I have an old Dell XPS 420 desktop that wouldn't come on some weeks ago.
I was finally able to get it working by cutting the green wire
connecting the power supply to motherboard and then using a switch to
ground the green wire (on the power supply side) to ground and then
pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed to
come on.

First thing I did was remove all the cards within (video card, USB 3
card, etc). Still no power up. I then decided to unplug the large 24
pin connector from power supply to motherboard and, when I grounded the
power supply green wire, the power supply came on and I believe all the
SSD drives did too as I didn't disconnect those. I decided to try
opening the front DVD drive and it opened right up electronically. So,
the power supply was now working as long as the 24 pin connector was
unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the
motherboard somewhere, since the power supply won't start when the 24
pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to keep flipping the external toggle switch to ground while noting whether or
not the power supply would start and the blue light would illuminate.
Much to my surprise, it finally did, I depressed the front blue
illuminated power button and the PC booted up. In fact, I am sending
this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll be
back to no power up. Any idea what could be causing what appears to be
an intermittent no power up issue? Can we rule out the power supply
since it seems to activate if not plugging in the 24 pin connector (but
it is now)?

Any ideas would be welcome. This old 2008 desktop is my workhorse and I
am basically lost without it. My PC and Internet demands are light, so
that's why I am still able to use it. There are a lot of electrolytics
on the motherboard and I understand that those fail with age and can
cause a short. I do have an electrolytic tester, but I'd have to remove
the motherboard to access the caps and it won't be easy. If that's even
what could be wrong.

Thanks in advance for any help.

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks ago. ÿI was finally able to get it working by cutting the green wire connecting the power supply to motherboard and then using a switch to ground the green wire (on the power supply side) to ground and then pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed to come on.

First thing I did was remove all the cards within (video card, USB 3 card, etc).ÿ Still no power up.ÿ I then decided to unplug the large 24 pin connector from power supply to motherboard and, when I grounded the power supply green wire, the power supply came on and I believe all the SSD drives did too as I didn't disconnect those.ÿ I decided to try opening the front DVD drive and it opened right up electronically.ÿ So, the power supply was now working as long as the 24 pin connector was unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the motherboard somewhere, since the power supply won't start when the 24 pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to keep flipping the external toggle switch to ground while noting whether or not the power supply would start and the blue light would illuminate. Much to my surprise, it finally did, I depressed the front blue illuminated power button and the PC booted up.ÿ In fact, I am sending this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll be back to no power up.ÿ Any idea what could be causing what appears to be an intermittent no power up issue?ÿ Can we rule out the power supply since it seems to activate if not plugging in the 24 pin connector (but it is now)?

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse and I am basically lost without it.ÿ My PC and Internet demands are light, so that's why I am still able to use it.ÿ There are a lot of electrolytics on the motherboard and I understand that those fail with age and can cause a short.ÿ I do have an electrolytic tester, but I'd have to remove the motherboard to access the caps and it won't be easy.ÿ If that's even what could be wrong.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

Chipset is X38. CPU is Core2 Quad, LGA775.
CPU is missing POPCNT instruction in SSE4 and
so is not a good candidate for a Windows 11 attempt
with Rufus. RAM is DDR2 presumably, but there were also
DDR3 versions (the Q45 had DDR3 and it allowed up to
4x4GB DDR3.

PSU might well have regular ATX pinout. You can compare the
colours, to a formfactors.org document and the order of colors
on the 24 pin should be the same. If you're lazy, a PSU at
Best Buy would be on the order of $150 (as they only stock
"mid-range" components, like $600 video cards and $150 supplies).

Try to get an 80+ supply - the Dell 420 is a BTX, and the airflow
in the Dell designs is "shockingly low". It doesn't turn the
fan up enough under load.

Also, in checking my Optiplex 780 refurb, I notice my BTX setup,
the mounting screws for the PSU are flipped 180 degrees for BTX.
If you stick a new ATX supply in there, the hole pattern will
make the new PSU intake vent face the top of the
computer case ("blocked"). Versus sucking in air from inside the case,
while the intake points downward (on a BTX PSU). Keep that
visualization in mind when shopping for a PSU. Some computer
cases are quite dependent on the four rear screws, for adequate
mechanical support for the PSU weight. The screws have custom
offsets, enforcing a single orientation.

*******

As there is a component shortage ongoing, refurbs should rise in
price as new computers rise in price. Normally, some of the
quad core refurbs would be champs right now, but the price
will tell you what kind of "demand" exists. I really don't
think all the scalpers will buy up the refurbs, but ordinary
users might partake.

BTX was a jolly mistake of an idea. I've got one, and enjoy all
the silly highlights of the thing. Like a potted plant, you could attempt
to move the motherboard to another case but... who are we fooling.
That never works. You'd have to make a custom case out of 2x4's
and saw out the bits that don't match :-) It's possible even
an open ATX lab bench would not have dual ATX/BTX capability.
BTX has been out of production for a while, so after market
conveniences might not be there for you.

Nothing says you cannot mount a PSU outside the case and
run the cables inside. But this will take a PSU with
longer cabling, to reach everything in the Dell. For example,
Fortron/Sparkle puts "precise/short" cables on their PSUs
and would only be a candidate for a ATX computer case, with
all load connectors being "conveniently placed" for the PSU to work.

A few ATX PSU manufacturers, make a block diagram of their
supply, with the length of the cables marked in the diagram.
This is handy for jobs where you know the reach will be longer.
My largest computer case, has the ATX PSU in the bottom of the case.
If instead I had to mount that outside the case, the wires might not be
long enough. The machine also has an "open top", which is "death for modders", no place to screw down a bodge up there.

You could attempt to buy a Dell PSU, but what shape would one
of those be in ? The PCPC company used to make "Dell Lookalikes",
and then those could be newer designs. PCPC was bought up or something.

Firepower (still in business... maybe) - research this

https://en.wikipedia.org/wiki/PC_Power_and_Cooling

Paul

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Pragmatic Individual wrote:

On 12/03/2026 00:02, JR Dolobson wrote:

Any ideas would be welcome.

Simply replace it with a refurbished laptop. There's no point wasting
time on that rubbish.

Don't ring me on: +44 (0) 798 4357 488 as I don't take support calls on this number.

I have several such PCs in a lockup, left over from when I used to run a
small computer support business. I'm working though them to establish
whether they work, and if they do I advertise them on Gumtree in the
Norwich area, for œ50 or less. See for example:

<https://www.gumtree.com/p/desktop-workstation-pcs/old-pc-needs-good-home/1511084692>


--
Graham J

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 2026-03-12 03:55, Paul wrote:

At this point, it is replace ATX PSU.

Yes, sounds like it, although I'm with yourself & others in doubting
whether it is worth the expense.

It sounds as though it could be yet another example of capacitor rot,
would the hardware be the right era for that? Let's see ...

https://www.google.com/search?client=firefox-b-e&q=would+a+del+xps+420+be+likely+to+suffer+from+capacitor+rot

"AI Overview

Yes, a Dell XPS 420 is highly likely to suffer from capacitor rot (also
known as capacitor plague or failed, bulging, or leaking capacitors).
Released around 2007?2008, this model falls directly within the era when
many Dell systems used inferior capacitors that are now notorious for
failing, sometimes with failure rates estimated to be quite high.
Key Factors for the Dell XPS 420:

Age and Technology: As a machine from the late 2000s, the
electrolytic capacitors are well past their expected lifespan, making
failure common.

Component Failure: The motherboard in the XPS 420 is specifically
noted as being prone to failure, often linked to these capacitor issues.

Heat & Usage: The system was designed for high-performance tasks
like video editing, which often results in higher heat, accelerating the degradation of these components.

Symptoms: If the unit has trouble booting, has a flashing power
button, or has been sitting unused for a long time, it is highly likely
to have capacitor issues.

It is highly recommended to inspect the motherboard for any swollen,
bulging, or leaking capacitors (often brown or white gunk) before trying
to run the machine, especially if it has been in storage."

So yes, it might well be worth examining the mobo and the internals of
the PSU looking for bulging, leaking, or opened up caps and replacing
any found, but the OP would need to be able to wield a soldering iron
and de-soldering tool to do the job. Last time I tried fixing a rotten
cap it was in a very confined space and very fiddly, and ultimately unsuccessful in that the 'repaired' unit still didn't work, so there
must have been something else wrong with it in addition to the misshapen capacitor. However, I have done such cap rot repairs successfully in
the past, for example the PSU for an old VHS recorder.

