-
Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.
Help for a nubus Radius PrecisionColor Pro 24X ROM/firmware upgrade
- Thread starter Classicman
- Start date
Phipli
Well-known member
I think that is a 24XP to be honest. I'll double check.
Edit - yup, it is a Radius PrecisionColor 24XP. It has 1.5MB of VRAM and does up to 832x624 resolution.
Phipli
Well-known member
The jumper isn't useful for end users. It isn't documented, but it is connected to a pin on the EEPROM that doesn't do anything. It is probably from development, or an option for things they never did, such as larger ROMs.
Professor 362
Member
Hello Phipli, Thanks for your reply. I am humbled by the span of your knowledge, as you have weighed in on many of my questions. But I am confused by some conflicting info I am seeing. 1. in the thread (linked below) a card identical to the one I have is identified as a Pro 24X. 2. If it is a 24XP, shouldn't it support higher resolutions than 832x624 at 24bit? Gamba's site suggests. Clipping attached... not sure what the solid lines mean.
What's the difference between the various Radius 24 cards?
So I happen to have a Radius 24X, 24XP, and 24XK. According to LEM, the 24XP appears to be the budget card. It doesn't do the higher 1152 resolution, though it it's still 24-bit and accelerated. The 24X and the 24XK, on the other hand, I can't tell the difference. I tried both on my IIci...
68kmla.org
Attachments
Phipli
Well-known member
I think you're misreading the table? Your card is the Pro version, and there is an error on the non Pro row.Hello Phipli, Thanks for your reply. I am humbled by the span of your knowledge, as you have weighed in on many of my questions. But I am confused by some conflicting info I am seeing. 1. in the thread (linked below) a card identical to the one I have is identified as a Pro 24X. 2. If it is a 24XP, shouldn't it support higher resolutions than 832x624 at 24bit? Gamba's site suggests. Clipping attached... not sure what the solid lines mean.
What's the difference between the various Radius 24 cards?
So I happen to have a Radius 24X, 24XP, and 24XK. According to LEM, the 24XP appears to be the budget card. It doesn't do the higher 1152 resolution, though it it's still 24-bit and accelerated. The 24X and the 24XK, on the other hand, I can't tell the difference. I tried both on my IIci...
All of these cards use the same PCB and look absolutely identical, but with a different grade RAMDAC (big chip near the video out port) and amounts of VRAM. The board colours don't imply the card model.
Yours has a 66MHz RAMDAC chip which isn't rated fast enough to do 1024x768 or 1152x870. I haven't looked the VRAM chips up today but only one of the board models was fitted with the 66MHz RAMDAC, and that was the 24XP (and 24XP Pro which is basically identical), which is a 1.5MB board. I've previously done a load of work on these cards. My dad has a 24XP like yours. But blue.
Last edited:
Professor 362
Member
Thanks so much. Kind of disappointing news... I was/am not sophisticated enough to know about RAMDAC chips. I thought it would be a 24 bit card up to 1024 or even 1152. I paid $130... probably too much. I really wanted to drive at least 1024 on my M2494 Multiple Scan 17 Trinitron. I thought it would be more like this Radius card in my other computer... Quadra 700 (pictured)... but that appears now to be a superior card. THANKS!
Attachments
Phipli
Well-known member
That card (24X) is a very good card.
They're both as fast as each other and 832x624 is quite a good resolution. The 24XP is still a nice card and given its speed and that it is an accelerated 24bit card, I don't think you really overpaid compared to some other recent listings.
If you get adventurous, you could make a 24AC ROM (about $2) using a ROM programmer (about $40) for your Quadra 700 card and swap it in. It would then be possible to run it using Apple's version of the drivers, and it would also work accelerated alongside a PPC 601 Upgrade card. With the 24x v2.0 ROM it works with the PPC cards, but not accelerated.
Professor 362
Member
Very Interesting! Thanks for your information.
eharmon
Well-known member
Oh interesting, I didn't know the Radius drivers wouldn't accelerate it. If I get my 601 card working again I'll give that a try on my 24X.
Phipli
Well-known member
I'm not sure, there might be two versions of the card? An Apple and a Radius 24AC? LEM says there is, but LEM is often wrong, and I have never found a 24AC ROM that isn't from an Apple sold card. Plus the "Apple" cards have a Radius copyright and a Radius PCB.
Basically, that could be true, but it isn't quite what I meant. I meant the 24AC ROM I have a copy of uses an Apple provided driver. Probably still written by Radius.
