Upgrading Ipaq C500

P

paulmd

We were gifted a whole boatload of these cute little computers. They're
suprizingly easy to work on, if you happen to have a tamper-resistant
torx screwdriver.

I was wondering, what the max processor for an Ipaq C500 Is? They start
out life as a 500mhz celeron, but with a Bios upgrade, I've gotten one
of them to a 1Ghz p3 (133/256/1.7v), and another to 1100Mhz
(100/something/1.7v) celeron. (Yes, I upgraded the 'sink to an active
cooling model, the board comes with 2 fan headers). I'm wondering if
these things actually do have a limit in the socket 370 world. So far,
no crashing.

I happen to have some faster celerons, up to 1400, in socket 370, but
they're 1.5v. I suppose I could risk killing one (one has a missing
panel anyway), but I thought I'd ask if anyone's tried.
 
K

kony

We were gifted a whole boatload of these cute little computers. They're
suprizingly easy to work on, if you happen to have a tamper-resistant
torx screwdriver.

I was wondering, what the max processor for an Ipaq C500 Is? They start
out life as a 500mhz celeron, but with a Bios upgrade, I've gotten one
of them to a 1Ghz p3 (133/256/1.7v), and another to 1100Mhz
(100/something/1.7v) celeron. (Yes, I upgraded the 'sink to an active
cooling model, the board comes with 2 fan headers). I'm wondering if
these things actually do have a limit in the socket 370 world. So far,
no crashing.

I happen to have some faster celerons, up to 1400, in socket 370, but
they're 1.5v. I suppose I could risk killing one (one has a missing
panel anyway), but I thought I'd ask if anyone's tried.


Occasionally I'd tried them (Tualatins) without success on
non-Tualatin "supportive" boards but nothing was damaged in
the exercise. Sometimes a simple pin-adapter socket adapter
was enough to get them to work but other times nothing would
(typically on Sis 6?0 chipsets the pin-adapter wouldn't
work).

Hard to say if Compaq did anything limiting with their bios
but in general a bios that supported Coppermine would run a
Tualatin.

I'd be more concerned with overall lifespan as related to
heat and power... would put the fastest P3 w/133MHz FSB
(presuming the system can do 133MHz FSB) that it can run
with vcore limited to <= 1.6V.
 
P

paulmd

kony said:
Occasionally I'd tried them (Tualatins) without success on
non-Tualatin "supportive" boards but nothing was damaged in
the exercise. Sometimes a simple pin-adapter socket adapter
was enough to get them to work but other times nothing would
(typically on Sis 6?0 chipsets the pin-adapter wouldn't
work).

Hard to say if Compaq did anything limiting with their bios
but in general a bios that supported Coppermine would run a
Tualatin.

I'd be more concerned with overall lifespan as related to
heat and power... would put the fastest P3 w/133MHz FSB
(presuming the system can do 133MHz FSB) that it can run
with vcore limited to <= 1.6V.

It can do 133 FSB, The system info in the bios says it can anyway. But
I have ZERO control of things like Vcore in the compaq. Seems to be
pretty much Plug and Pray. Compaq did sell the same machine, with a p3
733, presumably 133 fsb, also presumably 1.7v, so I'm not overly
worried.

Anyway, the Tualatins are a no go, they don't work, and I don't have a
pin adapter. Darnit. A p3 1Ghz IS an improvement.
 
K

kony

It can do 133 FSB, The system info in the bios says it can anyway. But
I have ZERO control of things like Vcore in the compaq. Seems to be
pretty much Plug and Pray.

Do some web searches for "Coppermine VID pins". By
soldering bridges across motherboard pins (back of the board
is usually easier), or at the VRM controller on the board,
or wrapping wires around the CPU pins themselves, and/or
putting insulation (like CAT5 wire insulation expanded a
little) over some of the VID pins (then using a small ream
to enlarge the socket hole for that insulated pin just
enough to slip the CPU in), you can get pretty much any
voltage you want.