--

Fake news kills!

I may be contacted via the contact address given on my website: www.macfh.co.uk


--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks ago. ÿI was finally able to get it working by cutting the green wire connecting the power supply to motherboard and then using a switch to ground the green wire (on the power supply side) to ground and then pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed to come on.

First thing I did was remove all the cards within (video card, USB 3 card, etc).ÿ Still no power up.ÿ I then decided to unplug the large 24 pin connector from power supply to motherboard and, when I grounded the power supply green wire, the power supply came on and I believe all the SSD drives did too as I didn't disconnect those.ÿ I decided to try opening the front DVD drive and it opened right up electronically.ÿ So, the power supply was now working as long as the 24 pin connector was unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the motherboard somewhere, since the power supply won't start when the 24 pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to keep flipping the external toggle switch to ground while noting whether or not the power supply would start and the blue light would illuminate. Much to my surprise, it finally did, I depressed the front blue illuminated power button and the PC booted up.ÿ In fact, I am sending this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll be back to no power up.ÿ Any idea what could be causing what appears to be an intermittent no power up issue?ÿ Can we rule out the power supply since it seems to activate if not plugging in the 24 pin connector (but it is now)?

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse and I am basically lost without it.ÿ My PC and Internet demands are light, so that's why I am still able to use it.ÿ There are a lot of electrolytics on the motherboard and I understand that those fail with age and can cause a short.ÿ I do have an electrolytic tester, but I'd have to remove the motherboard to access the caps and it won't be easy.ÿ If that's even what could be wrong.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

Chipset is X38. CPU is Core2 Quad, LGA775.
CPU is missing POPCNT instruction in SSE4 and
so is not a good candidate for a Windows 11 attempt
with Rufus. RAM is DDR2 presumably, but there were also
DDR3 versions (the Q45 had DDR3 and it allowed up to
4x4GB DDR3.

PSU might well have regular ATX pinout. You can compare the
colours, to a formfactors.org document and the order of colors
on the 24 pin should be the same. If you're lazy, a PSU at
Best Buy would be on the order of $150 (as they only stock
"mid-range" components, like $600 video cards and $150 supplies).

Try to get an 80+ supply - the Dell 420 is a BTX, and the airflow
in the Dell designs is "shockingly low". It doesn't turn the
fan up enough under load.

Also, in checking my Optiplex 780 refurb, I notice my BTX setup,
the mounting screws for the PSU are flipped 180 degrees for BTX.
If you stick a new ATX supply in there, the hole pattern will
make the new PSU intake vent face the top of the
computer case ("blocked"). Versus sucking in air from inside the case,
while the intake points downward (on a BTX PSU). Keep that
visualization in mind when shopping for a PSU. Some computer
cases are quite dependent on the four rear screws, for adequate
mechanical support for the PSU weight. The screws have custom
offsets, enforcing a single orientation.

*******

As there is a component shortage ongoing, refurbs should rise in
price as new computers rise in price. Normally, some of the
quad core refurbs would be champs right now, but the price
will tell you what kind of "demand" exists. I really don't
think all the scalpers will buy up the refurbs, but ordinary
users might partake.

BTX was a jolly mistake of an idea. I've got one, and enjoy all
the silly highlights of the thing. Like a potted plant, you could attempt
to move the motherboard to another case but... who are we fooling.
That never works. You'd have to make a custom case out of 2x4's
and saw out the bits that don't match :-) It's possible even
an open ATX lab bench would not have dual ATX/BTX capability.
BTX has been out of production for a while, so after market
conveniences might not be there for you.

Nothing says you cannot mount a PSU outside the case and
run the cables inside. But this will take a PSU with
longer cabling, to reach everything in the Dell. For example,
Fortron/Sparkle puts "precise/short" cables on their PSUs
and would only be a candidate for a ATX computer case, with
all load connectors being "conveniently placed" for the PSU to work.

A few ATX PSU manufacturers, make a block diagram of their
supply, with the length of the cables marked in the diagram.
This is handy for jobs where you know the reach will be longer.
My largest computer case, has the ATX PSU in the bottom of the case.
If instead I had to mount that outside the case, the wires might not be
long enough. The machine also has an "open top", which is "death for modders",
no place to screw down a bodge up there.

You could attempt to buy a Dell PSU, but what shape would one
of those be in ? The PCPC company used to make "Dell Lookalikes",
and then those could be newer designs. PCPC was bought up or something.

Firepower (still in business... maybe) - research this

https://en.wikipedia.org/wiki/PC_Power_and_Cooling

Paul

Thanks, Paul, I was hoping you would respond as you responded some weeks
back after I added the toggle to ground the power supply green wire independently of motherboard. That worked... for a while.

After a lot of toggle flipping this morning, after about 90 min, the 420 booted up and I am actually writing this from it now. As long as I keep
it on, once on, I'll be good until shut down.

So, are we sure it's the power supply? I didn't think it was because
one of the latest symptoms was power supply shutdown the moment the 24
pin connector was plugged into the motherboard. Yet, it's plugged in
now obviously and working.

My budget is very limited. A $150 Best Buy version would be way, way
beyond my budget, so I was thinking something used. Even if it went out
after a year, might be the best alternative for me for now.

I have unused power supplies here, but almost all of them were
reconfigured for other purposes. I power an electric blanket with one;
have another set up as a general power supply; and have two spare ones
that appeared pulled and kept in storage. I brought one out that seemed
to have similar wattage, plugged it in, and a minute later, a big spark
and the odor of destroyed components. The other spare is only 150 W and
too weak for use with the 420. I would appreciate some power supply
models that would work, if you would be kind enough to share. Dell or
not, just having same configuration and wattage. Thanks in advance.


--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 2026/3/12 14:38:23, JR Dolobson wrote:

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

[]

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse and I am basically lost without it.ÿ My PC and Internet demands are light, so that's why I am still able to use it.ÿ There are a lot of electrolytics on the motherboard and I understand that those fail with age and can cause a short.ÿ I do have an electrolytic tester, but I'd have to remove the motherboard to access the caps and it won't be easy.ÿ If that's even what could be wrong.

My understanding re electrolytics is that, yes, anything decays with
age, but specifically for electrolytics, there was a "bad batch", which
did not show up on original testing, so got bought by most manufacturers
and built into lots of things. The AI enquiry someone has done suggests
that your machine is from the appropriate era, though I'm nowadays
dubious that AIs say what the enquirer wants to hear.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

[]

So, are we sure it's the power supply? I didn't think it was because
one of the latest symptoms was power supply shutdown the moment the 24

that does make it sound like the mobo has the fault; or, it could just
be that the extra load on plugging it in causes the PSU to shut down. I
suppose you could test that to _some_ extent by using another supply, if
you have one that is plug-compatible - you wouldn't have to actually fit
it just to test (whether the mobo is at fault).

pin connector was plugged into the motherboard. Yet, it's plugged in
now obviously and working.

My budget is very limited. A $150 Best Buy version would be way, way
beyond my budget, so I was thinking something used. Even if it went out after a year, might be the best alternative for me for now.

I was going to say it might be time to bite the bullet and just get a refurbished laptop, but then I looked at the price of mine and it was
199.99 (and that's pounds); however, it had been upped to 8G and a 480G
SSD, so maybe you could find something cheaper (especially if you're
willing to transfer your existing SSD). But I'm with you in wanting to
keep the old war-horse going if it's something simple.

I have unused power supplies here, but almost all of them were
reconfigured for other purposes. I power an electric blanket with one;

?!?!? - home-made electric blanket? (Here, electric blankets all run
from the mains, at least bought ones do!)

have another set up as a general power supply; and have two spare ones
that appeared pulled and kept in storage. I brought one out that seemed
to have similar wattage, plugged it in, and a minute later, a big spark
and the odor of destroyed components. The other spare is only 150 W and

During that minute, was it powering the PC?
[]
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()ALIS-Ch++(p)Ar++T+H+Sh0!:`)DNAf

A true-born Englishman does not know any language. He does not speak
English too well either but, at least, he is not proud of this. He is,
however, immensely proud of not knowing any foreign languages.
(George Mikes, "How to be Inimitable" [1960].)

