Thread
Unknown card in PDS slot
It's not 030 to 040. It's 040 33mhz to 50 MHz 040 , it needs a driver.
It's usually an init and/or a control panel to activate all the options.
It's usually an init and/or a control panel to activate all the options.
At least you have a chance of finding it, my Plus's video board doesn't even appear to have ever existed. 
I'm talking about the Radius16 I had in the SE's PDS and PowerCache Cards, all of which require drivers to tell the CPU to take a nap/control panels to turn them on and off and to tweak settings like cache and Math processes.. In the 68k World, the only processor card upgrade that would not need a driver that I can think of offhand would be for the Blackbirds. I've never gotten my BlackBird bits and pieces up running, but it appears to be the same plug-n-play deal as the G3 ProcCard Upgrade in my 1400c.
edit: PDS is not a substitute for the native processor, it's full access to that processor's I/O at the rate of the system clock for all but a couple of oddball cases like the IIfx and the Portable.
Are we still absolutely postive that's what it is? I'm serious about not being 100% sold on that. Does *anyone* have another example of a '040 "upgrade" that went into the PDS slot of a Quadra? There were PPC accellerators that fit in the PDS slot, but I haven't seen any evidence that they made 68040 cards for it, and for good reason: any such card would be *barely* faster than the built-in CPU. (And heck, even if you did think it was worth the effort, I'd think a better way to design it would be as a clock-multiplying daughterboard that plugs into the CPU socket; every desktop '040 machine has a socketed CPU and that was standard fare for 486 upgrades.) It still seems to me this *could* be some sort of math coprocessor/DSP/"Photoshop Accellerator" card.
I don't want to suggest prying the heat sink off to see what that chip really is, but I'm still saying it would be a worthwhile exercise to count pins on the socket (it looks like it's through-hole so the pincount should be visible from the back) and verify that it's at least the same socket as a 68040.
No need for drivers to find out what it might be. Download and run TattleTech, the PDS/NuBus report should give you the information about the class of the card and maybe the Rev.Level of the firmware right from from ROM on board. Don't know that I've ever checked out an Accelerator with TattleTech, but it's a good bet.
PPC is a possibility, but I've never even seen a heat sink on a DSP, much less active cooling. Proc card upgrades for the 486 didn't entail overclocking the crap out of a standard part. Powering the fan from the PDS power budget makes sense to me, those wires look fairly hefty. Running a 25% overclocked 68040 (RTFM skipping bozo field installation) without the fan hooked up = imminent catastrophic failure.
Running the 40MHz part at 50MHz would be a 25% speed bump for the 840AV and a 51.5% bump for the stock Q650's 33MHz clock and 100% for the Centris version, IIRC. Overall system performance ought to be very noticeable even though system performance/clock bump is never proportional, even on an 840AV.
Just sayin' :
PPC is a possibility, but I've never even seen a heat sink on a DSP, much less active cooling. Proc card upgrades for the 486 didn't entail overclocking the crap out of a standard part. Powering the fan from the PDS power budget makes sense to me, those wires look fairly hefty. Running a 25% overclocked 68040 (RTFM skipping bozo field installation) without the fan hooked up = imminent catastrophic failure.
Running the 40MHz part at 50MHz would be a 25% speed bump for the 840AV and a 51.5% bump for the stock Q650's 33MHz clock and 100% for the Centris version, IIRC. Overall system performance ought to be very noticeable even though system performance/clock bump is never proportional, even on an 840AV.
Just sayin' :
The Sonnet Quaddoubler was an overclocked 50mhz upgrade that stuck into the CPU socket, so if you're going to do something stupid like this you don't *need* to use the PDS slot...
Actually, interesting point here: I don't see a crystal on that board, and the board was in a Q650. The PDS slot in that machine is going to be running at 33mhz. The QuadDoubler doesn't need a crystal because it derives its clock from the system clock simply enough by being phased 2:1 with it (which is why it only works in 25mhz '040s), if this runs at 50mhz it's going to have a much rougher time deriving the clock from the 33mhz signal it's getting from the motherboard. And... on top of that, the performance is going to be *lousy* because there would be a non-integer relationship between the card CPU's bus cycles and the memory bus, meaning you're going to be getting a buttload of wait states. (Unless the card magically switches to 33mhz when doing external bus cycles and only runs at 50mhz when it's not doing anything external... which I think is going to be "never" on an '040 because of the onboard cache doing line fills all the time?)
