[gentoo-user] Filthy oscilloscope picture! =P

[gentoo-user] Filthy oscilloscope picture! =P

[Alan Grimes]

[-- Attachment #1: Type: text/plain, Size: 1129 bytes --]

Isn't this the filthiest oscilloscope u've seen recently?

The only bare metal contact that I could safely use to get a reading off
was a +12v line on a spare PCI-E gpu plug. The ground reference is the
chassis.

You can see the machine's settings in the photo clearly enough. The
waveform is fairly constant, it stays in this mode most of the time but
sometimes goes into a "low ripple" mode where the ripple falls to +/-
20mv and holds tight. The scaling indicates the upward spikes are around
0.120 volts and the downward spikes are about 0.22 volts.  This
__SHOULD__ be within the input tolerances of the motherboard's regulators.

I would call this PSU marginal, it absolutely does power the machine but
it's noise output is a bit larger than what I would prefer.

Given that i'm flat on my ass broke with a foreclosure over my head, I
am powerfully inclined to continue to live with the PSU the way it is
now until it is no longer possible to do so.

I had to use my windows 7 machine to get the photo off my camera because
digikam does not compile. =|

-- 
IQ is a measure of how stupid you feel.

Powers are not rights.


[-- Attachment #2: IMG_0012.JPG --]
[-- Type: image/jpeg, Size: 394567 bytes --]

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Alan McKinnon]

On 22/08/2015 21:19, Alan Grimes wrote:
> Isn't this the filthiest oscilloscope u've seen recently?

I've seen cleaner. And dirtier.

> The only bare metal contact that I could safely use to get a reading off
> was a +12v line on a spare PCI-E gpu plug. The ground reference is the
> chassis.
> 
> You can see the machine's settings in the photo clearly enough. The
> waveform is fairly constant, it stays in this mode most of the time but
> sometimes goes into a "low ripple" mode where the ripple falls to +/-
> 20mv and holds tight. The scaling indicates the upward spikes are around
> 0.120 volts and the downward spikes are about 0.22 volts.  This
> __SHOULD__ be within the input tolerances of the motherboard's regulators.
> 
> I would call this PSU marginal, it absolutely does power the machine but
> it's noise output is a bit larger than what I would prefer.

I would call that PSU on it's last legs, and highly likely to be the
root cause for the recent difficulties you've posted about and possibly
more too.  +100mV/-200mV is excessive

> Given that i'm flat on my ass broke with a foreclosure over my head, I
> am powerfully inclined to continue to live with the PSU the way it is
> now until it is no longer possible to do so.

Well now you put it that way, you don't have many options other than use
what you've got.

But do realise that the next time you run into some weird issue, that
PSU is most likely what you are dealing with as root cause.


> I had to use my windows 7 machine to get the photo off my camera because
> digikam does not compile. =|



-- 
Alan McKinnon
alan.mckinnon@gmail.com

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Fernando Rodriguez]

On Saturday, August 22, 2015 3:19:50 PM Alan Grimes wrote:
> Isn't this the filthiest oscilloscope u've seen recently?
> 
> The only bare metal contact that I could safely use to get a reading off
> was a +12v line on a spare PCI-E gpu plug. The ground reference is the
> chassis.
> 
> You can see the machine's settings in the photo clearly enough. The
> waveform is fairly constant, it stays in this mode most of the time but
> sometimes goes into a "low ripple" mode where the ripple falls to +/-
> 20mv and holds tight. The scaling indicates the upward spikes are around
> 0.120 volts and the downward spikes are about 0.22 volts.  This
> __SHOULD__ be within the input tolerances of the motherboard's regulators.

Regulators don't filter noise, they introduce it. Capacitors do that as 
somebody pointed on the other thread.

So if you're on a tight budget and you have an electronics surplus store 
nearby you can replace all the capacitors on your mobo and PSU (except the big 
bulky ones on the PSU) for about $3.
 