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/12/26 11:53 AM, J. P. Gilliver wrote:

On 2026/3/12 14:38:23, JR Dolobson wrote:

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

[]

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse and I am basically lost without it.ÿ My PC and Internet demands are light, so that's why I am still able to use it.ÿ There are a lot of electrolytics on the motherboard and I understand that those fail with age and can cause a short.ÿ I do have an electrolytic tester, but I'd have to remove the motherboard to access the caps and it won't be easy.ÿ If that's even what could be wrong.

My understanding re electrolytics is that, yes, anything decays with
age, but specifically for electrolytics, there was a "bad batch", which
did not show up on original testing, so got bought by most manufacturers
and built into lots of things. The AI enquiry someone has done suggests
that your machine is from the appropriate era, though I'm nowadays
dubious that AIs say what the enquirer wants to hear.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

[]

So, are we sure it's the power supply? I didn't think it was because
one of the latest symptoms was power supply shutdown the moment the 24

that does make it sound like the mobo has the fault; or, it could just
be that the extra load on plugging it in causes the PSU to shut down. I suppose you could test that to _some_ extent by using another supply, if
you have one that is plug-compatible - you wouldn't have to actually fit
it just to test (whether the mobo is at fault).

Exactly. That's what I'm working on right now. I found another power
supply that was in storage. This is a 700 W one and modular in design.
I think that's the term anyway. Has a bunch of connectors available for plugging in various wired configurations. Fortunately, the wires and
plugs are with it. The only thing wrong is that I must have tried using
it for some experiment one time and hacked the 24 pin connector. I am
in the process right now of correctly restoring the wiring to that
connector, then I'll be able to try it.
>> pin connector was plugged into the motherboard. Yet, it's plugged in

now obviously and working.

My budget is very limited. A $150 Best Buy version would be way, way
beyond my budget, so I was thinking something used. Even if it went out
after a year, might be the best alternative for me for now.

I was going to say it might be time to bite the bullet and just get a refurbished laptop, but then I looked at the price of mine and it was
199.99 (and that's pounds); however, it had been upped to 8G and a 480G
SSD, so maybe you could find something cheaper (especially if you're
willing to transfer your existing SSD). But I'm with you in wanting to
keep the old war-horse going if it's something simple.

I have unused power supplies here, but almost all of them were
reconfigured for other purposes. I power an electric blanket with one;

?!?!? - home-made electric blanket? (Here, electric blankets all run
from the mains, at least bought ones do!)

I hadn't had an electric blanket for years, but a few years ago ran
short on cash with oil heating being almost impossible due to the
expense. I went out and bought a Sunbeam brand, which was the same
brand I had many years ago. However, over the next four years, I was
unable to have any of the ones I purchased last more than half a season.
I don't know what's happened to that brand as my original lasted
better than 15 years, but no longer. After becoming thoroughly
disgusted, I designed my own using one of the defunct Sunbeams as a
template. 60 feet of teflon coated wire later and a PC power supply
modified for variable current, and I had my blanket. Has been going
strong for a decade. I could have designed for mains voltage, but much
safer to put some unused PC power supplies to use. The blanket draws a maximum of 80 watts at 12V, which is nothing for most PC supplies to
handle.

have another set up as a general power supply; and have two spare ones
that appeared pulled and kept in storage. I brought one out that seemed
to have similar wattage, plugged it in, and a minute later, a big spark
and the odor of destroyed components. The other spare is only 150 W and

During that minute, was it powering the PC?
[]

No. I just plugged it in and was about to jumper the green to ground
for a moment to see if it worked. Never had the chance.


--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks
ago. I was finally able to get it working by cutting the green wire connecting the power supply to motherboard and then using a switch to
ground the green wire (on the power supply side) to ground and then
pressing the front power button on the desktop.

My sister has an old Dell that would not come up. CMOS battery?
--
<Bill>

Brought to you from Anchorage, Alaska



--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

JR Dolobson wrote:

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks

(snip)
So, are we sure it's the power supply?ÿ I didn't think it was because
one of the latest symptoms was power supply shutdown the moment the 24
pin connector was plugged into the motherboard.ÿ Yet, it's plugged in
now obviously and working.

My budget is very limited.ÿ A $150 Best Buy version would be way, way
beyond my budget, so I was thinking something used.ÿ Even if it went out after a year, might be the best alternative for me for now.

I have unused power supplies here, but almost all of them were
reconfigured for other purposes.ÿ I power an electric blanket with one;
have another set up as a general power supply; and have two spare ones
that appeared pulled and kept in storage.ÿ I brought one out that seemed
to have similar wattage, plugged it in, and a minute later, a big spark
and the odor of destroyed components.ÿ The other spare is only 150 W and
too weak for use with the 420.ÿ I would appreciate some power supply
models that would work, if you would be kind enough to share.ÿ Dell or
not, just having same configuration and wattage.ÿ Thanks in advance.

You might want to take the PSU apart and resolder every solder point on
it. If the PSU caps look bad then replace them. Those caps are
expensive though. Beware - If good they will hold a significant charge
for a long time. Drain with a 1k resistor or incandescent light bulb
before touching. Check Ebay for a PSU.

I've replaced motherboard caps (and also on other electronics) by
crushing and cutting existing caps, removing the debris, leaving the old
cap wires in place, and then soldering new caps onto the existing wires.
It is sometimes easier and safer than desoldering existing caps on old
style boards... probably won't work though on newer equipment due to RF, harmonics, etc.

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 12/03/2026 17:17, Bill Bradshaw wrote:

My sister has an old Dell that would not come up. CMOS battery?

Don't think so, certainly it may have lost its CMOS settings, but AFAIAA
PCs suffering from that will usually boot, although they may boot very
slowly as they will do a complete POST and their speed settings will
have reverted to a failsafe. Further, as they do so usually they will
give a message that the date & time are wrong, and suggest going into
the BIOS to correct them. They will preserve any settings in the CMOS
that the user changes, such as the date and time, until they are powered
off. Changing a CMOS battery is usually fairly straightforward, it's
just a CR2032 coin cell, which will either be in a hard mount on the
mobo, or else connected via leads to the mobo - in the latter case,
you must ensure that you buy a like for like replacement: same type and
size of connector, same orientation of leads within it.

But for the actual non-booting problem, capacitor rot is again a
possibility, see my other reply and eyeball the mobo.

--

Fake news kills!

I may be contacted via the contact address given on my website: www.macfh.co.uk

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 2026/3/12 17:34:34, Paul in Houston TX wrote:
[]

You might want to take the PSU apart and resolder every solder point on

Yes, you never know; I was once about to give up and go home after
trying to repair a neighbour's TV (I'd turned the lights off), when I
saw sparking around a pin: the solder had sheared in a conical way
around the pin, leaving a perfect cone of solder that looked fine.

it. If the PSU caps look bad then replace them. Those caps are
expensive though. Beware - If good they will hold a significant charge
for a long time. Drain with a 1k resistor or incandescent light bulb

Yes, they can be (peak) mains voltage.

Reminds me of a friend who was working with high voltages; he isolated
the capacitor, then carefully grounded the positive terminal. Then
touched the negative one ... minus X volts hurts just as much as plus X! Voltage is relative!

before touching. Check Ebay for a PSU.

I've replaced motherboard caps (and also on other electronics) by
crushing and cutting existing caps, removing the debris, leaving the old
cap wires in place, and then soldering new caps onto the existing wires.
It is sometimes easier and safer than desoldering existing caps on old style boards... probably won't work though on newer equipment due to RF, harmonics, etc.

Plus a lot of modern ones will be surface-mount - temporarily stuck in
place by masked solder paste, then reflowed in an oven, meaning the only
pads are underneath the can, virtually un-unsolderable without a very directional hot-air gun, sometimes not even then. Your crunch method
might work, as even surface-mount electrolytics usually do have leads of
sorts, just curled under.
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()ALIS-Ch++(p)Ar++T+H+Sh0!:`)DNAf

When I went to see Biddy Baxter [Blue Peter's editor] and told her I
was pregnant, her first reaction was 'Oh good, another viewer'.
- Janet Ellis, RT 2016/2/27-3/4

--- PyGate Linux v1.5.12
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Thu, 3/12/2026 1:48 PM, Java Jive wrote:

On 12/03/2026 17:17, Bill Bradshaw wrote:

My sister has an old Dell that would not come up.ÿ CMOS battery?