Again, totally not going to rule out it being an '040 acellerator, but it's a very weird one if that's what it is.
Just wondering about the release/ship date comparison between the 40MHz part and the 25/50MHz part. The latter makes sense in a Centris.68LC040 limited system's PDS, especially if the former was released before the clock doubled latter. Twice the clock and full '040/CoPro makes a world of sense as a Centris->Quadra upgrade card. Dunno, gotta find the Accelerator Reviews.
Just finished doing a space saving installation of the wall hanging shop vac and overall cubic use efficiency upgrade in my Delta14 garage/apartment kitchen pantry. Time to look for the ScanDump box as a reward! [
] ]'>
Just finished doing a space saving installation of the wall hanging shop vac and overall cubic use efficiency upgrade in my Delta14 garage/apartment kitchen pantry. Time to look for the ScanDump box as a reward! [
] ]'>SUCKER! There's no such thing as a "25/50mhz" part.
The 25/50:33/66:40/80:50/100 "thing" as applied to Motorola's 68040 CPUs was entirely sleezy marketing, as I mentioned above. One technical difference between the 80486 and 68040 CPUs is the Intel CPUs clocked their internal bits at the same rate as the input clock, while the Motorola CPUs internal cadence required a clock input 2x the desired speed. Therefore a "33mhz" 80486 uses a 33Mhz input clock while the equivalent 68040 uses a 66mhz clock. This doesn't mean squat in terms of performance as no part of a 68040 that actually performs calculations "runs" at the double speed; in fact, Intel's *previous* design, the 80386, also had a 2x relationship between the input clock and the effective ALU speed. So...
When Intel came out with the *genuine* clock-multiplying parts, IE, the "DX2" CPUs, Apple had nothing to counter that with because, again, Motorola NEVER MADE AN EQUIVALENT PART. However, some sleezy marketing guy happened to notice that "double-speed" input clock and decided to start pasting it into advertising literature, and likewise of course the upgrade makers latched onto it as well. IT'S PURE HOKUM, and everyone involved should be ashamed of themselves. A "25/50mhz" 68040 is just a 25mhz 68040, period, and is *not* anything like a 50mhz DX2, which *legitimately* runs important parts of the chip, IE, the ALU/FPU/cache/etc, faster than other parts, IE, the bus drivers.
If Motorola *had* made real double-clocked 68040s then the "Quad Doubler" wouldn't exist; note that it has circuitry between the CPU and the socket that supplies the clock-doubling and bus arbitration functions. Similar socket upgrades for Intel 486s usually had nothing on them but voltage regulators, and that was only necessary if the upgrade incorporated something like a clock-tripling DX4 CPU. (Which used a 3.3v line for internal logic instead of 5v.)
HEH! :lol: Ancient history, but sleaze still oozes out of marketing groups wherever they can be found!
P.T.Barnum's spirit is alive, well, hiding under rocks everywhere and nigh on inexterminable . . .
. . . now if we could just set the nails in Murphy's coffin. }
P.T.Barnum's spirit is alive, well, hiding under rocks everywhere and nigh on inexterminable . . .
. . . now if we could just set the nails in Murphy's coffin. }
I think the OP said this was in a Centris 650. Didn't the Centris have a 25MHz bus? The Quadras were pretty much identical to the Centrises but with a 20 to 25 MHz bump for the 610 and a 25 to 33 bump for the 650.
Yep, according to Everymac.com, the Centris 650 has a 25MHz bus/CPU and the Quadra 650 has a 33 MHz bus/CPU. So this could be an early clock doubling CPU upgrade, but I agree with Gorgonops, it would be worth at least turning it over and counting the pins. Also, check along the top edge of the heat sink.
The 68040 has its part number printed on the top edge and often, folks would attach the heat sink so as to leave that labeling visible.
As to whether a PDS CPU upgrade **needs** an extension/control panel or not, I had a conversation in which I was told that one of the early daughter board upgrades on the 128K/512K used some trick in bus arbitration to shut down the host CPU, but in that case, the upgrade has full access to all of the original CPU's pins. So I could go either way. Sure seems like the Turbo040 and such are active whether you load Daystar's software or not, but my memories are old and crumbly.