> I would call this PSU marginal, it absolutely does power the machine but
> it's noise output is a bit larger than what I would prefer.
> 
> Given that i'm flat on my ass broke with a foreclosure over my head, I
> am powerfully inclined to continue to live with the PSU the way it is
> now until it is no longer possible to do so.
> 
> I had to use my windows 7 machine to get the photo off my camera because
> digikam does not compile. =|
> 
> 

-- 
Fernando Rodriguez

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Mick]

[-- Attachment #1: Type: Text/Plain, Size: 2583 bytes --]

On Sunday 23 Aug 2015 01:11:03 Fernando Rodriguez wrote:
> On Saturday, August 22, 2015 3:19:50 PM Alan Grimes wrote:
> > Isn't this the filthiest oscilloscope u've seen recently?
> > 
> > The only bare metal contact that I could safely use to get a reading off
> > was a +12v line on a spare PCI-E gpu plug. The ground reference is the
> > chassis.
> > 
> > You can see the machine's settings in the photo clearly enough. The
> > waveform is fairly constant, it stays in this mode most of the time but
> > sometimes goes into a "low ripple" mode where the ripple falls to +/-
> > 20mv and holds tight. The scaling indicates the upward spikes are around
> > 0.120 volts and the downward spikes are about 0.22 volts.  This
> > __SHOULD__ be within the input tolerances of the motherboard's
> > regulators.
> 
> Regulators don't filter noise, they introduce it. Capacitors do that as
> somebody pointed on the other thread.
> 
> So if you're on a tight budget and you have an electronics surplus store
> nearby you can replace all the capacitors on your mobo and PSU (except the
> big bulky ones on the PSU) for about $3.

It is quite likely that only the secondary circuit on the PSU needs to have 
its electrolytic capacitors replaced.  We're talking of anything between one 
to half a dozen of capacitors.  In all likelihood less than a $1 to $3.  If 
any are even slightly domed I'd start with those and spend no more than a few 
cents.

Primary circuit ceramic capacitors (transient protection) could have been 
affected if the PSU was submitted to high surges in the mains supply.  I had 
one go bad on me after sheet lightning hit the area once.  Its replacement 
along with a resistor fixed the PSU without any further problems and to much 
of my surprise - I thought it was a gonner!

Domed capacitors on the MoBo is a different story.  Quite likely other 
components would have been affected and many of them are surface mounted.  
You'll need a magnifying glass and steady hands for those.  It is not 
something I would attempt in haste, as it is easy to damage more components 
than what you fix on a MoBo.  YMMV.

PS. Noisy PSUs are nothing new.  The noise is can be caused by the capacitors, 
or the coils.  Although annoying it does not necessarily mean that there is an 
electrical problem with the components.  If the fan is rattling, then a drop 
of oil on its bearing should soon put a stop to this.  As Dale mentioned, a 
stalled fan will not help the longevity of the remaining components.  :-)

-- 
Regards,
Mick

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Dale]

Mick wrote:
> PS. Noisy PSUs are nothing new. The noise is can be caused by the
> capacitors, or the coils. Although annoying it does not necessarily
> mean that there is an electrical problem with the components. If the
> fan is rattling, then a drop of oil on its bearing should soon put a
> stop to this. As Dale mentioned, a stalled fan will not help the
> longevity of the remaining components. :-) 


This is true, they do sometimes make noise.  What got my attention is
that it stopped.  Usually if one makes noise, it will make it for the
remainder of its life cycle unless something changes, such has adding
additional load which may change a frequency.  If nothing changes and it
stops making that noise, then it likely stopped working as well.  I've
seen capacitors go bad and it not blow anything else out or stop the
rest of the circuit from working.  It all depends on how it is used in
the circuit.  Most likely in a case like this, if the capacitor just
went open circuit, the rest of the circuit will continue to work but the
capacitor is no longer doing its job of filtering out the noise and
ripple. 

