Problems with Dimension 4550?

Just inherited a Dimension 4550 not working: I plugged it in and the green
light on the MoBo went on. However the cords from the Mobo to the power
switch board (not obviously to the power switch itself) near the USB ports
and the cords from the power switch to that board were unplugged.

When I plugged them both in, the power switch doesn't work at all. No
click, no electronic noise and nothing happens.

Is this model known to have a problem with either of these components?

I also have two working 4400s. Is there an easy solution to getting the 4550
(which seems to have a vastly superior spec) working? Is my only realistic
option to assume that the 4550 mobo is working and that I should just do all
the work necessary to swap motherboards on the 4400? The interior layout of
all units looks very similar with similar heat sinking/cooling of the
processor, the same memory and connectors to the MoBo and the same power
supply.

Or does anyone have experience of it being easier to cannibalise the 4400
for either the power switch or the circuit board near the USB sockets?

First, the cases and motherboards are physically compatible between the 4400 and
the 4550. At the risk of messing up a working 4400, you could swap the 4550
motherboard into the 4400 case. That's too risky. Best thing to do is to
disconnect every device and board from the 4550, except for any video card. Then
see if it powers up. If not, get yourself a working Pentium 4 ATX12v power
supply and test the motherboard with it, rather than the power supply in the
4550. Before doing so, make sure that the memory is in its DIMM sockets and
remove the heat sink to make sure that a processor is in the Socket 478.
Sometimes people remove the processors, and that does cause a problem. If the
board does not power up with a CPU, memory, working power supply, and video
card, then you can assume that the board is dead... Ben Myers

On Wed, 27 Feb 2008 12:00:17 -0500, "news.rcn.com" <news.rnc.com> wrote:

>Just inherited a Dimension 4550 not working: I plugged it in and the green
>light on the MoBo went on. However the cords from the Mobo to the power
>switch board (not obviously to the power switch itself) near the USB ports
>and the cords from the power switch to that board were unplugged.
>
>When I plugged them both in, the power switch doesn't work at all. No
>click, no electronic noise and nothing happens.
>
>Is this model known to have a problem with either of these components?
>
>I also have two working 4400s. Is there an easy solution to getting the 4550
>(which seems to have a vastly superior spec) working? Is my only realistic
>option to assume that the 4550 mobo is working and that I should just do all
>the work necessary to swap motherboards on the 4400? The interior layout of
>all units looks very similar with similar heat sinking/cooling of the
>processor, the same memory and connectors to the MoBo and the same power
>supply.
>
>Or does anyone have experience of it being easier to cannibalise the 4400
>for either the power switch or the circuit board near the USB sockets?
>

"Ben Myers" <> wrote in message
news:...
> First, the cases and motherboards are physically compatible between the
> 4400 and
> the 4550. At the risk of messing up a working 4400, you could swap the
> 4550
> motherboard into the 4400 case. That's too risky. Best thing to do is
> to
> disconnect every device and board from the 4550, except for any video
> card. Then
> see if it powers up. If not, get yourself a working Pentium 4 ATX12v
> power
> supply and test the motherboard with it, rather than the power supply in
> the
> 4550. Before doing so, make sure that the memory is in its DIMM sockets
> and
> remove the heat sink to make sure that a processor is in the Socket 478.
> Sometimes people remove the processors, and that does cause a problem. If
> the
> board does not power up with a CPU, memory, working power supply, and
> video
> card, then you can assume that the board is dead... Ben Myers

Hi Ben
I am grateful to you for your well thought out response but I don't
necessarily understand two things: ( I CAN see that there is a possibility
that the PS may be sending enough voltage to the moboto light up the green
led but not enough to power it on)

Firstly I appreciate your not wanting me
to waste time ruining a working 4400 but I don't THINK I have access to a
Dell power supply. Does Dell still use different power supplies form other
computers or will any ATX with a standard connector and that square 4 pin
24 volt supply do? I have a Bestec ATX-1956D,- but isn't it easier (if
they are different) to put a 4400 one in the 4500?

Secondly why do you discount so completely the possibility that either the
switch itself or the board near the USB connector at the front may have
died?


