He looks a little tired in that picture and isn't dancing very much. Are you sure he's getting 5V?
SuperMac VidCards - Timeline Development - Links
- Thread starter Trash80toG4
- Start date
Thanks for thumbnailing!
I appreciate it. Also - love the enthusiasm! On the timing, it's not really working yet. I can get a Dell monitor to sync at 800x600, but it's more lucky than right.
I am hoping to make the 2 SVGA modes work at the stock oscillator frequencies. SuperVideo does not calculate oscillator values; VideoHelper 1.0 (the original HyperCard version) does do it (which was one of the reasons it came into being in 1988, esp. since Excel access was not universal in the early days).
I will have to put the configs on a scope to see what my Spectrum/8 Series II ("Spectrum C") is up to based on the starting values. The problem with the horizontal divider is that, with a /8 divider, for example, every +/- 1 horizontal value in SuperVideo translates to 8 horizontal clocks on the TMS34061 (coarse timing granularity). So, it might (or might not) be tricky to align/tweak the signal so that it is recognized as valid by the target monitor within its signal tolerance thresholds. But, I do need to see what the timing looks like in order to do that. The VideoHelper 2.0 spreadsheet shows the exact/expected timings per the spec. So, if my scope works, it should be easy to spot how the video output is varying from the expected timing.
Why are you not considering a clock replacement for a workable frequency? Seems like you're fighting with one hand tied behind your back?
Manual for SpecA says you can use freq 2x the /8 divider range (16MHx-52MHz) by using the /16 divider (52MHz-104MHz). WAG there would be that a better match for frequency required might be available in that range?
Just caught on to the Oscillator Frequency required being the reciprocal of Pixel Rate targeted. Off at lunch ATM so can't figure that one out, but it explains why no calculation of frequency would take place in SuperVideo I think?
edit: that's all in steps 3 and 4 of the setup guide.
Manual for SpecA says you can use freq 2x the /8 divider range (16MHx-52MHz) by using the /16 divider (52MHz-104MHz). WAG there would be that a better match for frequency required might be available in that range?
Just caught on to the Oscillator Frequency required being the reciprocal of Pixel Rate targeted. Off at lunch ATM so can't figure that one out, but it explains why no calculation of frequency would take place in SuperVideo I think?
edit: that's all in steps 3 and 4 of the setup guide.
Last edited:
Changing the oscillator to a different or higher frequency is cart before the horse. I don't yet know why the timing with the exact/expected SVGA clocks isn't working. If it is not possible to manipulate the timing with the expected clocks, then looking for alternate frequencies to shift the timing window is certainly possible, whether they are close to the target values (likely) or even above 52MHz (where the /16 divider would take effect).
If you are interested in understanding the timing math, have a look at the cells in the VideoHelper 2.0 spreadsheet. You may also be able to compare the SVGA timing with other published sources.
The /16 divider has less granularity than /8 (but also, with possible x2 harmonic overlap), where a change of 1 horizontal VIDCLK value shifts the timing window by 16 pixel clocks.
Anyway, don't worry. The scope doesn't lie.
Also, I realized that I actually have another 15" Multiscan Apple monitor in storage. So, that one may be working and will let me go forward with live testing after using the scope. It supports 60 and 72Hz SVGA 800x600 to act as a visual confirmation of the scope timing. It should sync up if the timing is correct.
If you are interested in understanding the timing math, have a look at the cells in the VideoHelper 2.0 spreadsheet. You may also be able to compare the SVGA timing with other published sources.
The /16 divider has less granularity than /8 (but also, with possible x2 harmonic overlap), where a change of 1 horizontal VIDCLK value shifts the timing window by 16 pixel clocks.
Anyway, don't worry. The scope doesn't lie.
Also, I realized that I actually have another 15" Multiscan Apple monitor in storage. So, that one may be working and will let me go forward with live testing after using the scope. It supports 60 and 72Hz SVGA 800x600 to act as a visual confirmation of the scope timing. It should sync up if the timing is correct.