Still, the mention that it stopped is what got my attention.  That
doesn't sound good, pardon the pun.  ;-) 

Dale

:-)  :-) 

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Fernando Rodriguez]

On Sunday, August 23, 2015 12:14:58 PM Mick wrote:
> On Sunday 23 Aug 2015 01:11:03 Fernando Rodriguez wrote:
> > On Saturday, August 22, 2015 3:19:50 PM Alan Grimes wrote:
> > > Isn't this the filthiest oscilloscope u've seen recently?
> > > 
> > > The only bare metal contact that I could safely use to get a reading off
> > > was a +12v line on a spare PCI-E gpu plug. The ground reference is the
> > > chassis.
> > > 
> > > You can see the machine's settings in the photo clearly enough. The
> > > waveform is fairly constant, it stays in this mode most of the time but
> > > sometimes goes into a "low ripple" mode where the ripple falls to +/-
> > > 20mv and holds tight. The scaling indicates the upward spikes are around
> > > 0.120 volts and the downward spikes are about 0.22 volts.  This
> > > __SHOULD__ be within the input tolerances of the motherboard's
> > > regulators.
> > 
> > Regulators don't filter noise, they introduce it. Capacitors do that as
> > somebody pointed on the other thread.
> > 
> > So if you're on a tight budget and you have an electronics surplus store
> > nearby you can replace all the capacitors on your mobo and PSU (except the
> > big bulky ones on the PSU) for about $3.
> 
> It is quite likely that only the secondary circuit on the PSU needs to have 
> its electrolytic capacitors replaced.  We're talking of anything between one 
> to half a dozen of capacitors.  In all likelihood less than a $1 to $3.  If 
> any are even slightly domed I'd start with those and spend no more than a 
few 
> cents.
> 
> Primary circuit ceramic capacitors (transient protection) could have been 
> affected if the PSU was submitted to high surges in the mains supply.  I had 
> one go bad on me after sheet lightning hit the area once.  Its replacement 
> along with a resistor fixed the PSU without any further problems and to much 
> of my surprise - I thought it was a gonner!
> 
> Domed capacitors on the MoBo is a different story.  Quite likely other 
> components would have been affected and many of them are surface mounted.  
> You'll need a magnifying glass and steady hands for those.  It is not 
> something I would attempt in haste, as it is easy to damage more components 
> than what you fix on a MoBo.  YMMV.

I don't think it's very likely to have damanged something else if it's just 
noise, but then again I'm not an electronics engineer, this is just a hobby of 
mine so you may be right. Though I can tell you that I have gotten a few 
damaged boards to work like new by just replacing the electrolitic caps.
 
> PS. Noisy PSUs are nothing new.  The noise is can be caused by the 
capacitors, 
> or the coils.  Although annoying it does not necessarily mean that there is 
an 
> electrical problem with the components.  If the fan is rattling, then a drop 
> of oil on its bearing should soon put a stop to this.  As Dale mentioned, a 
> stalled fan will not help the longevity of the remaining components.  :-)

Agreed.

-- 
Fernando Rodriguez

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Alan McKinnon]

On 23/08/2015 22:24, Fernando Rodriguez wrote:
> On Sunday, August 23, 2015 12:14:58 PM Mick wrote:
>> On Sunday 23 Aug 2015 01:11:03 Fernando Rodriguez wrote:
>>> On Saturday, August 22, 2015 3:19:50 PM Alan Grimes wrote:
>>>> Isn't this the filthiest oscilloscope u've seen recently?
>>>>
>>>> The only bare metal contact that I could safely use to get a reading off
>>>> was a +12v line on a spare PCI-E gpu plug. The ground reference is the
>>>> chassis.
>>>>
>>>> You can see the machine's settings in the photo clearly enough. The
>>>> waveform is fairly constant, it stays in this mode most of the time but
>>>> sometimes goes into a "low ripple" mode where the ripple falls to +/-
>>>> 20mv and holds tight. The scaling indicates the upward spikes are around
>>>> 0.120 volts and the downward spikes are about 0.22 volts.  This
>>>> __SHOULD__ be within the input tolerances of the motherboard's
>>>> regulators.
>>>
>>> Regulators don't filter noise, they introduce it. Capacitors do that as
>>> somebody pointed on the other thread.
>>>
>>> So if you're on a tight budget and you have an electronics surplus store
>>> nearby you can replace all the capacitors on your mobo and PSU (except the
>>> big bulky ones on the PSU) for about $3.
>>
>> It is quite likely that only the secondary circuit on the PSU needs to have 
>> its electrolytic capacitors replaced.  We're talking of anything between one 
>> to half a dozen of capacitors.  In all likelihood less than a $1 to $3.  If 
>> any are even slightly domed I'd start with those and spend no more than a 
> few 
>> cents.
>>
>> Primary circuit ceramic capacitors (transient protection) could have been 
>> affected if the PSU was submitted to high surges in the mains supply.  I had 
>> one go bad on me after sheet lightning hit the area once.  Its replacement 
>> along with a resistor fixed the PSU without any further problems and to much 
>> of my surprise - I thought it was a gonner!
>>
>> Domed capacitors on the MoBo is a different story.  Quite likely other 
>> components would have been affected and many of them are surface mounted.  
>> You'll need a magnifying glass and steady hands for those.  It is not 
>> something I would attempt in haste, as it is easy to damage more components 
>> than what you fix on a MoBo.  YMMV.
> 
> I don't think it's very likely to have damanged something else if it's just 
> noise, but then again I'm not an electronics engineer, this is just a hobby of 
> mine so you may be right. Though I can tell you that I have gotten a few 
> damaged boards to work like new by just replacing the electrolitic caps.