> On Wed, 27 Feb 2008 12:00:17 -0500, "news.rcn.com" <news.rnc.com> wrote:
>
>>Just inherited a Dimension 4550 not working: I plugged it in and the green
>>light on the MoBo went on. However the cords from the Mobo to the power
>>switch board (not obviously to the power switch itself) near the USB ports
>>and the cords from the power switch to that board were unplugged.
>>
>>When I plugged them both in, the power switch doesn't work at all. No
>>click, no electronic noise and nothing happens.
>>
>>Is this model known to have a problem with either of these components?
>>
>>I also have two working 4400s. Is there an easy solution to getting the
>>4550
>>(which seems to have a vastly superior spec) working? Is my only
>>realistic
>>option to assume that the 4550 mobo is working and that I should just do
>>all
>>the work necessary to swap motherboards on the 4400? The interior layout
>>of
>>all units looks very similar with similar heat sinking/cooling of the
>>processor, the same memory and connectors to the MoBo and the same power
>>supply.
>>
>>Or does anyone have experience of it being easier to cannibalise the 4400
>>for either the power switch or the circuit board near the USB sockets?
>>

I haven't followed this thread but I'll take a stab at answering you
based on things I've read in the past. Keep in mind tho, I may be
wrong but still perhaps worth considering. If I recall Dell power
supplies may be or may not be the same size as 3rd parties but the
dell connections likely are different. Inotherwords, even if the p/s
fits, be careful of the connections.

Hopefully Ben will jump back in here with the correct answer as I'm
not that knowledgeable on power supplies but I read a lot and try to
remember stuff for future knowledge for myself and to pass on to
others.













On Thu, 28 Feb 2008 22:36:32 -0500, "news.rcn.com" <news.rnc.com>
wrote:

>
>"Ben Myers" <> wrote in message
>news:.. .
>> First, the cases and motherboards are physically compatible between the
>> 4400 and
>> the 4550. At the risk of messing up a working 4400, you could swap the
>> 4550
>> motherboard into the 4400 case. That's too risky. Best thing to do is
>> to
>> disconnect every device and board from the 4550, except for any video
>> card. Then
>> see if it powers up. If not, get yourself a working Pentium 4 ATX12v
>> power
>> supply and test the motherboard with it, rather than the power supply in
>> the
>> 4550. Before doing so, make sure that the memory is in its DIMM sockets
>> and
>> remove the heat sink to make sure that a processor is in the Socket 478.
>> Sometimes people remove the processors, and that does cause a problem. If
>> the
>> board does not power up with a CPU, memory, working power supply, and
>> video
>> card, then you can assume that the board is dead... Ben Myers
>
>Hi Ben
>I am grateful to you for your well thought out response but I don't
>necessarily understand two things: ( I CAN see that there is a possibility
>that the PS may be sending enough voltage to the moboto light up the green
>led but not enough to power it on)
>
>Firstly I appreciate your not wanting me
>to waste time ruining a working 4400 but I don't THINK I have access to a
>Dell power supply. Does Dell still use different power supplies form other
>computers or will any ATX with a standard connector and that square 4 pin
>24 volt supply do? I have a Bestec ATX-1956D,- but isn't it easier (if
>they are different) to put a 4400 one in the 4500?
>
>Secondly why do you discount so completely the possibility that either the
>switch itself or the board near the USB connector at the front may have
>died?
>
>
>> On Wed, 27 Feb 2008 12:00:17 -0500, "news.rcn.com" <news.rnc.com> wrote:
>>
>>>Just inherited a Dimension 4550 not working: I plugged it in and the green
>>>light on the MoBo went on. However the cords from the Mobo to the power
>>>switch board (not obviously to the power switch itself) near the USB ports
>>>and the cords from the power switch to that board were unplugged.
>>>
>>>When I plugged them both in, the power switch doesn't work at all. No
>>>click, no electronic noise and nothing happens.
>>>
>>>Is this model known to have a problem with either of these components?
>>>
>>>I also have two working 4400s. Is there an easy solution to getting the
>>>4550
>>>(which seems to have a vastly superior spec) working? Is my only
>>>realistic
>>>option to assume that the 4550 mobo is working and that I should just do
>>>all
>>>the work necessary to swap motherboards on the 4400? The interior layout
>>>of
>>>all units looks very similar with similar heat sinking/cooling of the
>>>processor, the same memory and connectors to the MoBo and the same power
>>>supply.
>>>
>>>Or does anyone have experience of it being easier to cannibalise the 4400
>>>for either the power switch or the circuit board near the USB sockets?
>>>
>
>

Assuming the power supply (red) switch is already set to the correct 115v
rather than accidentally to 220v...... (which would cause the machine to
appear almost completely dead)....


If, with the power switch disconnected from the system board, the rear diag
LEDs (and maybe the case fan as well) "bump" briefly when plugging the
system to main, it's likely one of the components in the front I/O boards or
switch are bad.

