only) LMH1980,

https://community.lzxindustries.net/t/sync-extraction-circuit/3496/5

That looks like a very simple setup, because you can ignore all the input stuff from CSync IN (e.g BNC buffer and stuff) as we do not have a high frequency signal and it is also only a Sync signal without any video data in it.

This chip is VSSOP, so pretty small on the PCB and requires minmal other components. It actually requires 1 Resistor to pull Rext down and on decoupling capacitor (they put in two in the schematics) 10uF and 0.1uF.

Not sure about the output Resistors, maybe you want to put them in, but I thing they are there to get a 100Ohm output resistance as defined resitance.

https://www.ti.com/lit/ds/symlink/lmh1980.pdf?ts=1701767569476&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252Fde-de%252FLMH1980

[SerErris]

BTW: I still do not understand where the Power Supply is coming from.

[SerErris]

What Chip code do you read on the X-Picture above?

Ist it:

• 86A 72PM
• 86A 7CPM
• 8oA 7CI1 

?

I could not find anything on that chip.

[SerErris]

Here is a quick setup with the LMH1980, that should actually work and is very little efford. I am not sure about the 75Ohm resistors to be honest. The specification VGA mentions them in the signal path, but not for the logic ... so this could be either logic or signal - not sure.

So if they are not required, you can simply bridge them.

[Bryce]

What Chip code do you read on the X-Picture above?

Ist it:

• 86A 72PM
• 86A 7CPM
• 8oA 7CI1

?

I could not find anything on that chip.

Those are manufacturing dates and locations. Not the Chip part number.

Bryce.

[eto]

BTW: I still do not understand where the Power Supply is coming from.

At the moment I guess it comes from pin 9 of the VGA connector. I read somewhere that pin 9 sometimes carries 5V. Not sure if that's really the case. 

That finally could also explain why my NEC 1970 works with the ST but not with the CPC. I checked its schematics of the NEC and it expects V-sync and H-sync on separate pins and pin 9 does not supply 5V. Then the Aranet adapter would just be a pure passive adapter which only provides C-sync to pin 13 and the NEC will not receive the right signal.

For a community solution I personally would prefer to use components that are easily available. I can't find a source to buy the LMH1980 in small quantities. The LM1881 however is easily available. I think it might even be able to use through-hole components so the whole adapter still fits perfectly behind the CPC.

[Bryce]

Pin 9 might be connected on the monitor side in some monitors, but it not standard and relatively rare. If that's the source, then the adapter will only work on a few carefully chosen monitors. It has to get power from somewhere though. The CPC LUM can be used (with a capacitor) to build up a voltage, but that could never supply enough current to power even a very low power 5V IC.

Thru-hole LM1881 are still easy to find.

Bryce.

[SerErris]

Thruhole is to large for this connector. 

Here you can buy it in small and large quantities:

https://www.digikey.de/de/products/base-product/texas-instruments/296/LMH1980/62143

[eto]

Here you can buy it in small and large quantities:

in theory yes, in practice it's 18€ for shipping which makes it no longer feasible.

Thruhole is to large for this connector.
 

I found a very simple circuit which has been reported to work in a similar context. If that works for the CPC, it should fit onto a PCB. Even including a 5.5/2.1mm connector. The 5V from the power supply go into the adapter and from there a short power cable into the CPC.