I've seen a few things out there in looking around, but they seem pricey and
questionable if they even would work. Seems like many of them are designed
to convert the output of old console arcade games to work with VGA monitors.
They are labeled as converting CGA to VGA, but there is some question in my
mind (based on searching the net) whether they would work with a PC source -
TTL digital CGA as opposed to analog RGB VGA (if there is such a thing and
this makes any sense).
Obviously the obvious route would be to put a VGA card in the old PC in
question, but I just think it would be nice to be able to use old PC's
without having to swap around a video card.
Thanks,
Wesley
http://www.buy.com/retail/product.asp?sku=203067985&listingid=10093700
http://www.cablestogo.com/product.asp?cat_id=113&sku=02458
--
-- Marten Kemp (Fix ISP to reply)
You can't help being ignorant 'cause there's always
something you don't know; what you can't be is stupid.
It wouldn't cost much to try this:
http://www.monoprice.com/products/product.asp?c_id=104&cp_id=10414&cs_id=1041401&p_id=1144
Or has someone had success with such adapters?
Wesley
"Marten Kemp" <marte...@thisplanet-link.net> wrote in message
news:hd6ei0$4mh$1...@aioe.org...
>
> http://www.buy.com/retail/product.asp?sku=203067985&listingid=10093700
> http://www.cablestogo.com/product.asp?cat_id=113&sku=02458
> http://www.buy.com/retail/product.asp?sku=203067985&listingid=10093700
> http://www.cablestogo.com/product.asp?cat_id=113&sku=02458
These adapters will not allow you to use a CGA monitor with a VGA monitor.
CGA output is TTL with a line rate of ~15kHz.
VGA output is analogue RGB with a line rate of ~31kHz.
You need a purpose-built circuit/device that converts CGA TTL to analogue,
and also does scan-rate conversion (doubling). And yes, it is possible - but
not easy - to find a VGA-compatible monitor that will synz at 15kHz.
Regards,
--
| Mark McDougall | "Electrical Engineers do it
| <http://members.iinet.net.au/~msmcdoug> | with less resistance!"
They appear to work but they don't support the 4th "intensity" pin, so
you get only the lower 8 colors.
> Obviously the obvious route would be to put a VGA card in the old PC in
> question, but I just think it would be nice to be able to use old PC's
> without having to swap around a video card.
One thing you can do to use a modern LCD panel with CGA is to simply
to hook it up via the CGA composite output of the CGA card (assuming
your LCD panel has composite input, which an increasing number do as
they're being used for more than just computer monitors). This will
actually downgrade your experience slightly, if you're used to CGA
RGB, but the upside is that you can now run all those games that
support composite color, giving you 16 colors in games. Flight
Simulator, all Sierra AGI games, California Games, and many more
support up to 16 colors in that mode, which is a nice surprise if
you've never seen your favorite games in 16 colors.
That won't work; that's a 9-pin-analog to 15-pin analog adapter. No
conversion of any kind.
In principle an analog signal like the line out from a TRS-80 model-I could be
compatible with VGA but the monitors don't like the timing. CGA is 60Hz so
creating a CGA interface for the TRS-80 is easy. I have done it, took one
GAL22V10 only and it emulates the LE15 colour graphics for Video Genie using the
standard 60Hz timing (Genie-I jumpers moved for NTSC timing).
Much of what you can buy is for converting a composite video signal to VGA
timing. These work with more or less success because they are aimed at
converting analog signals like movies and tv, smoothing out jaggedness and moire
effects by filtering.
A tv input card would do the same trick but you would need a pc (or other
computer) to run it. What they do is similar so with more or less success. I
have 3 computers that can input video and display the result on a VGA monitor.
They are all pretty bad, possibly slightly better than my old color tv through RF.
Knut
small correction, CGA is 15kHz, where VGA is 31kHz.
They both run at 60hz (though VGA's usually can go up higher.
> Anybody out there know of an adapter that will work to convert CGA
> (or EGA or mono even) to VGA so that a modern monitor can be used
> with a vintage PC?
Sorry Wesley, but there is no adapter to do that.
You will need a new graphics card that supports VGA!
> Obviously the obvious route would be to put a VGA card in the old PC
> in question, but I just think it would be nice to be able to use old
> PC's without having to swap around a video card.
NO, by _technical_ reasons there's no way to perform this!
You need a *new VGA graphics card* to perform this conversion.
Horst
> Sorry Wesley, but there is no adapter to do that.
Wrong.
> NO, by _technical_ reasons there's no way to perform this!