Don't think so, certainly it may have lost its CMOS settings, but AFAIAA PCs suffering from that will usually boot, although they may boot very slowly as they will do a complete POST and their speed settings will have reverted to a failsafe.ÿ Further, as they do so usually they will give a message that the date & time are wrong, and suggest going into the BIOS to correct them.ÿ They will preserve any settings in the CMOS that the user changes, such as the date and time, until they are powered off.ÿ Changing a CMOS battery is usually fairly straightforward, it's just a CR2032 coin cell, which will either be in a hard mount on the mobo, or else connected via leads to the moboÿ -ÿ in the latter case, you must ensure that you buy a like for like replacement: same type and size of connector, same orientation of leads within it.

But for the actual non-booting problem, capacitor rot is again a possibility, see my other reply and eyeball the mobo.

There are SuperIO chips where having VBAT drop to zero volts,
stops POST. The documentation never seems to hint at where
this function is hiding... Most annoying.

But the message here, is that Bill is right, and a small percentage
of boards have that problem. And a contributor, is the ability
of the SuperIO to measure VBAT. It has a connection to VBAT,
specifically for measuring it (high impedance input), but...
we're not told what happens with the info collected.

Most boards are like this. They don't measure VBAT.

|
LPC
|
3.3V SuperIO Southbridge (PCH)
5.0V (Hardware Mon
12.0V Volts) VBAT --+-- CMOS Clock # Uses a moat with T-gates.
+-- 256 byte CMOS RAM # Isolated from attempts to drain it.
# 10uA load.

It is when you see this, you have to wonder why that wire is connected.
For function purposes, just the Southbridge needs the VBAT connection.
If the SuperIO is "monitoring" the signal, we should see more reports
of "hey, my BIOS shows the VBAT voltage, neat". But we don't really
see reports like that, and instead we hear stories of "I have one
of those PCs that stops running when VBAT is flat". And there is in fact,
no reason to worry about VBAT, as a PC will "Restore Defaults"
if the checksum is wrong on the CMOS content, and a working clock is
not a requirement for achieving boot. If you're a cheapskate, you
can turn on the PC every day, enter the BIOS, fix up the boot order,
sync the clock to NIST later, and off you go.

|
LPC
|
3.3V SuperIO Southbridge (PCH)
5.0V (Hardware Mon
12.0V Volts) VBAT --+-- CMOS Clock # Uses a moat with T-gates.
^ | +-- 256 byte CMOS RAM # Isolated from attempts to drain it.
+----------------+ # 10uA load.
Makes no sense...

Paul

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks ago. ÿI was finally able to get it working by cutting the green wire connecting the power supply to motherboard and then using a switch to ground the green wire (on the power supply side) to ground and then pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed to come on.

First thing I did was remove all the cards within (video card, USB 3 card, etc).ÿ Still no power up.ÿ I then decided to unplug the large 24 pin connector from power supply to motherboard and, when I grounded the power supply green wire, the power supply came on and I believe all the SSD drives did too as I didn't disconnect those.ÿ I decided to try opening the front DVD drive and it opened right up electronically.ÿ So, the power supply was now working as long as the 24 pin connector was unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the motherboard somewhere, since the power supply won't start when the 24 pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to keep flipping the external toggle switch to ground while noting whether or not the power supply would start and the blue light would illuminate. Much to my surprise, it finally did, I depressed the front blue illuminated power button and the PC booted up.ÿ In fact, I am sending this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll be back to no power up.ÿ Any idea what could be causing what appears to be an intermittent no power up issue?ÿ Can we rule out the power supply since it seems to activate if not plugging in the 24 pin connector (but it is now)?

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse and I am basically lost without it.ÿ My PC and Internet demands are light, so that's why I am still able to use it.ÿ There are a lot of electrolytics on the motherboard and I understand that those fail with age and can cause a short.ÿ I do have an electrolytic tester, but I'd have to remove the motherboard to access the caps and it won't be easy.ÿ If that's even what could be wrong.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

Chipset is X38. CPU is Core2 Quad, LGA775.
CPU is missing POPCNT instruction in SSE4 and
so is not a good candidate for a Windows 11 attempt
with Rufus. RAM is DDR2 presumably, but there were also
DDR3 versions (the Q45 had DDR3 and it allowed up to
4x4GB DDR3.

PSU might well have regular ATX pinout. You can compare the
colours, to a formfactors.org document and the order of colors
on the 24 pin should be the same. If you're lazy, a PSU at
Best Buy would be on the order of $150 (as they only stock
"mid-range" components, like $600 video cards and $150 supplies).

Try to get an 80+ supply - the Dell 420 is a BTX, and the airflow
in the Dell designs is "shockingly low". It doesn't turn the
fan up enough under load.

Also, in checking my Optiplex 780 refurb, I notice my BTX setup,
the mounting screws for the PSU are flipped 180 degrees for BTX.
If you stick a new ATX supply in there, the hole pattern will
make the new PSU intake vent face the top of the
computer case ("blocked"). Versus sucking in air from inside the case,
while the intake points downward (on a BTX PSU). Keep that
visualization in mind when shopping for a PSU. Some computer
cases are quite dependent on the four rear screws, for adequate
mechanical support for the PSU weight. The screws have custom
offsets, enforcing a single orientation.

*******

As there is a component shortage ongoing, refurbs should rise in
price as new computers rise in price. Normally, some of the
quad core refurbs would be champs right now, but the price
will tell you what kind of "demand" exists. I really don't
think all the scalpers will buy up the refurbs, but ordinary
users might partake.

BTX was a jolly mistake of an idea. I've got one, and enjoy all
the silly highlights of the thing. Like a potted plant, you could attempt
to move the motherboard to another case but... who are we fooling.
That never works. You'd have to make a custom case out of 2x4's
and saw out the bits that don't match :-) It's possible even
an open ATX lab bench would not have dual ATX/BTX capability.
BTX has been out of production for a while, so after market
conveniences might not be there for you.

Nothing says you cannot mount a PSU outside the case and
run the cables inside. But this will take a PSU with
longer cabling, to reach everything in the Dell. For example,
Fortron/Sparkle puts "precise/short" cables on their PSUs
and would only be a candidate for a ATX computer case, with
all load connectors being "conveniently placed" for the PSU to work.

A few ATX PSU manufacturers, make a block diagram of their
supply, with the length of the cables marked in the diagram.
This is handy for jobs where you know the reach will be longer.
My largest computer case, has the ATX PSU in the bottom of the case.
If instead I had to mount that outside the case, the wires might not be
long enough. The machine also has an "open top", which is "death for modders",
no place to screw down a bodge up there.

You could attempt to buy a Dell PSU, but what shape would one
of those be in ? The PCPC company used to make "Dell Lookalikes",
and then those could be newer designs. PCPC was bought up or something.

Firepower (still in business... maybe) - research this

https://en.wikipedia.org/wiki/PC_Power_and_Cooling

Paul

I found an unused Ultra ult-vx700 I had in storage yesterday. Problem
with it was that I cut about half the 24 pin wires away because I had
tried using the PSU for another purpose. However, a lot of tedious
soldering and heat shrinking, and I have spliced on an unused 24 pin
connector from yet another unused (and burned out) PSU today. I was
extremely careful to match the original connector with the new (spliced)
one. Taking a break now, then will try it in the PC.

My only concern was the unused pin on Dell's 24 pin connector (pin 20)
yet there was a wire there, so at least wired, in the VX700's original connector. I decided not to use that wire and leave pin 20 unused. Not
sure that will be an issue. I wasn't able to find any voltage data for
the VX700's connector other than the PSU is ATX design.

Since the '700 is what I think they call "modular" in design, there are
about a dozen (or more) connectors included that terminate in the
various connectors needed for the Dell. These are all plugged in to
sockets on one side of the PSU.

I hope for success. Not going to install it yet, just try it out from
outside the case.