If you wanted the alternate approach, putting a pot between
3.3V and Gnd and the wiper going to the feedback pin on the
vrm controller will also allow quite a variation in voltage
adjustment (far more versatile, and dangerous). You might
also want shielded leads to cut down on noise pickup to/from
the pot if mounted at a distance. Generally I prefer the
soldering on back of board and insulation on the CPU pins
over any other combination.
Compaq did sell the same machine, with a p3
733, presumably 133 fsb, also presumably 1.7v, so I'm not overly
worried.

Maybe, but I've seen some systems that looked a little heat
stressed over time from running the stock CPU. Compaq has
often had pretty low airflow on some of their systems that
depend on a proprietary form PSU doing all the ventilation
work. At least your active cpu 'sink should help.

Anyway, the Tualatins are a no go, they don't work, and I don't have a
pin adapter. Darnit. A p3 1Ghz IS an improvement.


What's the use of the system?

I'd be tempted to just make a NAS out of it, making CPU
peformance a lot less important.
 
P

paulmd

kony said:
Do some web searches for "Coppermine VID pins". By
soldering bridges across motherboard pins (back of the board
is usually easier), or at the VRM controller on the board,
or wrapping wires around the CPU pins themselves, and/or
putting insulation (like CAT5 wire insulation expanded a
little) over some of the VID pins (then using a small ream
to enlarge the socket hole for that insulated pin just
enough to slip the CPU in), you can get pretty much any
voltage you want.

If you wanted the alternate approach, putting a pot between
3.3V and Gnd and the wiper going to the feedback pin on the
vrm controller will also allow quite a variation in voltage
adjustment (far more versatile, and dangerous). You might
also want shielded leads to cut down on noise pickup to/from
the pot if mounted at a distance. Generally I prefer the
soldering on back of board and insulation on the CPU pins
over any other combination.


Maybe, but I've seen some systems that looked a little heat
stressed over time from running the stock CPU. Compaq has
often had pretty low airflow on some of their systems that
depend on a proprietary form PSU doing all the ventilation
work. At least your active cpu 'sink should help.

I think the normal airflow on this machine comes from the top vent on
this machine, down out the PSU on the bottom. There should be the
normal heat rises thing going on, But the arflow is top-down. I'll have
to double check tomorrow which way the PSU fan is blowing, to be sure.
I might have it exactly backwards. You're right about the default
cooling being inadequite, though. The crappy passive sink these things
came with is hot to the touch even with the stock chip. Though it
probably cools better with the panels ON.
What's the use of the system?

Basicly, The organization I belong to is a non profit computer
refurbisher. In fact, we are a Microsoft Authorised Refurbisher, though
we are also very friendly to Macs. Some systems are regifted to
Qualifying Individuals, and to other nonprofits. Others are resold at
the thrift store, to pay the rent, and other expenses. Since these
machines are eye-candy, and compact, they'll be resold. But we can get
a higher value if they're faster.

As for parts, the upgraded CPUs are usually salvaged from machines with
dead motherboards, or those that are otherwise non-viable, or were
themselves upgraded. The Celerons salvaged from these will probably be
an uprade for some old Etower 400s.

The collection of upgrade parts, and machines is utterly random. And we
see pretty oddball stuff from time to time. Any one want an
unauthorized Mac Clone, or A neXTstation, or an Indy? We also have a
fair idea of just how MUCH a system can be abused (not by us) before
it's worthless. In the case of these machines, they were tossed in a
large cardboard box (3 ft^3), among dead VCRs and misc Crap, with the
side panels distributed among 3 other similar boxes (Which did not
arrive the same week) , and mostly stripped RAM and hard drives.
Cordwood gets better treatment. And so far every last one has been 100%
stable. The 20th time you see this you learn not to worry so much. :)

Since we do materials recovery (copper, gold, alumimum, metal,
stainless steel, and plastics). Dead electronics pay the rent, too.
I'd be tempted to just make a NAS out of it, making CPU
peformance a lot less important.

There's 24 of these machines. (I did say Boatload) Most of them are the
c500, and about 5 or 6 are a newer Celeron, that I haven't opened up
yet to see what's inside. (I think 700 or 800)
 
Top