It could be the any of them...

power switch board>>>>flat power cable>>>>secondary I/O (USB, Audio, Power)
board >>>> secondary flat cable from I/O board to system board.

-The power switch board could be bad.
-The power cable to the secondary board could be bad, mis-pinned, or
connector pins bent on either board.
-The secondary I/O power board could be bad or damaged.
-The secondary flat cable could be damaged (look for crimps/damage in the
cable insulation).
-The connector pins on the motherboard could be damaged.


All of the components are interchangeable with this assembly from the 4400
with one caveat; the 4400 is a USB 1.1 machine while the 4550 is USB 2.0.
It could be (and probably should be that the system board will dictate those
speeds in any event.)

Getting to the above components is extremely difficult, particularly the
power switch board, as the front bezel plastics (and other base plastics)
have to be removed to get to it. The secondary I/O board is a pain, but
could be carefully swapped and re-cabled as a test if you want to go that
far.

The power supplies in both machines are interchangeable.

No disrepect to Ben, but I'd be willing to swap the boards over to test the
4550 board in the 4400 just to confirm POST and video. You could then focus
on either the power switch assembly or the power supply.

From what you're describing it sounds like some portion of the 4550's switch
assembly. I've seen them do this many times, but it's been several years
since the last one.


Good luck.

Stew

I responded similarly to the other thread which started this one. The entire
chain of hardware leading from the plastic on-off button to the motherboard is
overly complicated, perhaps the weakest point in the design of the Dell black
chassis Pentium 4 systems with motherboards mounted on a metal plate.

In terms of probabilities, I would still test the 4550 motherboard with a
different power supply as my first attempt to isolate the problem. (Actually,
in my environment, I might first remove the 4550 board, put it on the bench and
hook up the whole cable contraption with momentary on-off switch.) But,
because of the complex on-off mechanism, yes, any one of the items in the chain
could be the cause of the problem. You never know, and that is one of the joys
of troubleshooting. Being able to say "Eureka, I've got it!" or something
similar when the cause of the problem is found and fixed.

To respond directly to the title of THIS posting, beginning with the Socket 478
desktop systems (Dimension 2300 and later, Dimension 4300 and later, Dimension
8200 and later, Optiplex GX240 and later, Precision 340 and later), Dell finally
got religion and stopped using non-standard power supplies. Any and all tower
models in the families above use a standard ATX-12v power supply, same size,
shape, mounting holes, etc. With the LGA775 systems (4700, 8400, GX280, 370),
the power supply remains standard, including the newer 24-pin ATX power
connector.

Previously, Dell had screwed up everyone including themselves by imposing on its
OEM suppliers to manufacture motherboards and power supplies with 20-pin power
connectors that LOOK exactly like ATX, but with a different pinout. That made
it impossible to swap in a 3rd party motherboard unless the power supply was
changed out, too. This practice actually began with some of the late Pentium
systems which introduced this totally stupid and idiotic 20-pin connector when
everyone else was still using the pair of old-time ATX connectors... Ben Myers

On Thu, 28 Feb 2008 22:59:27 -0600, "S.Lewis" <> wrote:

>
>Assuming the power supply (red) switch is already set to the correct 115v
>rather than accidentally to 220v...... (which would cause the machine to
>appear almost completely dead)....
>
>
>If, with the power switch disconnected from the system board, the rear diag
>LEDs (and maybe the case fan as well) "bump" briefly when plugging the
>system to main, it's likely one of the components in the front I/O boards or
>switch are bad.
>
>It could be the any of them...
>
>power switch board>>>>flat power cable>>>>secondary I/O (USB, Audio, Power)
>board >>>> secondary flat cable from I/O board to system board.
>
>-The power switch board could be bad.
>-The power cable to the secondary board could be bad, mis-pinned, or
>connector pins bent on either board.
>-The secondary I/O power board could be bad or damaged.
>-The secondary flat cable could be damaged (look for crimps/damage in the
>cable insulation).
>-The connector pins on the motherboard could be damaged.
>
>
>All of the components are interchangeable with this assembly from the 4400
>with one caveat; the 4400 is a USB 1.1 machine while the 4550 is USB 2.0.
>It could be (and probably should be that the system board will dictate those
>speeds in any event.)
>
>Getting to the above components is extremely difficult, particularly the
>power switch board, as the front bezel plastics (and other base plastics)
>have to be removed to get to it. The secondary I/O board is a pain, but
>could be carefully swapped and re-cabled as a test if you want to go that
>far.
>
>The power supplies in both machines are interchangeable.
>
>No disrepect to Ben, but I'd be willing to swap the boards over to test the
>4550 board in the 4400 just to confirm POST and video. You could then focus
>on either the power switch assembly or the power supply.
>
>From what you're describing it sounds like some portion of the 4550's switch
>assembly. I've seen them do this many times, but it's been several years
>since the last one.
>
>
>Good luck.
>
>Stew
>