--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/13/26 5:31 AM, JR Dolobson wrote:

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks
ago. ÿI was finally able to get it working by cutting the green wire
connecting the power supply to motherboard and then using a switch to
ground the green wire (on the power supply side) to ground and then
pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed
to come on.

First thing I did was remove all the cards within (video card, USB 3
card, etc).ÿ Still no power up.ÿ I then decided to unplug the large
24 pin connector from power supply to motherboard and, when I
grounded the power supply green wire, the power supply came on and I
believe all the SSD drives did too as I didn't disconnect those.ÿ I
decided to try opening the front DVD drive and it opened right up
electronically.ÿ So, the power supply was now working as long as the
24 pin connector was unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the
motherboard somewhere, since the power supply won't start when the 24
pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to
keep flipping the external toggle switch to ground while noting
whether or not the power supply would start and the blue light would
illuminate. Much to my surprise, it finally did, I depressed the
front blue illuminated power button and the PC booted up.ÿ In fact, I
am sending this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll be
back to no power up.ÿ Any idea what could be causing what appears to
be an intermittent no power up issue?ÿ Can we rule out the power
supply since it seems to activate if not plugging in the 24 pin
connector (but it is now)?

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse
and I am basically lost without it.ÿ My PC and Internet demands are
light, so that's why I am still able to use it.ÿ There are a lot of
electrolytics on the motherboard and I understand that those fail
with age and can cause a short.ÿ I do have an electrolytic tester,
but I'd have to remove the motherboard to access the caps and it
won't be easy.ÿ If that's even what could be wrong.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

Chipset is X38. CPU is Core2 Quad, LGA775.
CPU is missing POPCNT instruction in SSE4 and
so is not a good candidate for a Windows 11 attempt
with Rufus. RAM is DDR2 presumably, but there were also
DDR3 versions (the Q45 had DDR3 and it allowed up to
4x4GB DDR3.

PSU might well have regular ATX pinout. You can compare the
colours, to a formfactors.org document and the order of colors
on the 24 pin should be the same. If you're lazy, a PSU at
Best Buy would be on the order of $150 (as they only stock
"mid-range" components, like $600 video cards and $150 supplies).

Try to get an 80+ supply - the Dell 420 is a BTX, and the airflow
in the Dell designs is "shockingly low". It doesn't turn the
fan up enough under load.

Also, in checking my Optiplex 780 refurb, I notice my BTX setup,
the mounting screws for the PSU are flipped 180 degrees for BTX.
If you stick a new ATX supply in there, the hole pattern will
make the new PSU intake vent face the top of the
computer case ("blocked"). Versus sucking in air from inside the case,
while the intake points downward (on a BTX PSU). Keep that
visualization in mind when shopping for a PSU. Some computer
cases are quite dependent on the four rear screws, for adequate
mechanical support for the PSU weight. The screws have custom
offsets, enforcing a single orientation.

*******

As there is a component shortage ongoing, refurbs should rise in
price as new computers rise in price. Normally, some of the
quad core refurbs would be champs right now, but the price
will tell you what kind of "demand" exists. I really don't
think all the scalpers will buy up the refurbs, but ordinary
users might partake.

BTX was a jolly mistake of an idea. I've got one, and enjoy all
the silly highlights of the thing. Like a potted plant, you could attempt
to move the motherboard to another case but... who are we fooling.
That never works. You'd have to make a custom case out of 2x4's
and saw out the bits that don't match :-) It's possible even
an open ATX lab bench would not have dual ATX/BTX capability.
BTX has been out of production for a while, so after market
conveniences might not be there for you.

Nothing says you cannot mount a PSU outside the case and
run the cables inside. But this will take a PSU with
longer cabling, to reach everything in the Dell. For example,
Fortron/Sparkle puts "precise/short" cables on their PSUs
and would only be a candidate for a ATX computer case, with
all load connectors being "conveniently placed" for the PSU to work.

A few ATX PSU manufacturers, make a block diagram of their
supply, with the length of the cables marked in the diagram.
This is handy for jobs where you know the reach will be longer.
My largest computer case, has the ATX PSU in the bottom of the case.
If instead I had to mount that outside the case, the wires might not be
long enough. The machine also has an "open top", which is "death for
modders",
no place to screw down a bodge up there.

You could attempt to buy a Dell PSU, but what shape would one
of those be in ? The PCPC company used to make "Dell Lookalikes",
and then those could be newer designs. PCPC was bought up or something.

ÿÿÿ Firepower (still in business... maybe) - research this

ÿÿÿ https://en.wikipedia.org/wiki/PC_Power_and_Cooling

ÿÿ Paul

I found an unused Ultra ult-vx700 I had in storage yesterday.ÿ Problem
with it was that I cut about half the 24 pin wires away because I had
tried using the PSU for another purpose.ÿ However, a lot of tedious soldering and heat shrinking, and I have spliced on an unused 24 pin connector from yet another unused (and burned out) PSU today.ÿ I was extremely careful to match the original connector with the new (spliced) one.ÿ Taking a break now, then will try it in the PC.

My only concern was the unused pin on Dell's 24 pin connector (pin 20)
yet there was a wire there, so at least wired, in the VX700's original connector. I decided not to use that wire and leave pin 20 unused.ÿ Not
sure that will be an issue.ÿ I wasn't able to find any voltage data for
the VX700's connector other than the PSU is ATX design.

Since the '700 is what I think they call "modular" in design, there are about a dozen (or more) connectors included that terminate in the
various connectors needed for the Dell.ÿ These are all plugged in to
sockets on one side of the PSU.

I hope for success.ÿ Not going to install it yet, just try it out from outside the case.

Working, normal boot up now, no extra switches or tricks! I guess is
was the PSU. There is only one problem.... I was unable to find a 6 pin
video card cable! Since the old PSU had one, and it worked without
plugging in the 24 pin connector, I actually have it running
simultaneously with the '700 and just powering the video card with it.

Obviously, I can't keep it this way so.... can you or anyone suggest a 4
pin to 6 pin adapter. I'd rather hack up one of these unused 4 pin
connectors and create a 4 pin to 6 pin plug if possible....




--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/13/26 6:25 AM, JR Dolobson wrote:

On 3/13/26 5:31 AM, JR Dolobson wrote:

On 3/11/26 11:55 PM, Paul wrote:

On Wed, 3/11/2026 8:02 PM, JR Dolobson wrote:

I have an old Dell XPS 420 desktop that wouldn't come on some weeks
ago. ÿI was finally able to get it working by cutting the green wire
connecting the power supply to motherboard and then using a switch
to ground the green wire (on the power supply side) to ground and
then pressing the front power button on the desktop.

Five days ago, even the switch workaround failed and the unit failed
to come on.

First thing I did was remove all the cards within (video card, USB 3
card, etc).ÿ Still no power up.ÿ I then decided to unplug the large
24 pin connector from power supply to motherboard and, when I
grounded the power supply green wire, the power supply came on and I
believe all the SSD drives did too as I didn't disconnect those.ÿ I
decided to try opening the front DVD drive and it opened right up
electronically.ÿ So, the power supply was now working as long as the
24 pin connector was unplugged from the motherboard.

I'm not sure, but it seems to me like there is a short on the
motherboard somewhere, since the power supply won't start when the
24 pin connector is plugged in.

Tonight, after plugging back in the 24 pin connector, I decided to
keep flipping the external toggle switch to ground while noting
whether or not the power supply would start and the blue light would
illuminate. Much to my surprise, it finally did, I depressed the
front blue illuminated power button and the PC booted up.ÿ In fact,
I am sending this message from it now.

Now I'm not sure whether I want to turn it off again for fear I'll
be back to no power up.ÿ Any idea what could be causing what appears
to be an intermittent no power up issue?ÿ Can we rule out the power
supply since it seems to activate if not plugging in the 24 pin
connector (but it is now)?

Any ideas would be welcome.ÿ This old 2008 desktop is my workhorse
and I am basically lost without it.ÿ My PC and Internet demands are
light, so that's why I am still able to use it.ÿ There are a lot of
electrolytics on the motherboard and I understand that those fail
with age and can cause a short.ÿ I do have an electrolytic tester,
but I'd have to remove the motherboard to access the caps and it
won't be easy.ÿ If that's even what could be wrong.

Thanks in advance for any help.

At this point, it is replace ATX PSU.