> You need a *new VGA graphics card* to perform this conversion.
Wrong again.
<http://www.highway.net.au/parts/converters/1488.html>
<http://www.ambery.com/rgbcgatovgac.html>
FWIW my colleague and I are currently designing an all-in-one video
converter for the retro-bobbyist market. Unfortunately, it won't be cheap.
But it will handle composite/s-video/component/analogue-rgb/digital-rgb. CGA
comes under the latter category. It will also do full frame-rate conversion,
eg. from 50Hz PAL to 60Hz DVI/VGA. Output is analogue/digital VGA.
Unfortunately these devices don't convert actual/true CGA because they
don't handle the intensity pin, so you only get 8 colors out of it
instead of the full 16. The wiring diagram that proves this is here:
http://site.ambery.com/webgraph/CGA-VGA-Adapter.gif
> FWIW my colleague and I are currently designing an all-in-one video
> converter for the retro-bobbyist market. Unfortunately, it won't be cheap.
> But it will handle composite/s-video/component/analogue-rgb/digital-rgb. CGA
> comes under the latter category. It will also do full frame-rate conversion,
> eg. from 50Hz PAL to 60Hz DVI/VGA. Output is analogue/digital VGA.
I'd be happy to speak to you offline about things you'll need to
include if you want to truly convert CGA properly. For one thing:
http://www.oldskool.org/pc/cgacal (check under the text "color #6 is
brown" for why you not only need to handle the intensity pin but also
why you need to treat one specific combination differently)
> Unfortunately these devices don't convert actual/true CGA because they
> don't handle the intensity pin, so you only get 8 colors out of it
> instead of the full 16. The wiring diagram that proves this is here:
> http://site.ambery.com/webgraph/CGA-VGA-Adapter.gif
Fair enough, I didn't look into this close enough. It also looks to be the
case with _both_ converters. But my point was, Horst's claim that it was not
technically possible is simply incorrect.
> I'd be happy to speak to you offline about things you'll need to
> include if you want to truly convert CGA properly.
I appreciate the offer, thanks, and when the time comes, if we can't get it
sorted I'll take you up on that. But it would seem your reference should be
sufficient - thanks again for that!
If I seem somewhat unconcerned with the details atm, it's because a large
part of the design is implemented within an FPGA, and the TTL pins in
particular are simply fed into the FPGA to be handled internally. So in some
respects, "it's just software" handling how those pins are interpreted. In
reality, and in light of your revelation, there'll probably be a look-up
table in there somwhere. It will also have to handle non-CGA TTL inputs,
such as the Sord M23, so there'll also be an (menu) option specifically for
CGA I guess...
You could, of course, do CGA->VGA conversion much cheaper than what we're
desiging. I suspect the fact that you can't buy an off-the-shelf product is
simply because no-one wants to do it! I mean - retro-hobbyist aside - why
would you do it when a much cheaper and better option is a VGA card!?!
'Cause a VGA card would use up segment A,
which I want to use for 704k base memory.
I'm kinda hoping that a 9-pin-to-15-pin
adapter will alleviate the need for a
monochrome adapter on an overly-crowded
workbench.
But only a retro-hobbyist would want to do it :-) And the reason is
because there are some PC games that only display properly on true CGA.
If the card is CGA, no. The signals are incompatible (TTL vs.
analog). There is an analog 9-pin VGA standard, but you usually see
it on the monitor, not the card (ie. I've never seen a VGA card with 9-
pin analog output, only monitors with 9-pin analog input). It is
those monitors that those adapters are built for.
No, a retro-hobbyist wouldn't be interested because it's too recent.
They'd want a composite NTSC to VGA adapter, since that's what most small
computers used before (and even after) the IBM PC came along.
Michael
> No, a retro-hobbyist wouldn't be interested because it's too recent.
CGA graphics adapters first appeared in 1981. The Amiga 1000 was introduced
in 1985. Are you saying that Amiga enthusiasts aren't retro-hobbyists???
> They'd want a composite NTSC to VGA adapter, since that's what most small
> computers used before (and even after) the IBM PC came along.
True, but there are also examples that use s-video, TTL and (more-so
consoles) using component video as well as analogue RGB with 15kHz line
rates. And don't forget hobbyists in PAL regions either!
> 'Cause a VGA card would use up segment A,
> which I want to use for 704k base memory.
There is http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=310179198136
*Arcade RGB CGA to TV and Video Composite converter NEW*
IT IS A RGB, CGA to TV Set and S-Video converter.