On Thu, 28 Feb 2008 22:36:32 -0500, "news.rcn.com" <news.rnc.com> wrote:
<SNIP>
>
>"Ben Myers" <> wrote in message
>news:.. .
>> First, the cases and motherboards are physically compatible between the
>> 4400 and
>> the 4550. At the risk of messing up a working 4400, you could swap the
>> 4550
>> motherboard into the 4400 case. That's too risky. Best thing to do is
>> to
>> disconnect every device and board from the 4550, except for any video
>> card. Then
>> see if it powers up. If not, get yourself a working Pentium 4 ATX12v
>> power
>> supply and test the motherboard with it, rather than the power supply in
>> the
>> 4550. Before doing so, make sure that the memory is in its DIMM sockets
>> and
>> remove the heat sink to make sure that a processor is in the Socket 478.
>> Sometimes people remove the processors, and that does cause a problem. If
>> the
>> board does not power up with a CPU, memory, working power supply, and
>> video
>> card, then you can assume that the board is dead... Ben Myers
>
>Hi Ben
>I am grateful to you for your well thought out response but I don't
>necessarily understand two things: ( I CAN see that there is a possibility
>that the PS may be sending enough voltage to the moboto light up the green
>led but not enough to power it on)
>
>Firstly I appreciate your not wanting me
>to waste time ruining a working 4400 but I don't THINK I have access to a
>Dell power supply. Does Dell still use different power supplies form other
>computers or will any ATX with a standard connector and that square 4 pin
>24 volt supply do? I have a Bestec ATX-1956D,- but isn't it easier (if
>they are different) to put a 4400 one in the 4500?
>
>Secondly why do you discount so completely the possibility that either the
>switch itself or the board near the USB connector at the front may have
>died?
>
The power supplies used in Dell Socket 478 Pentium 4 and Celeron systems are
absolutely industry standard in all respects, except one small one. The
Dimension 4xxx and 8xxx series require that the power supply NOT have an on-off
rocker switch which is found on most power supplies. Otherwise, there are the
same 20-pin ATX connector, the 4-pin 12v motherboard connector, and some number
of 5v-12v connectors for disk drives, optical drives and diskette drives. And
the screw mounting holes are the same. So why run the risk of making a 4400
non-functional? Use a standard power supply and do not even mount it inside
the case for testing purposes.

Bestec power supplies are used extensively by HP (for both Compaq and HP brands)
and Gateway (for Gateway and eMachines). The ATX-1956D is a smaller microATX
power supply made for HP. I cannot determine if it has the 4-pin 12v connector
required by a Socket 478 system. If it does, go ahead and use it for testing
purposes. BTW, Bestec power supplies are noted for often taking down a
motherboard when they fail, because they apparently do not have enough circuitry
to protect the motherboard from a power surge. But if the Bestec is known to be
working, test with it, by all means.

I discounted almost entirely the failure of the on-off switch and its wiring on
the front of the 4550 because it is far less likely to fail than either a
motherboard or a power supply, even though Dell uses a complicated set of parts
to simply turn the power on. The other cables to USB and audio on the front of
the system have nothing to do with how the system works. But, check the several
elements here which create a long chain of interrelated parts between on-off
switch and the motherboard.

Leading from the motherboard is a 34-pin ribbon cable which mates with very
small pins at both ends. At the other end is a small circuit board with USB
ports, a hole for an audio jack AND another cable, this one with 12 pins. The
12-pin cable leads to yet another smaller circuit board with on-off momentary
switch and two LEDs. When you press the button on the front of the chassis, it
puts pressure on the momentary switch... Ben Myers