That's quite normal - electrolytic caps are the only electronic
components that can be considered to "wear out". Apart from batteries of
course :-)

Getting the caps off modern motherboards is a real PITA though - surface
mount caps need semi-specialized equipment: a proper soldering iron or
hot air pencil with a very fine tip, desolder braid, a magnifier and a
very steady hand

>  
>> PS. Noisy PSUs are nothing new.  The noise is can be caused by the 
> capacitors, 
>> or the coils.  Although annoying it does not necessarily mean that there is 
> an 
>> electrical problem with the components.  If the fan is rattling, then a drop 
>> of oil on its bearing should soon put a stop to this.  As Dale mentioned, a 
>> stalled fan will not help the longevity of the remaining components.  :-)

I recall an ancient TV from the mid '70s (Blaupunkt) that would
sometimes develop a rattle in one of the drive circuit coils. Damn thing
would sound like a hive of bees inside the cabinet!

> 
> Agreed.



-- 
Alan McKinnon
alan.mckinnon@gmail.com

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Fernando Rodriguez]

On Sunday, August 23, 2015 10:34:20 PM Alan McKinnon wrote:
> On 23/08/2015 22:24, Fernando Rodriguez wrote:
> > On Sunday, August 23, 2015 12:14:58 PM Mick wrote:
> >> On Sunday 23 Aug 2015 01:11:03 Fernando Rodriguez wrote:
> >>> On Saturday, August 22, 2015 3:19:50 PM Alan Grimes wrote:
> >>>> Isn't this the filthiest oscilloscope u've seen recently?
> >>>>
> >>>> The only bare metal contact that I could safely use to get a reading off
> >>>> was a +12v line on a spare PCI-E gpu plug. The ground reference is the
> >>>> chassis.
> >>>>
> >>>> You can see the machine's settings in the photo clearly enough. The
> >>>> waveform is fairly constant, it stays in this mode most of the time but
> >>>> sometimes goes into a "low ripple" mode where the ripple falls to +/-
> >>>> 20mv and holds tight. The scaling indicates the upward spikes are 
around
> >>>> 0.120 volts and the downward spikes are about 0.22 volts.  This
> >>>> __SHOULD__ be within the input tolerances of the motherboard's
> >>>> regulators.
> >>>
> >>> Regulators don't filter noise, they introduce it. Capacitors do that as
> >>> somebody pointed on the other thread.
> >>>
> >>> So if you're on a tight budget and you have an electronics surplus store
> >>> nearby you can replace all the capacitors on your mobo and PSU (except 
the
> >>> big bulky ones on the PSU) for about $3.
> >>
> >> It is quite likely that only the secondary circuit on the PSU needs to 
have 
> >> its electrolytic capacitors replaced.  We're talking of anything between 
one 
> >> to half a dozen of capacitors.  In all likelihood less than a $1 to $3.  
If 
> >> any are even slightly domed I'd start with those and spend no more than a 
> > few 
> >> cents.
> >>
> >> Primary circuit ceramic capacitors (transient protection) could have been 
> >> affected if the PSU was submitted to high surges in the mains supply.  I 
had 
> >> one go bad on me after sheet lightning hit the area once.  Its 
replacement 
> >> along with a resistor fixed the PSU without any further problems and to 
much 
> >> of my surprise - I thought it was a gonner!
> >>
> >> Domed capacitors on the MoBo is a different story.  Quite likely other 
> >> components would have been affected and many of them are surface mounted.  
> >> You'll need a magnifying glass and steady hands for those.  It is not 
> >> something I would attempt in haste, as it is easy to damage more 
components 
> >> than what you fix on a MoBo.  YMMV.
> > 
> > I don't think it's very likely to have damanged something else if it's 
just 
> > noise, but then again I'm not an electronics engineer, this is just a 
hobby of 
> > mine so you may be right. Though I can tell you that I have gotten a few 
> > damaged boards to work like new by just replacing the electrolitic caps.
> 
> That's quite normal - electrolytic caps are the only electronic
> components that can be considered to "wear out". Apart from batteries of
> course :-)
> 
> Getting the caps off modern motherboards is a real PITA though - surface
> mount caps need semi-specialized equipment: a proper soldering iron or
> hot air pencil with a very fine tip, desolder braid, a magnifier and a
> very steady hand