The cpu has 179 pins so seems it is to be an 68040.But I haven't found any details on the internet yet.I even mailed Formac for details without any succeed answer...
My SE accelerator needs software to use, it just ignores the card without the software
OP says Quadra 650 @ 33MHz
Daystar's Universal PowerCache Install Manual says you've gotta install Power Central in the System Folder. Dunno, I've used several PDS Accelerators and Cache Cards from the 68000 on and all needed an INIT to shut down the native proc. to work. 2x CPU active on one I/O bus = BAD Juju, but this Card is definitely an odd duck.
ScanDump Box tearsheet search turned up bopkes. I've got various MW, MU & Byte Accelerator Review Articles from 91 - early 93, so nothing for the 650 and no products mentioned from Formac. No Formac listed in the Advertiser Directory of the May '94 PowerMac Intro, Macworld issue, nothing in the Star Ratings and I didn't see anything in the Mail Order House ads for accelerators.
If I find the intact Magazine collection, maybe something will turn up, but right now I got nuthin! :mellow:
edit: once again, ROM spelunking with TattleTech highly recommended. [
] ]'>On my SE Sytem Profiler detects the CPU part of the Hypercharger 020 PDS upgrade fine, no extension needed. If I were the op I'd try and find something that can detect the speed of the processor and run that, see if maybe the card does run extensionless.
Failing that something which lists devices on the PDS would work.
Or, load a heavy program while timing it, shut down, pull the card, reboot, load the same program and see if it takes longer?
So... maybe it is something like a QuadDoubler stuck in a PDS slot instead of the CPU socket, but that raises the other question again: are you *positive* your machine is a 33mhz Quadra 650 instead of a Centris 650? There is almost *no way* a 68040 is going to tolerate an overclock up to 66mhz (remember, even the "50mhz" ones used in the Quad Doubler were actually overclocked 40mhz marked chips) so if that is something like a PDS QuadDoubler it was almost certainly meant for machines like the Quadra 700/900, Centris 610/650, etc. If you run something like Speedometer 4.0 on your system does it correctly identify it as a Q650 and does it benchmark as expected? (Both with and without the card installed?)
(Now I wonder if it's possible, given this has a ROM chip on it unlike the QuadDoubler, if this card is trixy enough to detect if it's plugged into a compatible Mac and disables itself if not?)
I'm not sure it'd fit in the 700's PDS slot, would it?
Interesting, Tattletech tells both tales. ASP(?) or Tattletech would poll PDS devices and detect the Proc in your SE and the classification of the card, but that doesn't necessarily mean it's the active processor does it? From your advice, I assume you've noticed the acceleration.
I was under the vague impression that all 68040 "real" Quadras (IE, excluding the LC based ones like the 605 and 630/640) had the same PDS slot, but I can't say definitively.
I meant physically, the slot is in a bad spot, I think it'd interfere with the case.
The machine is a Quadra 650 for sure.(68040/33 and I've tested with several utilities like Speedometer,Apple Personal Diagnostics) The card it is very strange,thats for sure.We need a German guy who has old local Mac magazines.The card has a date on it. 7-3-1994
Looking at the photos, it seems like the main ceramic chip is socketed, if so, can you pull it out? sometimes they write some details on the bottom as well
Also, it seems that the other socketed chips on the card are AMD Flash erasable PALs. what are they doing? why so many of them?
Also, it seems that the other socketed chips on the card are AMD Flash erasable PALs. what are they doing? why so many of them?
Yeah, that's a possibility; I was thinking they're *electrically* the same because things like DOS cards seem to be mostly shuffleable between them but mechanically they differ. (Like the Quadra 900/950 can take that absolutely huge WGS95 card that might not fit in any other model.)
I'd like to reiterate my suggestion of checking the top edge of the chip which is not covered by the heat sink. The chip label might be high enough that it's not covered by the heat sink. It's hard to tell how much lap there is from the photo.
The name on the ROM, "proquadra 50" does seem to indicate a 50 MHz 68040 upgrade. No one ever actually shipped a 68060 upgrade, right?
The Centris 650 was easily upgradable to a Quadra 650 equivalent by moving a few resistors....