On Feb 29, 9:55 am, Ben Myers <ben_myers_spam_me_...@charter.net>
wrote:
> On Thu, 28 Feb 2008 22:36:32 -0500, "news.rcn.com" <news.rnc.com> wrote:
>
> <SNIP>
>
>
>
> >"Ben Myers" <ben_myers_spam_me_...@charter.net> wrote in message
> >news:.. .
> >> First, the cases and motherboards are physically compatible between the
> >> 4400 and
> >> the 4550. At the risk of messing up a working 4400, you could swap the
> >> 4550
> >> motherboard into the 4400 case. That's too risky. Best thing to do is
> >> to
> >> disconnect every device and board from the 4550, except for any video
> >> card. Then
> >> see if it powers up. If not, get yourself a working Pentium 4 ATX12v
> >> power
> >> supply and test the motherboard with it, rather than the power supply in
> >> the
> >> 4550. Before doing so, make sure that the memory is in its DIMM sockets
> >> and
> >> remove the heat sink to make sure that a processor is in the Socket 478.
> >> Sometimes people remove the processors, and that does cause a problem. If
> >> the
> >> board does not power up with a CPU, memory, working power supply, and
> >> video
> >> card, then you can assume that the board is dead... Ben Myers
>
> >Hi Ben
> >I am grateful to you for your well thought out response but I don't
> >necessarily understand two things: ( I CAN see that there is a possibility
> >that the PS may be sending enough voltage to the moboto light up the green
> >led but not enough to power it on)
>
> >Firstly I appreciate your not wanting me
> >to waste time ruining a working 4400 but I don't THINK I have access to a
> >Dell power supply. Does Dell still use different power supplies form other
> >computers or will any ATX with a standard connector and that square 4 pin
> >24 volt supply do? I have a Bestec ATX-1956D,- but isn't it easier (if
> >they are different) to put a 4400 one in the 4500?
>
> >Secondly why do you discount so completely the possibility that either the
> >switch itself or the board near the USB connector at the front may have
> >died?
>
> The power supplies used in Dell Socket 478 Pentium 4 and Celeron systems are
> absolutely industry standard in all respects, except one small one. The
> Dimension 4xxx and 8xxx series require that the power supply NOT have an on-off
> rocker switch which is found on most power supplies. Otherwise, there are the
> same 20-pin ATX connector, the 4-pin 12v motherboard connector, and some number
> of 5v-12v connectors for disk drives, optical drives and diskette drives. And
> the screw mounting holes are the same. So why run the risk of making a 4400
> non-functional? Use a standard power supply and do not even mount it inside
> the case for testing purposes.
>
> Bestec power supplies are used extensively by HP (for both Compaq and HP brands)
> and Gateway (for Gateway and eMachines). The ATX-1956D is a smaller microATX
> power supply made for HP. I cannot determine if it has the 4-pin 12v connector
> required by a Socket 478 system. If it does, go ahead and use it for testing
> purposes. BTW, Bestec power supplies are noted for often taking down a
> motherboard when they fail, because they apparently do not have enough circuitry
> to protect the motherboard from a power surge. But if the Bestec is known to be
> working, test with it, by all means.
>
> I discounted almost entirely the failure of the on-off switch and its wiring on
> the front of the 4550 because it is far less likely to fail than either a
> motherboard or a power supply, even though Dell uses a complicated set of parts
> to simply turn the power on. The other cables to USB and audio on the front of
> the system have nothing to do with how the system works. But, check the several
> elements here which create a long chain of interrelated parts between on-off
> switch and the motherboard.
>
> Leading from the motherboard is a 34-pin ribbon cable which mates with very
> small pins at both ends. At the other end is a small circuit board with USB
> ports, a hole for an audio jack AND another cable, this one with 12 pins. The
> 12-pin cable leads to yet another smaller circuit board with on-off momentary
> switch and two LEDs. When you press the button on the front of the chassis, it
> puts pressure on the momentary switch... Ben Myers

Ben, you are certainly from the old school...sharing what you have
learned in great detail. Thanks much!

> Ben, you are certainly from the old school...sharing what you have
> learned in great detail. Thanks much!

Allow me to second that, and thank you to SLewis as well!

Actually one of my 4400s isn't in fact working, it goes into a continuous
crash (even the mouse) about ten seconds after the systray comes up both on
normal and in safe mode. No one on the MS forum seems to know why or how to
fix it. I wasn't going to ask here as the problem does seem to be with the
OS but if it doesn't get corrected soon, I might try the PS as I got the
BEstec from CompUSA in the fake liquidation for ten bucks. So if this was
the ONLY actual deal anyone got from those liquidators, it might well be
dead and take the 4500 MoBo with it when plugged in!

No problem. In ways that Michael Dell and friends never imagined, I make part
of my living doing a lot of work on Dell computers. After a while, things
start to come together. Things that Dell never intended to come together,
except that they do, because Dell has done some decent high level planning and
industrial design with its products.

Oh, I guess I am sounding like a Dellbot again, Irene (if she's reading)... Ben
Myers

On Fri, 29 Feb 2008 16:05:38 -0800 (PST), wrote:

<SNIP>
>
>Ben, you are certainly from the old school...sharing what you have
>learned in great detail. Thanks much!