Chipset is X38. CPU is Core2 Quad, LGA775.
CPU is missing POPCNT instruction in SSE4 and
so is not a good candidate for a Windows 11 attempt
with Rufus. RAM is DDR2 presumably, but there were also
DDR3 versions (the Q45 had DDR3 and it allowed up to
4x4GB DDR3.

PSU might well have regular ATX pinout. You can compare the
colours, to a formfactors.org document and the order of colors
on the 24 pin should be the same. If you're lazy, a PSU at
Best Buy would be on the order of $150 (as they only stock
"mid-range" components, like $600 video cards and $150 supplies).

Try to get an 80+ supply - the Dell 420 is a BTX, and the airflow
in the Dell designs is "shockingly low". It doesn't turn the
fan up enough under load.

Also, in checking my Optiplex 780 refurb, I notice my BTX setup,
the mounting screws for the PSU are flipped 180 degrees for BTX.
If you stick a new ATX supply in there, the hole pattern will
make the new PSU intake vent face the top of the
computer case ("blocked"). Versus sucking in air from inside the case,
while the intake points downward (on a BTX PSU). Keep that
visualization in mind when shopping for a PSU. Some computer
cases are quite dependent on the four rear screws, for adequate
mechanical support for the PSU weight. The screws have custom
offsets, enforcing a single orientation.

*******

As there is a component shortage ongoing, refurbs should rise in
price as new computers rise in price. Normally, some of the
quad core refurbs would be champs right now, but the price
will tell you what kind of "demand" exists. I really don't
think all the scalpers will buy up the refurbs, but ordinary
users might partake.

BTX was a jolly mistake of an idea. I've got one, and enjoy all
the silly highlights of the thing. Like a potted plant, you could
attempt
to move the motherboard to another case but... who are we fooling.
That never works. You'd have to make a custom case out of 2x4's
and saw out the bits that don't match :-) It's possible even
an open ATX lab bench would not have dual ATX/BTX capability.
BTX has been out of production for a while, so after market
conveniences might not be there for you.

Nothing says you cannot mount a PSU outside the case and
run the cables inside. But this will take a PSU with
longer cabling, to reach everything in the Dell. For example,
Fortron/Sparkle puts "precise/short" cables on their PSUs
and would only be a candidate for a ATX computer case, with
all load connectors being "conveniently placed" for the PSU to work.

A few ATX PSU manufacturers, make a block diagram of their
supply, with the length of the cables marked in the diagram.
This is handy for jobs where you know the reach will be longer.
My largest computer case, has the ATX PSU in the bottom of the case.
If instead I had to mount that outside the case, the wires might not be
long enough. The machine also has an "open top", which is "death for
modders",
no place to screw down a bodge up there.

You could attempt to buy a Dell PSU, but what shape would one
of those be in ? The PCPC company used to make "Dell Lookalikes",
and then those could be newer designs. PCPC was bought up or something.

ÿÿÿ Firepower (still in business... maybe) - research this

ÿÿÿ https://en.wikipedia.org/wiki/PC_Power_and_Cooling

ÿÿ Paul

I found an unused Ultra ult-vx700 I had in storage yesterday.ÿ Problem
with it was that I cut about half the 24 pin wires away because I had
tried using the PSU for another purpose.ÿ However, a lot of tedious
soldering and heat shrinking, and I have spliced on an unused 24 pin
connector from yet another unused (and burned out) PSU today.ÿ I was
extremely careful to match the original connector with the new
(spliced) one.ÿ Taking a break now, then will try it in the PC.

My only concern was the unused pin on Dell's 24 pin connector (pin 20)
yet there was a wire there, so at least wired, in the VX700's original
connector. I decided not to use that wire and leave pin 20 unused.
Not sure that will be an issue.ÿ I wasn't able to find any voltage
data for the VX700's connector other than the PSU is ATX design.

Since the '700 is what I think they call "modular" in design, there
are about a dozen (or more) connectors included that terminate in the
various connectors needed for the Dell.ÿ These are all plugged in to
sockets on one side of the PSU.

I hope for success.ÿ Not going to install it yet, just try it out from
outside the case.

Working, normal boot up now, no extra switches or tricks!ÿ I guess is
was the PSU.ÿ There is only one problem.... I was unable to find a 6 pin video card cable!ÿ Since the old PSU had one, and it worked without
plugging in the 24 pin connector, I actually have it running
simultaneously with the '700 and just powering the video card with it.

Obviously, I can't keep it this way so.... can you or anyone suggest a 4
pin to 6 pin adapter.ÿ I'd rather hack up one of these unused 4 pin connectors and create a 4 pin to 6 pin plug if possible....

Done! Made a DIY adapter after finding plug data online. All working
fine now. After weeks, the PC is booting up again normally with no
extra switch. I don't believe this PSU quite matches the specs you
mentioned before, but I believe it was originally for "elementary" gamer systems.

I'll keep it on top of the case for a week or so just to make sure all
is well, then pull the existing one and replace it with this one after
taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame. You did mention it weeks
back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin connector to motherboard. At some point, I'll have to open it and see
if there is a bad electrolytic or two. I have a tester for
electrolytics in circuit, but hasn't been used since purchase a couple
of years back. Will be a good test for it.

Thanks again for all the help, both you, Paul, and the others.



--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 2026/3/13 11:52:40, JR Dolobson wrote:
[]

Done! Made a DIY adapter after finding plug data online. All working
fine now. After weeks, the PC is booting up again normally with no
extra switch. I don't believe this PSU quite matches the specs you mentioned before, but I believe it was originally for "elementary" gamer systems.

Glad to hear all is well.

I'll keep it on top of the case for a week or so just to make sure all
is well, then pull the existing one and replace it with this one after taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame. You did mention it weeks
back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin connector to motherboard.

That has elements of "apart from that, Mrs. Lincoln, did you enjoy the
play?"

At some point, I'll have to open it and see

Be careful; as well as the high voltages (even after off for a while, in theory), a few decades ago I cut myself; I think the normal insides of a
PC they're reasonably careful as people might be expected to poke around inside, but the insides of a power supply they seemed to think fewer
people would go, and were more careless about designing how the parts of
its case go together. At least, that's what I assumed. Things may have improved.

if there is a bad electrolytic or two. I have a tester for
electrolytics in circuit, but hasn't been used since purchase a couple
of years back. Will be a good test for it.

Sounds like a useful toy!

Thanks again for all the help, both you, Paul, and the others.

YW.
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()ALIS-Ch++(p)Ar++T+H+Sh0!:`)DNAf

Being punctual makes people think you have nothing to do.

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Fri, 3/13/2026 10:39 AM, J. P. Gilliver wrote:

On 2026/3/13 11:52:40, JR Dolobson wrote:
[]

Done! Made a DIY adapter after finding plug data online. All working
fine now. After weeks, the PC is booting up again normally with no
extra switch. I don't believe this PSU quite matches the specs you
mentioned before, but I believe it was originally for "elementary" gamer
systems.

Glad to hear all is well.

I'll keep it on top of the case for a week or so just to make sure all
is well, then pull the existing one and replace it with this one after
taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame. You did mention it weeks
back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin
connector to motherboard.

That has elements of "apart from that, Mrs. Lincoln, did you enjoy the
play?"

My PSU from my very first build failed that way. They can become "weak".

In my description some time ago, I pointed out the special properties
of the PS_ON# signal. The spec for it was written as if a logic gate
would measure the voltage on the pin. You could for example, use a
Schmidt Trigger to clean up the signal, and give crispy ON/OFF behavior.

Instead, the circuit uses transistors and has a measure of "analog"
behavior. There have been cases, where the PSU was "half-on" due to a
defect in the PS_ON# interface on the PSU. When these things happen,
the problem can actually be that the PS_ON# signal is at 1.25V instead
of 0V, "making the supply seem weak" when it is not weak.

On my HEC supply, I could power one or maybe two fans (12V @ 200mA total) before the 12V output voltage would start to drop. The supply would be
a 12V @ 15A output normally. The supply is not "dead" as such -- it's
not a primary side switching transistor failure. But the feedback loop
does not seem interested in making the regulation of the thing a
voltage source on output.

And this is how some of the supplies go. They "die of weakness".
You can run two fans off them, but not an entire computer.