Basically, that could be true, but it isn't quite what I meant. I meant the 24AC ROM I have a copy of uses an Apple provided driver. Probably still written by Radius.
slomacuser
Well-known member
True. There is no Radius PrecisionColor 24AC listed anywhere on Radius web page or any price list anywhere advertised. It was just a mistaken by Apple 24AC card which was made by Radius, only a back label identified it as Apple Display Card 24AC, see picture, front and back:
Where is this info from? For System 7.5 and Powermac compatibility Radius recommended Dynamic Desktop version 3.3 or later, QuickColor 3.3 or later, and ROM 2.0
Phipli
Well-known member
Personal testing with about 30 tests of different cards, bit depths processors and OS versions.
ROM version 2.0 works with PPC, in that it doesn't crash, but it runs at the same speed as the 24AC does /without/ drivers. This means it is unaccelerated.
Phipli
Well-known member
Are you sure there is a quickcolor 3.3? Do you mean RadiusWare 3.3? Don't trust LEM.
Edit - no there is. I'll double check what version I was using in the future.
Phipli
Well-known member
I was using version 2.62 of the Control Panel it seems. I'll try a newer version and see if acceleration works with V2.0 ROMs and PPC.
eharmon
Well-known member
I also wonder if there's any difference in performance with the 24AC v. 24X ROM on PPC assuming acceleration works. That is, did the acceleration drivers diverge between what shipped in the OS versus Radius?
Phipli
Well-known member
They were identical under 68k, so I suspect not.
I'm confusing myself because I just booted my IIci with my 24XK, forgetting that it is flashed with the 24AC ROM
QuickColor didn't recognise it because it is an "Apple" card. Apparently.
I have my jumper mod on the card so I'll switch it back over when I power down next.
Phipli
Well-known member
@slomacuser
You're right it is accelerated with 3.3. see results for all of my tests in my IIci below.
You're right
it is accelerated with 3.3. see results for all of my tests in my IIci below.They're basically the same (OS version does make a difference, with OS8.5 being the fastest - don't ask why my IIci has 8.5 installed, I got carried away one day).
Phipli
Well-known member
@lemketron
I have a couple of cheeky questions that I really hope you don't mind my asking. I didn't want to hijack the other thread, and realise many years have passed. But...
1 - Do you happen to know the purpose of the resistor straps I've circled as "1" in the following photo? They do seem to change between the lower end cards and the high end cards. Are they to do with VRAM organisation? Or something else?
2 - The highlighted jumper (marked "2")seems to connect and disconnect one of the pins on the ROM to the left most ASIC. The pin is either a write enable pin, or an address pin depending on what size ROM you have fitted (I've been fitting a larger ROM and rewiring this jumper slightly so I can manually bank the ROM between say, the PrecisionColor 24X and 24AC ROM, so I can experiment with the Apple driver or the Radius driver (they're the same speed, I'm sure you knew!)), but I was wondering what it was originally for.
3 - What... cough... are your thoughts on overclocking the Bt473? Or the second source part from Analog Devices? Did you find that an 80MHz part would happily run at 100MHz?
It would be nice to open up some higher resolutions at lower bit depths on some of the lower end cards if the RAMDAC is happy to overclock. Although I think I saw someone having issues where it looked like the cards with less VRAM had issues even if you had enough VRAM for a resolution + colour depth, you'd end up with VRAM being written to twice due to the chips responding to writes to higher addresses being generated by the ROM from a card with more VRAM. I assume because it was using the higher ranges of the lower order chips first, rather than filling all chips equally. It would be a lot of soldering to replace the VRAM!
I have a couple of cheeky questions that I really hope you don't mind my asking. I didn't want to hijack the other thread, and realise many years have passed. But...
1 - Do you happen to know the purpose of the resistor straps I've circled as "1" in the following photo? They do seem to change between the lower end cards and the high end cards. Are they to do with VRAM organisation? Or something else?
2 - The highlighted jumper (marked "2")seems to connect and disconnect one of the pins on the ROM to the left most ASIC. The pin is either a write enable pin, or an address pin depending on what size ROM you have fitted (I've been fitting a larger ROM and rewiring this jumper slightly so I can manually bank the ROM between say, the PrecisionColor 24X and 24AC ROM, so I can experiment with the Apple driver or the Radius driver (they're the same speed, I'm sure you knew!)), but I was wondering what it was originally for.
3 - What... cough... are your thoughts on overclocking the Bt473? Or the second source part from Analog Devices? Did you find that an 80MHz part would happily run at 100MHz?
It would be nice to open up some higher resolutions at lower bit depths on some of the lower end cards if the RAMDAC is happy to overclock. Although I think I saw someone having issues where it looked like the cards with less VRAM had issues even if you had enough VRAM for a resolution + colour depth, you'd end up with VRAM being written to twice due to the chips responding to writes to higher addresses being generated by the ROM from a card with more VRAM. I assume because it was using the higher ranges of the lower order chips first, rather than filling all chips equally. It would be a lot of soldering to replace the VRAM!