It can Allows displaying Simultaneous RGB,CGA output signal to your
TV PAL/NTSC selectable
FEATURE
*PURE HARDWARE DESIGN,JUST PLUG AND PLAY
*SUPPORT RGB SIGNAL CONVERTING TO VIDEO SIGNAL AND S-VIDEO
SIGNAL
*SWITCHHABLE PAL/NTSC VIDEO SYSTEM
*ADJUSTABLE 3 COLOR (RGB) BRIGHTNESS
I'm curious why running 704K free DOS RAM is a requirement. Is it
because you're curious to see how it works? If it's because you want
to run a lot of TSRs, there are ways to do that without significantly
impacting DOS RAM.
There are a handful of games that only run and/or display properly on
a true CGA card connected via RGB. Composite doesn't work with them.
So yes, there is definitely interest in retro-hobbyists. I'm one, and
I have been looking for such a device for over a decade as I only have
two RGB monitors left (more if you count the 400-line proprietary AT&T
stuff and yet more if you consider the horrible PCjr monitors in my
collection).
Over a decade ago there was a device called the Video Turtle that
claimed to convert TTL from CGA and Apple IIGS to S-Video. I bought
one for $150, it never worked right, and by the time I called for
support a few months later the company was gone. I've also looked at
very early devices meant for hooking computers up to coax projectors
(ie. the output is R/G/B/H/V component analog, which can be scan-
converted) and I have an early Extron unit that should work, but I
have been unsuccessful finding the custom cable it needs for CGA/TTL
input.
So yes, there is definitely interest.
Because I have a modified U44 PROM that'll map memory
into segments A, C, D and E, plus one device driver
that'll turn up tp a full segment into an LRU disk
sector buffer and another that'll turn the rest into
XMS UMBs so I can use LOADHI and DEVICEHI.
I'll copy the floppy and SCSI controller BIOSen into
a single EPROM and put it in the socket where ROM
BASIC usually lives; might have to tweak the BIOS a
bit to scan F000-F800 for device BIOSen and totally
ignore BASIC.
Version 2 will be to change the address decode on an
IBM MDP card to segment E and have 896k base memory.
Just to see if I can.
> Horst Franke wrote:
>> Sorry Wesley, but there is no adapter to do that.
> Wrong.
>> NO, by _technical_ reasons there's no way to perform this!
>> You need a *new VGA graphics card* to perform this conversion.
> Wrong again.
Sorry Mark, but explain your "wrong statements" more precisely!
> FWIW my colleague and I are currently designing an all-in-one video
> converter for the retro-bobbyist market. Unfortunately, it won't be cheap.
Why do you want to design a converter while native design is appropriate?
And you state *won't be cheap*?
Any converter will loose something of the original!
OK, you want to make money - but there's no technical reason behind.
See also the later issues that provide more details about possible problems.
My issue is that native presentation by an adapter (VGA card) will be
more cost sensitive than any converter. See above about loosing details.
So my previous statement about adapter=no is obsolete.
By adaptor I also unterstood a device to convert something.
It's always best to provide native levels to the required output (VGA).
An other issue may be if the motherboard has no interface for an
external adapter card. Then look for another motherboard.
I don't see any reason why to use such old boards any more.
If they have a CGA interface then connect a CGA monitor.
If you want VGA output then add a VGA adapter card.
A converter will normally be much more expensive (but not even
solve all of the converting problems). See the later issues.
Horst
>>> NO, by _technical_ reasons there's no way to perform this!
> Sorry Mark, but explain your "wrong statements" more precisely!
You state emphatically that it is technically _impossible_ to convert CGA to
VGA. That statement is simply incorrect. I don't really understand what you
are asking me to explain?
> Why do you want to design a converter while native design is appropriate?
I don't understand this statement.
> And you state *won't be cheap*?
> Any converter will loose something of the original!
> OK, you want to make money - but there's no technical reason behind.
> See also the later issues that provide more details about possible
> problems.
The video converter was never driven by any particular desire to convert CGA
to VGA - in fact I didn't even have CGA in mind when specifying the TTL
interface - and we started on it well before this thread arose. I mentioned
the converter here only because it will happen to support CGA-VGA.
Make money? No, defintitely not. This converter was borne out of a need by
myself and my colleague to support a large variety of retro computers on VGA
monitors. My 4th 1084S recently died, and it's impractical to keep repairing
them. They also take up a lot of room. So we decided to design a single box
that we could use on the various TRS-80's, Apples, Commodores, Sinclairs,
Sord M23 etc, not to mention the retro consoles we have as well. If we sell
a few to help fund the development costs, then great!