This is why, in my answers to the OP (the history of this one),
we were already having trouble just getting PS_ON# to respond at
all. The OP has been using a bypass (ground PS_ON#) to even get his supply to run.
Once the "weak supply" symptom shows up, that's strike two against
the PSU. And the quickest fix for a home tech, is a PSU swap
when we're seeing a "Theme" to what is going on.

The outcome of this one could have been different. I was playing
percentages.

For example, I could be sitting here a week from now and seeing
a message indicating yet another failure. This sort of thing
has happened before (a dribbling failure case).

Paul

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/13/26 10:39 AM, J. P. Gilliver wrote:

On 2026/3/13 11:52:40, JR Dolobson wrote:
[]

Done! Made a DIY adapter after finding plug data online. All working
fine now. After weeks, the PC is booting up again normally with no
extra switch. I don't believe this PSU quite matches the specs you
mentioned before, but I believe it was originally for "elementary" gamer
systems.

Glad to hear all is well.

I'll keep it on top of the case for a week or so just to make sure all
is well, then pull the existing one and replace it with this one after
taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame. You did mention it weeks
back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin
connector to motherboard.

That has elements of "apart from that, Mrs. Lincoln, did you enjoy the
play?"

At some point, I'll have to open it and see

Be careful; as well as the high voltages (even after off for a while, in theory), a few decades ago I cut myself; I think the normal insides of a
PC they're reasonably careful as people might be expected to poke around inside, but the insides of a power supply they seemed to think fewer
people would go, and were more careless about designing how the parts of
its case go together. At least, that's what I assumed. Things may have improved.

One time, I was "hit" with an unexpected charge from a camera flash that
had been disassembled. The electrolytic obviously wasn't discharged
before I touched it. Sent me about halfway across the room. I know PC
and indeed other power supplies can harbor nasty charges long after deactivation.

if there is a bad electrolytic or two. I have a tester for
electrolytics in circuit, but hasn't been used since purchase a couple
of years back. Will be a good test for it.

Sounds like a useful toy!

An ESR tester:

https://www.peakelec.co.uk/downloads/esr70-datasheet-en.pdf

I got this several years ago when I had an amateur radio that began to
act oddly. Being about three decades old, some amateurs suggested electrolytic testing and replacement. I just haven't had the chance. A tedious job that will take a lot of time, but I was recommended the
tester for efficient checking.

Thanks again for all the help, both you, Paul, and the others.

YW.

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Partially defunct Dell PSU removed from casing within the last hour and replaced with the Ultra ult-vx700.

Dell PSU removal was not easy. There is an extra long bunch of wires
that wrap around the motherboard, go underneath the air shroud, and
terminate in various connectors. So, it wasn't as simple as just the 24
pin connector and 4 pin CPU connector. Others with a similar
predicament either removed the air shroud (which also apparently
separates the CPU heatsink, something I did not want to do), or cut the
group of wires near the PSU. Since I hope to maybe resurrect the PSU at
some point and not have to solder a zillion splices and heat shrink them again, I worked the connector end back and forth until it finally pulled through. So, Dell PSU removal is totally intact.

Ultra installation went ok. Not a perfect fit, from an attachment to
casing perspective, as I was only able to attach using two out of the
four screws. I believe this PSU has two fans, one at the front and
another at the rear, but of course not certain of this. That's the way
the ventilation will have to do. There are some small side vents, but
they'll be right up against the sides of the desktop casing once closed.
Hope that is not an issue. The original Dell PSU did not even have
the side vents only front and rear, so I think all will be ok.

Dell uses some small unknown screws to keep the PSU assembled. Anyone
know what they are? Definitely not phillips heads, maybe some sort of
star heads.

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

JR Dolobson wrote:

Partially defunct Dell PSU removed from casing within the last hour and replaced with the Ultra ult-vx700.

Dell PSU removal was not easy.ÿ There is an extra long bunch of wires
that wrap around the motherboard, go underneath the air shroud, and terminate in various connectors.ÿ So, it wasn't as simple as just the 24
pin connector and 4 pin CPU connector.ÿ Others with a similar
predicament either removed the air shroud (which also apparently
separates the CPU heatsink, something I did not want to do), or cut the group of wires near the PSU.ÿ Since I hope to maybe resurrect the PSU at some point and not have to solder a zillion splices and heat shrink them again, I worked the connector end back and forth until it finally pulled through.ÿ So, Dell PSU removal is totally intact.

Ultra installation went ok.ÿ Not a perfect fit, from an attachment to
casing perspective, as I was only able to attach using two out of the
four screws.ÿ I believe this PSU has two fans, one at the front and
another at the rear, but of course not certain of this.ÿ That's the way
the ventilation will have to do.ÿ There are some small side vents, but they'll be right up against the sides of the desktop casing once closed.
ÿHope that is not an issue.ÿ The original Dell PSU did not even have
the side vents only front and rear, so I think all will be ok.

Dell uses some small unknown screws to keep the PSU assembled.ÿ Anyone
know what they are?ÿ Definitely not phillips heads, maybe some sort of
star heads.

Dell uses a lot of T-5 torx, metric thread.
I have a few left over from one repair... threw the rest away and
installed Phillips instead.

https://www.parts-people.com/index.php?action=category&id=162&subid=684&refine=motherboard&srsltid=AfmBOooJlbT_khSEeWux-0M5RUpQ_OJnI_N60Nl8DqK3nK8Y0z5wzOv7

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On 3/15/26 8:18 AM, Daniel70 wrote:

On 15/03/2026 4:03 am, JR Dolobson wrote:

On 3/13/26 10:39 AM, J. P. Gilliver wrote:

On 2026/3/13 11:52:40, JR Dolobson wrote:
[]

Done!ÿ Made a DIY adapter after finding plug data online.ÿ All working >>>> fine now.ÿ After weeks, the PC is booting up again normally with no
extra switch.ÿ I don't believe this PSU quite matches the specs you
mentioned before, but I believe it was originally for "elementary"
gamer
systems.

Glad to hear all is well.

I'll keep it on top of the case for a week or so just to make sure all >>>> is well, then pull the existing one and replace it with this one after >>>> taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame.ÿ You did mention it weeks >>>> back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin >>>> connector to motherboard.

That has elements of "apart from that, Mrs. Lincoln, did you enjoy the
play?"

At some point, I'll have to open it and see

Be careful; as well as the high voltages (even after off for a while, in >>> theory), a few decades ago I cut myself; I think the normal insides of a >>> PC they're reasonably careful as people might be expected to poke around >>> inside, but the insides of a power supply they seemed to think fewer
people would go, and were more careless about designing how the parts of >>> its case go together. At least, that's what I assumed. Things may have
improved.

One time, I was "hit" with an unexpected charge from a camera flash
that had been disassembled.ÿ The electrolytic obviously wasn't
discharged before I touched it.ÿ Sent me about halfway across the
room.ÿ I know PC and indeed other power supplies can harbor nasty
charges long after deactivation.

I can remember, back in about 1980, I was working on a 20kW PEP,
Automatic Tune H.F. Radio Transmitter (with 7.500 Volt H.T.) that wasn't auto-tuning correctly. I wanted to make an adjustment in the high power
RF cabinet so I switched off the HT. I then switched off the LT. I
walked around the back of the 3mtr by 1mtr cabinet and slide open the panelling there and climbed a three or four step ladder then removed the internal cabinet cover. When I removed that cover, a knife-switch was SUPPOSED to short the 7500 or 8000 HT Filter Capacitors to Deck

Sounds like you are (or were) a broadcast engineer. Although I possess
a GROL, I never worked in the communications field. I've done a lot of electronic DIY projects over the years to at least try and stay current
on very basic skills, but even that homebrew project was at least a
decade ago. The 27 solder joint connections made restore and use PC PSU
was the largest solder job I've done in about two decades. One issue
over that time has been my eyes. Could no longer do the job without
"helping hands" unlike 20 years ago. Let me say that I'm glad you
weren't killed no doubt due to working slowly and methodically.

When I decide to open up the partially defunct Dell PSU, I'll also
proceed slowly. Although my ESR tester can discharge caps, it can only
do so up to a certain point, so I'll take the extra step and manually discharge the electrolytics first.