Was going off the early step definition in the manual/integrating hazy memories of gorgonops/eudimorphodon saying that freq/pixel clock was the baseline from which the remainder of the inputs for SuperVideo would be based?
I still haven't looked at the Spec Sheet @Bolle posted.
What's an oscilloscope?
I still haven't looked at the Spec Sheet @Bolle posted.
What's an oscilloscope?
Square is s but a special case, an equilateral rectangle.Do you mean that oscillators are not square boxes?
Be sure to skew Oscar around to face front and back somewhere in his dance routine so he can Walk Like an Egyptian.
Last edited:
Yes - if you know the resolution, the dot clock and the front porch, back porch and sync values, you can calculate everything (which you will note are the blue input boxes in the VideoHelper 2.0 spreadsheet attached above). Similarly, you can calculate one of the following missing values if you have the other 3: horizontal frequency, horizontal blanking, horizontal resolution, oscillator value.
Thanks, eventually some of this stuff will sink in. It sounded to me like you were trying to fix the out of spec dot clock input to the Controller from a "standard" oscillator by tweaking the remainder of the timings?
Is the CRT you're feeding a multiple fixed resolution display as opposed to a true multisync?
I have a 21" MAG Innovision CRT that reads out frequency input on the front panel, discriminating between VGA, Macintosh and Out of Spec on its LCD. Might feeding it questionable freqs help you out in any way?
Standard operating procedure here: muck things up as apprentice to our resident sorcerers.
Is the CRT you're feeding a multiple fixed resolution display as opposed to a true multisync?
I have a 21" MAG Innovision CRT that reads out frequency input on the front panel, discriminating between VGA, Macintosh and Out of Spec on its LCD. Might feeding it questionable freqs help you out in any way?
Standard operating procedure here: muck things up as apprentice to our resident sorcerers.
Last edited:
I am going to look at the signal on a scope, see what the card is generating based on calculated timing to look for issues/errors/deviations, and see if it is possible to adjust the signal using the expected clock frequencies. The calculations should be correct, but obviously aren't since there wasn't any sync on the monitor, but it also could have been because the monitor failed. So, I am going to try a replacement after checking the card output.
Suggestion again would be to post a screen shot of the SuperVideo inputs window with SVGA timings and freq of oscillator I can order up for parallel testing?
Yes...I will be showing scope timing and, of course, updating the spreadsheet with new values for SuperVideo vs. the ones that are there now.
Back onto the IP of this thread: I have hi res shots of five cards and a medium res shot of another so far. If anyone has one of these cards, and especially the daughtercards to add some pics to the collection it would be greatly appreciated. Need these:
Spectrum/24 v.___ - undocumented - Rev.B
Spectrum/24 Series IV
Spectrum/24 Series V
Spectrum/24 PDQ
Spectrum/24 PDQ+
Spectrum/8
Spectrum/8 Series III
Spectrum/8*24 PDQ
Spectrum Power 1152
ColorCard 24
Thunder II GX
Spectrum/8 Series II - @BFEXTU you're working with this one, no?
Thunder II GX - @siliconinsider (?) you sent me a medium res shot of this one a while back. can your upload a better pic?
OK, back to pixel partying with Oscar!
Spectrum/24 v.___ - undocumented - Rev.B
Spectrum/24 Series IV
Spectrum/24 Series V
Spectrum/24 PDQ
Spectrum/24 PDQ+
Spectrum/8
Spectrum/8*24 PDQ
Spectrum Power 1152
ColorCard 24
Thunder II GX
Spectrum/8 Series II - @BFEXTU you're working with this one, no?
Thunder II GX - @siliconinsider (?) you sent me a medium res shot of this one a while back. can your upload a better pic?
OK, back to pixel partying with Oscar!
Last edited:
This morning I came across the SuperMac video output card that hooks up a Mac with LC-style PDS to a TV.
I understand it is not a graphics card. Yet doesn't it changes frequencies to be able to display on a TV CRT?
