Subject: Re: Off topic macppc hardware help
To: John Klos <john@ziaspace.com>
From: Jeff Walther <trag@io.com>
List: port-macppc
Date: 03/14/2003 21:39:39
--============_-1164422908==_============
Content-Type: text/plain; charset="us-ascii" ; format="flowed"

At 22:03 -0500 03/14/2003, John Klos wrote:

>Because the 604ev is a smaller mask than the 604e, it could go faster,
>draw less power, and therefore it took a different voltage. While the 8600
>and 9600 which came with the 604ev can take other CPU cards, the 604ev
>cards won't work in a standard older motherboard. However, the only
>difference between the motherboards, I am told, is that it makes a
>different voltage available to the CPU (perhaps in addition to the 3.3
>volts), and that it does not come with a socket for 50 MHz motherboard L2
>cache.
>
>So, if someone has some Apple docs which talk about the 604ev systems,
>perhaps the pinout of the card connector, maybe I can manage to set up a
>regulator in line to provide what it needs and use the 350 MHz with 1 meg
>of L2 in place of the 200 MHz.

This is something I have wondered about for some time, but I do not 
have a Kansas motherboard, nor a Mach V card available on which to 
experiement.

Attached is the pinout for the standard Mac CPU card.  Starting with 
that it shouldn't be too difficult to identify the differences in the 
Mach V cards.   After all, regular cards work in the Kansas machines 
(at least, third party upgrades mostly do) so most of the things like 
the address and data lines cannot be different.   This leaves few 
candidates for change.

The card has two levels of connectors and the attached file only 
identifies the connections on the upper level of pins (pins 1 - 146). 
The lower level (147 292) is 3.3V on the first six pins and then GND 
across the rest of the card.    It's usually obvious which six pins, 
as they're often merged into a single line of copper, so there will 
be a short length of copper, a break and then a long length for the 
the GND pins.

If your card has individual pins connectors on the lower level, a few 
minutes with an ohmmeter will tell you where the six 3.3V pins are on 
each side of the card, as they're all common to each other.

I would look at pin 82 as it is an NC on the standard CPU card and so 
a good candidate for additional power connections.  However, rumor 
has it that installing a Mach V in a regular motherboard can damage 
the card and the MB, so if those rumors are true pin 82 is a less 
likely candidate.

So, you may also wish to test the secondary arbitration pins.   Those 
are there solely to provide the arbitration signals to a second CPU 
on dual processor cards.  If Apple had no intention of building dual 
processor Kansas machines, they might have used them.    I don't 
think they did because I believe the dual processor G4 cards a few 
manufacturers abortively released did work on Kansas machines. 
Anyway, those would be pins 56 - 58, SBR, SDBG and SBG.   You can 
check to see if those pins connect to pins 121, 135 and 141 on 
Hammerhaed.   If they do, then they're still providing secondary 
arbitration signals.

I guess the other candidates might be some of the 5V pins on the 
regular card ( 1 - 6 and 74 - 79)  Maybe one or more of those was 
diverted to another voltage on the Mach V and that's why putting a 
Mach V card in a regular PowerSurge motherboard can cause problems. 
These are easy to check as there should be direct connectivity (0 
resistance) to the 5V power supply connector pins.   Other easy test 
points are the 5V pins in the PCI slots.

Please let us (or at least me) know what you find.

Jeff Walther
--============_-1164422908==_============
Content-Id: <a04310100ba984d95938f@[199.170.89.48].0.0>
Content-Type: multipart/appledouble; boundary="============_-1164422908==_D============"

--============_-1164422908==_D============
Content-Transfer-Encoding: base64
Content-Type: application/applefile; name="%CPU_Slot_Pinout"
Content-Disposition: attachment; filename="%CPU_Slot_Pinout"
 ; modification-date="Fri, 27 Dec 2002 01:24:55 -0600"
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--============_-1164422908==_D============
Content-Type: application/octet-stream; name="CPU_Slot_Pinout"
Content-Disposition: attachment; filename="CPU_Slot_Pinout"
Content-Transfer-Encoding: base64
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=
--============_-1164422908==_D============--
--============_-1164422908==_============--