The name on the ROM, "proquadra 50" does seem to indicate a 50 MHz 68040 upgrade. No one ever actually shipped a 68060 upgrade, right?
The Centris 650 was easily upgradable to a Quadra 650 equivalent by moving a few resistors....
You can change what it identifies as, but don't you have to swap a crystal too? (IE, I know if you fiddle with the ID bits you can make the board identify as anything from a Centris 650 to a Quadra 800, but the speed is still dictated by the oscillator.)
Again: Download TattleTech and run the NuBus/PDS report to find out what you've got. No drivers or chip pulling necessary, it polls the onboard ROM, identifies the CPU a/o determines the class of the card: accelerator, NIC, Graphics card etc.
If you can't find that, download the (freeware) Gauge Utilities and run SlotInfo!
Way too much guessing going on here, no utilities have been employed per long established SOP this kind of thing: :
See: Peripherals: LINKS Project v2
If you can't find that, download the (freeware) Gauge Utilities and run SlotInfo!
Way too much guessing going on here, no utilities have been employed per long established SOP this kind of thing: :
See: Peripherals: LINKS Project v2
Agree 100%, Trash. It's time for the OP to run some software probes/diagnostics.
Formac had a few mentions in UK Mac magazines in the 1990s but I don't have many. MacUser UK 6 Jan 1995 mentions an older accelerator for the LC family. It was the Formac Pro 50 (68030 50Mhz).
Formac had a straightforward naming convention. ProNitron = Trinitron monitor. ProDrive = SCSI HD. ProGraph = PDS graphics card. ProOpt = SCSI MO drive. ProVision = PCI graphics card. ProLegend = NuBus graphics card.
Formac had a few mentions in UK Mac magazines in the 1990s but I don't have many. MacUser UK 6 Jan 1995 mentions an older accelerator for the LC family. It was the Formac Pro 50 (68030 50Mhz).
Formac had a straightforward naming convention. ProNitron = Trinitron monitor. ProDrive = SCSI HD. ProGraph = PDS graphics card. ProOpt = SCSI MO drive. ProVision = PCI graphics card. ProLegend = NuBus graphics card.
I just have a question: How does the machine act with and without the accelerator?
I'm saying that because of my DayStar 040 in my IIci. When in the PDS slot, you turn on the machine and the little Smiling Mac Icon comes up with "Turbo 040" on its belly and the memory test only last a few seconds- LITERALLY! Without it, the Smiling Mac icon comes up as it normally does and the memory test (for 20megs of RAM) takes over 3 minutes!
I'm thinking along those lines - if the Formac is in the PDS Slot, it should activate without software, though there should be a control panel to control a few options like the 040 cache. (BTW: Apple's 040 Cache control panel does work on the DayStar 040.)
As to why one would want to put in a 50MHz 040 into a 33MHz 040, the answer if simple - for the work being done, faster is better. Things like multimedia (audio/video/images) are processed a lot faster on 50MHz than it is on 333MHz, enough to make or break a deal.
I'm saying that because of my DayStar 040 in my IIci. When in the PDS slot, you turn on the machine and the little Smiling Mac Icon comes up with "Turbo 040" on its belly and the memory test only last a few seconds- LITERALLY! Without it, the Smiling Mac icon comes up as it normally does and the memory test (for 20megs of RAM) takes over 3 minutes!
I'm thinking along those lines - if the Formac is in the PDS Slot, it should activate without software, though there should be a control panel to control a few options like the 040 cache. (BTW: Apple's 040 Cache control panel does work on the DayStar 040.)
As to why one would want to put in a 50MHz 040 into a 33MHz 040, the answer if simple - for the work being done, faster is better. Things like multimedia (audio/video/images) are processed a lot faster on 50MHz than it is on 333MHz, enough to make or break a deal.
Hell I'd be happy with a 50mhz overclock today on my Quadra 650 let alone back in 1994 when it was still relevant.
Well, as I've mentioned before, my memory is old and crumbly. Does the 25 => 33 upgrade require an oscillator swap? Some of them just needed a resistor change on a clock generator, but the Centris/Quadra may still have been oscillator controlled.
Okay, I went and looked at the mirror/preserve of Marc Schrier's Clock Chipping Home Page hosted at applefool, and it was indeed an oscillator swap. Thank you.
So, an oscillator swap and a couple of resistors....