Vintage SuperMac Simply TV Video Output Card, LC Color Classic, Performa Quadra | eBay
<p>Vintage SuperMac Simply TV Video Output Card. Part Number 1003839-0001. </p><p>Device Will Turn TVs Into Big Macs: A color graphics card will allow computer owners to view Macintosh programs on the larger screens of their television sets. E-Machines, a division of SuperMac Technology Inc. of...
www.ebay.com
I understand it is not a graphics card. Yet doesn't it changes frequencies to be able to display on a TV CRT?
I'm surprised it'll fit into the LC/605 form factors mentioned with that socketed FPU(?) and ROM on the solder side/top side of the board. I've never seen that before, don't think the lids would close. Seems like a run-of-the-mill presentation card with a CoPro Socket for the 5x0 FuglyMacs and marginally less appearance challenged CC?
VideoSpigot LC seems a far more interesting card to me.
VideoSpigot LC seems a far more interesting card to me.
It is a SuperMac card and should work to connect the LC or related Mac to the TV.
SuperMac Simply TV - Higher Intellect Vintage Wiki
wiki.preterhuman.net
Yep, says that, but SuperMac also sold the VideoSpigot LC, though it may have been developed elsewhere?
But you're missing my point. The LC/Q605 form factor's PDS spec allows for no components topside (solder side) of the PCB. I've never before seen an LC PDS card in violation of that spec. They expand horizontally, never vertically.
Just measured it up and that socketed ROM would all but, if not touch the RFI shielding of my Q605. If for the LC form factor, that card's configuration would have been flip-flopped with the four lower profile SMT packages topside and the tall components underside where they belong. That would still be out of spec, but not wildly so in terms of cubic limitations.
Both are for the the LC 020 PDS connector, but only the VideoSpigot LC meets the physical spec for that form factor. I've got that same boxed set for VideoSpigot LC, but can't get at it to check the copyright dates. %$^#$+*%& FLOODS!
The 1993 copyright date of Simply TV matches release of the LC520, which has the head room for it's oddball configuration. It looks to have been updated with S-Video output, proper DeclROM/Slot Manager architecture setup and Socket for CoPro. VideoSpigot LC lacks any ROM at all, depending entirely upon drivers for detection and function as far ask I could tell back when it was a new toy many years ago.
There are very curious things about both cards.
Last edited:
SimplyTV (1994) is just TV out at different resolutions AFAIK, maybe with virtual desktops, etc. -- and from SuperMac's post-merger days. It comes from the multimedia cross-over era when people were (more commonly) starting to use their computers to show content on TVs (and covered a number of different machine types). But, they could have done the same thing in many other machines with SuperMac cards via NTSC support with a 14.31818Mhz oscillator (and did). However, I guess in 1994, there was a market case for a dedicated product so that users could have a cheap, "big-screen" monitor.
VideoSpigot LC (1992) is for video input/capture only and is a PDS variant of the original VideoSpigot card.
Conceptually, SimplyTV (output only) and VideoSpigot (input only) are inverses of each other and far apart in time, market focus, company mergers and engineering efforts.
VideoSpigot LC (1992) is for video input/capture only and is a PDS variant of the original VideoSpigot card.
Conceptually, SimplyTV (output only) and VideoSpigot (input only) are inverses of each other and far apart in time, market focus, company mergers and engineering efforts.
Last edited:
Curious Cards indeed. What the heck is up with the form factor of SimplyTV? Is that thing really supposed to fit in the pizza box LC-Q605 case or was it developed specifically for the LC520 and up? I've never seen a SuperMac product (other than the ROM deficient VideoSpigot LC) so wildly out of Apple spec. Was that card picked up from a company who never bothered to read the DCaDftMF series or entirely missed/ignored the specifications for LC expansion cards?
Here's my Spectrum/8 Series III. If anyone has a good way to photograph cards of this size with better angles or such, I'd be more than happy to take more.