A for _losing_ detail (note spelling) - that's not necessarily the case. You
keep mentioning "loss of detail", but neglect to supply any technical
explanation of what you're referring to. So I will.
One issue is sampling of the CGA signal in sync with (and centred on) the
dot clock from the CGA card. Unfortunately CGA, unlike some other TTL video
outputs, does not supply the dot-clock on the interface. So in the case of
CGA it is more problematic. It is possible to do clock-recovery given a
suitable output, or it is possible to super-sample the output and work back
from there. I am yet to decide what method(s) we will employ - but most
likely will simply lock a clock to each hsync, which is how most video
sampling ICs work.
A 2nd issue will be up-scaling the display to the native resolution of the
monitor. Obviously not an issue when the native resolution is an integer
multiple of the CGA output, but this won't always be the case. But we will
be using sophisticated scaling algorithms that do a good job of scaling.
So the problem won't be _loss_ of detail, but rather loss of definition, if
you like. However, given the relatively low resolution of CGA, scaling
algorithms and relatively high resolution of modern VGA monitors, I very
much doubt that anyone could complain about the quality. This is not about
reproducing an absolutely _perfect_ CGA output (_any_ conversion by its very
nature is going to fall short here), it's about producing the best possible
output on a VGA monitor.
To be clear, I'm not proposing at all that our box is cost-effective for
anyone who wants to _only_ do CGA-VGA - far from it! But if you own a range
of *PAL* and NTSC computers/consoles that have composite, s-video,
component, analogue rgb and digital rgb outputs (as I do) and want to be
able to display them _all_ on a single VGA/DVI monitor - scaled to fit the
whole screen, or scaled approriately on a wide-screen monitor - then you may
wish to consider our box.
> My issue is that native presentation by an adapter (VGA card) will be
> more cost sensitive than any converter.
Well that's pretty obvious to anyone I would've thought.
> I don't see any reason why to use such old boards any more.
That's not the point at all. Wesley _wants_ to use an old computer for his
own reasons - you can't just say "I don't see any reason to use such an old
board, get a newer one".
Hope this makes things clearer.
> Fair enough, I didn't look into this close enough. It also looks to
> be the case with _both_ converters. But my point was, Horst's claim
> that it was not technically possible is simply incorrect.
Sorry Mark
I didn't talk about technical possibilities but about feasibility!
Say you "speak" chinese by help of a translator - then YOU really
don't speak chinese. And the translator will not be cheap.
Horst
> Sorry Mark
> I didn't talk about technical possibilities but about feasibility!
I quote...
"NO, by _technical_ reasons there's no way to perform this!
You need a *new VGA graphics card* to perform this conversion.
Horst "
So I'm not sure how you can make this statement?!?
Anyway, it's not my intention to be pedantic or specifically take you to
task on your opinion - no-one else will benefit from such a discussion.
Perhaps we're simply arguing semantics here, hindered by the very nature of
email exchanges and (I suspect) that German is your 1st language.
I'm merely offering my opinion - for *Wesley's* benefit - that it is indeed
_technically_ possible to convert CGA to VGA. Whether or not it can be done
cost-effectively is another matter (and to some extent, subjective).
My other recent post on this subject may clear up my position somewhat.
> Horst Franke wrote:
> "NO, by _technical_ reasons there's no way to perform this!
> You need a *new VGA graphics card* to perform this conversion.
> Horst
> So I'm not sure how you can make this statement?!?
>
> Anyway, it's not my intention to be pedantic or specifically take you
> to task on your opinion - no-one else will benefit from such a
> discussion. Perhaps we're simply arguing semantics here, hindered by
> the very nature of email exchanges and (I suspect) that German is
> your 1st language.
YES
> I'm merely offering my opinion - for *Wesley's* benefit - that it is
> indeed _technically_ possible to convert CGA to VGA. Whether or not
> it can be done cost-effectively is another matter (and to some
> extent, subjective).
Ok, ACK.
But to convert is the 2nd choice from my point of view.
Horst
> Horst Franke wrote:
>>>> NO, by _technical_ reasons there's no way to perform this!
>> Sorry Mark, but explain your "wrong statements" more precisely!
>
> You state emphatically that it is technically _impossible_ to convert
> CGA to VGA. That statement is simply incorrect. I don't really
> understand what you are asking me to explain?
>> Why do you want to design a converter while native design is
>> appropriate?
