Jump to content
Trash80toHP_Mini

UNDERClocking a Video Card??????

Recommended Posts

rcp2iisicrystalcan2post.jpg

KnuckleDraggin'HardwareHacker's Brute Force Approach to a, crystal can swap, down-clock of 16" 75Hz Fixed Res Output to a 60Hz LCD friendly Output solution for the RCP2IIsi Card.

 

I've also got a pending request for VGA 640 x 480 x 8 bit @ 60 Hz.

 

Any suggestions from my electron-pushin' comrades? :?:

 

This is a call for help in the IIsiColorPivotII_PDS_Card_HackProject™ thread.

Share this post


Link to post
Share on other sites

Just a thought: Some cards had jumper settings to compensate for different I/O bus speeds of different machines. Idk if this card does; but if it does, you might be able to set the jumpers for a higher bus speed, thus increasing the divider and down-clocking the chip.

Share this post


Link to post
Share on other sites

Thanks for the thought, but historically, your theory makes sense for the PC VidCards, but probably not for Macs. That said, unless one of the two jumpers on this particular Pivot Card might be compensation for the differential between the IIsi's and SE/30's bus clocks . . .

 

HRMMM!!!! another test in the works, my ASSumptIon was that they were both for Pseudo Slot addressing, as are the two jumpers on the MacCon NIC.

 

The Color Pivot II Cards were made for specific monitors to be used on specific Macs running at specific bus speeds in their PDS implementations. This was way back in the Neolithic, if not the Paleolithic DTP ERA which drove the development of FPD and TPD monitors for less expensive multi/large monitor workstations than those of the Jurassic 's Mega$ MiniComputer CAD License "seats" of the day.

 

The IIsi/SE/30 and the LC slots . . . :-/

 

. . . dunno, good point, none of many LC or IIsi/SE/30 Cards I've tried over the years have needed bus speed compensation, IIRC.

 

Maybe the Crystal Can on this card is independent of the PDS bus speed . . . I don't know if this makes swapping crystals more or less viable a solution. :?:

 

Hey! Electronics & Clock Chipping Boffins, how 'bout a little help here! :o)

Share this post


Link to post
Share on other sites

iCrap!!!!!! It looks like I found a second crystal . . . is that what CR1, the three legged thingamajig next to the cable connector is?

rcp2iisicr12p0st.jpg

 

 

If CR1 a crystal, these tiny bast . . . . erm . . . components . . . marked CR are probably crystals too! :-/

rcp2iisitinycrystals2p.jpg

 

I really need help here . . . :I

Share this post


Link to post
Share on other sites

Thanks guys, that takes a load off.

 

My electronics development days/education ended at about the time this card was made, it topped out at the CMOSS Cookbook Second Edition. The crystal can and ROM are about the only components similar to what we worked with back in the day. I just spec'd the project, did the PCB layouts, made the PCB prototypes and RubyLith patterns for the PCB House. I hardly even touched a soldering iron.

 

Now if somebody could noodle out what crystal to replace the stock one with to underclock this card to give 60 Hz LCD outputs . . . or tell me it's not gonna work at all, I can drop this line of investigation.

 

I'm cool with the card as is, but they'd be a LOT more useful to the rest of the gang if we can get 'em to be more LCD friendly! :approve:

 

edit: [:D]]'> me likes the auto-linking to WikiWhat'sIs! Interesting that the underclocking page link hadn't turned up yet in either thread!

Share this post


Link to post
Share on other sites

Hey Trash!

 

I've got some bad news for you. I was reading about interference generated at 14.318MHz - and noticed that it's very common around NTSC equipment. Why, you ask, is this important?

 

14318180Hz is exactly four times 3579545Hz - the NTSC Chrominance encoding reference. (two signals that are 90° out of phase, but I digress...)

 

The good news? You don't have to desolder that can.

 

The bad news? You need to look elsewhere for how the card generates the vertical synchronisation signal - that can is just a colour timing reference thingy. You might find some manner of phase-locked-loop circuitry on the card to interpolate a higher frequency from the 14.318MHz - that could be a starting point.

 

Let us know how you go, if I can think of anything regarding this I'll be sure to pipe up.

Share this post


Link to post
Share on other sites

The bad news? You need to look elsewhere for how the card generates the vertical synchronisation signal - that can is just a colour timing reference thingy. You might find some manner of phase-locked-loop circuitry on the card to interpolate a higher frequency from the 14.318MHz - that could be a starting point.

 

--------------------------------------- GAME OVER! ---------------------------------------

 

That's waaayyyyyyyyyyyyyyyyy outta my league! :disapprove:

 

A couple of possible leads:

 

Could it be done in the BT478KPJ50 - Digital Analog Converter, Triple, 8 Bit, 44 Pin, PLCC?

 

The Dallas DS1010-100 - Clock/Timing - Delay Lines IC sounds promising to this electronics noob.

 

Dunno, there are a few ICs that aren't Radius labeled ASICs, but not many . . . :-/

 

 

I'll post HI-Res pics of the board's surfaces for all of you electronics proficient type comrades to take a look at.

If somebody wants to tackle the project in hardware, I know where I can find another card and get it to you. :approve:

Share this post


Link to post
Share on other sites

Maybe I'm a bit late, but...

 

Don't give up! Unless your card is able to make an actual composite NTSC signal (instead of the common RGB output with NTSC timing on many Macs) that 14.318 MHz crystal won't be used as the NTSC Chrominance encoding reference...

 

More good news: the official dot clock for 16" mode (832x624 @ 75 Hz) is 57.2832 MHz -- divide that by 4, and you get about 14.32 MHz, which is very very close to your crystal. That slight underclocking (less than 0.02%) won't affect any monitor at all, and allows the use of a very common (read: cheap) crystal. Besides its use in colour TVs, 14.318 MHz divided by three makes... 4.77 MHz. Does that sound familiar? ;)

 

So, we can assume that crystal is the timebase for 16" at 75 Hz refresh... in order to underclock for 60 Hz refresh, you want a crystal 0.8 times (60/75) the frequency -- something like 11.45 MHz, a bit higher or lower won't harm, just check what's available.

 

PLEASE NOTE: I haven't actually tried this, so act at your own risk. Anyway, in case it won't work as intended, I think it's very unlikely that it would damage anything.

Share this post


Link to post
Share on other sites
So, we can assume that crystal is the timebase for 16" at 75 Hz refresh... in order to underclock for 60 Hz refresh, you want a crystal 0.8 times (60/75) the frequency -- something like 11.45 MHz, a bit higher or lower won't harm, just check what's available.

Cool! I just found your post now and this is definitely worth a try! Thanks for the suggestion.

 

I'd like to see a 960x600 60Hz resolution. Exactly 1:2 pixel ratio of my 1920x1200 display.

That's not gonna happen, what you'll get will be 832 x 624 x 8bit x 60Hz LetterBoxed within frame or interpolated to full height and letterboxed on the ends on most LCDs that are Mac fixed res/frequency hostile.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×