Are you getting the idea that I wasn't rushing to get at the problem
area?? So the H.T. should have drained away.

Anyway, after removing the cover and (supposedly) shorting the HT rail
to deck, I reached in with my left hand...... and WAAAACCCCKKKK!

My mind told me my Right Hand had reached in and lifted my left arm off
some unknown Voltage point.

The problem with that is that, a couple of years before hand, I was
involved in a Road Accident .... which resulted in my Right Arm being amputated above the elbow ..... so I had NO Right Hand .... BUT MY MIND
TOLD ME MY RIGHT HAND HAD REACHED IN AND LIFTED MY LEFT HAND.

I still have a spot on the inside of my Left Forearm indicating where I
was zapped. I'm guessing the 7500 or 8000 HT might have been developed across two capacitors in series and the knife-switch ONLY shorted the
top point to deck .... and I might have contacted the mid-point .... or something like that!!

Marked for life!! ;-P

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Sun, 3/15/2026 8:31 AM, JR Dolobson wrote:

When I decide to open up the partially defunct Dell PSU, I'll also proceed slowly.ÿ
Although my ESR tester can discharge caps, it can only do so up to a certain point,
so I'll take the extra step and manually discharge the electrolytics first.

Use a *resistor* when draining large caps.

In this schematic, R2 and R3 (220K ohm 1 watt) are the built-in drain resistors.
That's just to give you some idea what kind of resistor would be used for this. Notice the resistor also has a non-trivial power rating -- the more aggressively
you drain a capacitor, the hotter your drain resistor gets, the higher the power rating should be. Work out the tau, R*C time constant, of the 470uF times 220K
to determine how long it will take to discharge in this case.

http://www.pavouk.org/hw/en_atxps.html

Discharging large capacitors improperly, the noise level can be
so great, as to knock out your hearing for ten minutes. Always
estimate the joules stored in the cap, before holding a
shorting wire (with no resistor) near it. For energetic
enough capacitor banks (ultracap bank), your wire can
be blown to hell (3000 amps) and this will be your
shocked face.

Paul




--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

On Sun, 3/15/2026 8:18 AM, Daniel70 wrote:

On 15/03/2026 4:03 am, JR Dolobson wrote:

On 3/13/26 10:39 AM, J. P. Gilliver wrote:

On 2026/3/13 11:52:40, JR Dolobson wrote:
[]

Done!ÿ Made a DIY adapter after finding plug data online.ÿ All working >>>> fine now.ÿ After weeks, the PC is booting up again normally with no
extra switch.ÿ I don't believe this PSU quite matches the specs you
mentioned before, but I believe it was originally for "elementary" gamer >>>> systems.

Glad to hear all is well.

I'll keep it on top of the case for a week or so just to make sure all >>>> is well, then pull the existing one and replace it with this one after >>>> taking note of the ventalation as you mentioned prior.

I honestly didn't think the PSU was to blame.ÿ You did mention it weeks >>>> back after I had to start using the external switch to get boot up.

The original PSU is still working as long as I don't plug in the 24 pin >>>> connector to motherboard.

That has elements of "apart from that, Mrs. Lincoln, did you enjoy the
play?"

At some point, I'll have to open it and see

Be careful; as well as the high voltages (even after off for a while, in >>> theory), a few decades ago I cut myself; I think the normal insides of a >>> PC they're reasonably careful as people might be expected to poke around >>> inside, but the insides of a power supply they seemed to think fewer
people would go, and were more careless about designing how the parts of >>> its case go together. At least, that's what I assumed. Things may have
improved.

One time, I was "hit" with an unexpected charge from a camera flash that had been disassembled.ÿ The electrolytic obviously wasn't discharged before I touched it.ÿ Sent me about halfway across the room.ÿ I know PC and indeed other power supplies can harbor nasty charges long after deactivation.

I can remember, back in about 1980, I was working on a 20kW PEP, Automatic Tune H.F. Radio Transmitter (with 7.500 Volt H.T.) that wasn't auto-tuning correctly. I wanted to make an adjustment in the high power RF cabinet so I switched off the HT. I then switched off the LT. I walked around the back of the 3mtr by 1mtr cabinet and slide open the panelling there and climbed a three or four step ladder then removed the internal cabinet cover. When I removed that cover, a knife-switch was SUPPOSED to short the 7500 or 8000 HT Filter Capacitors to Deck

Are you getting the idea that I wasn't rushing to get at the problem area?? So the H.T. should have drained away.

Anyway, after removing the cover and (supposedly) shorting the HT rail to deck, I reached in with my left hand...... and WAAAACCCCKKKK!

My mind told me my Right Hand had reached in and lifted my left arm off some unknown Voltage point.

The problem with that is that, a couple of years before hand, I was involved in a Road Accident .... which resulted in my Right Arm being amputated above the elbow ..... so I had NO Right Hand .... BUT MY MIND TOLD ME MY RIGHT HAND HAD REACHED IN AND LIFTED MY LEFT HAND.

I still have a spot on the inside of my Left Forearm indicating where I was zapped. I'm guessing the 7500 or 8000 HT might have been developed across two capacitors in series and the knife-switch ONLY shorted the top point to deck .... and I might have contacted the mid-point .... or something like that!!

Marked for life!! ;-P

Each capacitor type, has a penchant for "memory effect".

This is something a technician has to be aware of. It is related
to the dielectric behavior. In fact, a technician should study the
capacitor on the bench, and notice that after you remove the
drain arrangement, the capacitor voltage "jumps back up".

When I noticed this effect while work in my basement as a kid,
it did elicit a "Weird!" declaration. That a drained capacitor
could have a voltage growing on it with time, afterwards. I learned
of the possibility of the effect ("don't turn your back on a capacitor")
in my basement, before taking any classes in it.

Being aware of how capacitors work, is what keeps you alive.

If you discharge something like a vacuum tube (your TV tube), and the
voltage is well and for sure "0 volts", and you remove your draining arrangement, the voltage will immediately start to grow on its own.
For some capacitor types, you have to "drain them five times",
and by then the dielectric is getting the message that we're
serious about draining you.

This means there is the possibility in some situations, that the
drain switch must stay in the "draining position". Then, after
the repair work is complete, you carefully return the switch
to the null position, so you don't "blow the supply when it comes
back on". Checking that the drain resistor is not open-circuit
("blown") is also part of some ceremonies.

Some capacitor types are perfect gentlemen. You discharge those
to zero, remove the drain arrangement, the voltage stays precisely
at zero. There is no dielectric memory on those. There are
other situations though, where a capacitor type is determined to
shock you, and even draining it five times, the capacitor
will have a small standing voltage on it later.

Any time you have an "accident" as an electrical person, it is
very very important to "root cause" what happened. We do this
as intelligent animals... so we don't make the same mistake twice :-/

*******

It took a couple minutes to track this down, but this is a sampling on the topic.

https://wiki.analog.com/university/courses/electronics/electronics-lab-capacitors

"Dielectric Absorption, RDA, CDA:

Monolithic ceramic capacitors are excellent for HF decoupling, but they have
considerable dielectric absorption, which makes them unsuitable for use as the
hold capacitor of a sample-hold amplifier (SHA).

Dielectric absorption is a hysteresis-like internal charge distribution that
causes a capacitor which is quickly discharged and then open-circuited to appear
to recover some of its charge. Since the amount of charge recovered is a function
of its previous charge, this is, in effect, a charge memory and will cause errors
in any SHA where such a capacitor is used as the hold capacitor.

Figure 4 Dielectric Absorption

Capacitors that are recommended for this type of application include the "poly"
type capacitors we spoke about earlier, i.e., polystyrene, polypropylene, or Teflon.
These capacitor types have very low dielectric absorption (typically <0.01%).
"

In some low voltage circuits, that's why you will find a more-expensive capacitor
type in a circuit (when a cheaper one would have worked). It is the low low dielectric absorption that the engineer selected for, and not just the "0.1uF" value the circuit called for. Some instrumentation circuits need this
kind of treatment. Audio circuits also have requirements like this.

We had an engineer at work, who did nothing but answer questions about capacitors. If I said "I need a capacitor with low dielectric absorption",
he would know the answer and not have to look it up on a wiki (like me) :-)

Paul

--- PyGate Linux v1.5.13
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)