> I don't understand this statement.
Ok. Then see my example with the chinese translator.
>> And you state *won't be cheap*?
>> Any converter will loose something of the original!
>> OK, you want to make money - but there's no technical reason behind.
>> See also the later issues that provide more details about possible
>> problems.
> The video converter was never driven by any particular desire to
> convert CGA to VGA - in fact I didn't even have CGA in mind when
> specifying the TTL interface - and we started on it well before this
> thread arose. I mentioned the converter here only because it will
> happen to support CGA-VGA.
> Make money? No, defintitely not. This converter was borne out of a
> need by myself and my colleague to support a large variety of retro
> computers on VGA monitors. My 4th 1084S recently died, and it's
> impractical to keep repairing them. They also take up a lot of room.
> So we decided to design a single box that we could use on the various
> TRS-80's, Apples, Commodores, Sinclairs, Sord M23 etc, not to mention
> the retro consoles we have as well. If we sell a few to help fund the
> development costs, then great!
OK, but I think there's no relation to IBM PC CLASSIC.
> A for _losing_ detail (note spelling) - that's not necessarily the
> case. You keep mentioning "loss of detail", but neglect to supply any
> technical explanation of what you're referring to. So I will.
> One issue is sampling of the CGA signal in sync with (and centred on)
> the dot clock from the CGA card. Unfortunately CGA, unlike some other
> TTL video outputs, does not supply the dot-clock on the interface. So
> in the case of CGA it is more problematic. It is possible to do
> clock-recovery given a suitable output, or it is possible to
> super-sample the output and work back from there. I am yet to decide
> what method(s) we will employ - but most likely will simply lock a
> clock to each hsync, which is how most video sampling ICs work.
Sorry but that technical detail was out of my mind.
I responded from an end user point of view.
> A 2nd issue will be up-scaling the display to the native resolution
> of the monitor. Obviously not an issue when the native resolution is
> an integer multiple of the CGA output, but this won't always be the
> case. But we will be using sophisticated scaling algorithms that do a
> good job of scaling.
NO. This is convertion issue but not native related.
> So the problem won't be _loss_ of detail, but rather loss of
> definition, if you like. However, given the relatively low resolution
> of CGA, scaling algorithms and relatively high resolution of modern
> VGA monitors, I very much doubt that anyone could complain about the
> quality. This is not about reproducing an absolutely _perfect_ CGA
> output (_any_ conversion by its very nature is going to fall short
> here), it's about producing the best possible output on a VGA monitor.
ACK from my site. But I don't like "definition" in this context.
And CGA low resolution should everyone understand.
> To be clear, I'm not proposing at all that our box is cost-effective
> for anyone who wants to _only_ do CGA-VGA - far from it! But if you
> own a range of *PAL* and NTSC computers/consoles that have composite,
> s-video, component, analogue rgb and digital rgb outputs (as I do)
> and want to be able to display them _all_ on a single VGA/DVI monitor
> - scaled to fit the whole screen, or scaled approriately on a
> wide-screen monitor - then you may wish to consider our box.
OK. ACK.
>> My issue is that native presentation by an adapter (VGA card) will be
>> more cost sensitive than any converter.
And why you then state it's not cheap?
A VGA adapter should be less than 20 USD.
>> I don't see any reason why to use such old boards any more.
>
> That's not the point at all. Wesley _wants_ to use an old computer
> for his own reasons - you can't just say "I don't see any reason to
> use such an old board, get a newer one".
Ok, but then he will need to look for "personal" solution.
Either he already have a CGA monitor or need to "upgrade".
Routed to comp.sys.pc.classic
Horst
> And why you then state it's not cheap?
> A VGA adapter should be less than 20 USD.
I never said this. I agree with you, using a VGA card is going to be - by
far - the cheapest solution. In fact, it shouldn't be difficult to find one
for _free_!
However, this is academic, as Wesley has stated why he can't use a VGA card.
Please keep us posted as to the progress of your converter. If it works as
well and with as many things as you are planning, I'd love to have one,
assuming I can justify whatever the cost ends up being.
Thanks,
Wesley
"Mark McDougall" <msmc...@no.spam.iinet> wrote in message
news:yNudnYt8VcpQ753W...@westnet.com.au...
> Please keep us posted as to the progress of your converter. If it works as
> well and with as many things as you are planning, I'd love to have one,
> assuming I can justify whatever the cost ends up being.
It has been on hold for a few months now but we're planning to get back to
it early in the new year. The PCB layout is almost complete!