For the tiny SMT ones I usually use an worn out iron tip (cause it may get 
plastic on it), heat the whole thing up and push it aside if there's room, 
then pull them off with twizzers and a little bit for force, clean up the 
contacts with braid. If they're many I use solder paste and an oven the get 
new ones on.

But usually there's still a few through hole electrolytics (at least on boards 
old enough to be failing) and those are the ones that fail. When they're SMT 
it's usually a relatively big one or an SMT can and I only seen those fail on 
homemade or dev boards when I do something stupid. For the canned ones I heat 
the can up until it comes off. The real PITA with those is that you usually 
don't find those at a local store.
 
> >  
> >> PS. Noisy PSUs are nothing new.  The noise is can be caused by the 
> > capacitors, 
> >> or the coils.  Although annoying it does not necessarily mean that there 
is 
> > an 
> >> electrical problem with the components.  If the fan is rattling, then a 
drop 
> >> of oil on its bearing should soon put a stop to this.  As Dale mentioned, 
a 
> >> stalled fan will not help the longevity of the remaining components.  :-)
> 
> I recall an ancient TV from the mid '70s (Blaupunkt) that would
> sometimes develop a rattle in one of the drive circuit coils. Damn thing
> would sound like a hive of bees inside the cabinet!

-- 
Fernando Rodriguez

[gentoo-user] Re: Filthy oscilloscope picture! =P

[Grant Edwards]

On 2015-08-23, Alan McKinnon <alan.mckinnon@gmail.com> wrote:

> That's quite normal - electrolytic caps are the only electronic
> components that can be considered to "wear out". Apart from batteries
> of course :-)

And various forms of non-volatile memory (EEPROM, NAND flash, etc.)
that can withstand a limited number erase/program cycles.

-- 
Grant Edwards               grant.b.edwards        Yow! I'm pretending that
                                  at               we're all watching PHIL
                              gmail.com            SILVERS instead of RICARDO
                                                   MONTALBAN!

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 427 bytes --]

On Sat, 22 Aug 2015 15:19:50 -0400, Alan Grimes wrote:

> I had to use my windows 7 machine to get the photo off my camera because
> digikam does not compile. =|

What's wrong with putting the camera in mass storage mode, or putting
it's memory card into your computer?

Compiling DigiKam to copy one file brings new meaning to overkill.

 
-- 
Neil Bothwick

Geordi, show these children the antimatter - Picard

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [gentoo-user] Filthy oscilloscope picture! =P

[Alan McKinnon]

On 25/08/2015 23:33, Neil Bothwick wrote:
> On Sat, 22 Aug 2015 15:19:50 -0400, Alan Grimes wrote:
> 
>> I had to use my windows 7 machine to get the photo off my camera because
>> digikam does not compile. =|
> 
> What's wrong with putting the camera in mass storage mode, or putting
> it's memory card into your computer?
> 
> Compiling DigiKam to copy one file brings new meaning to overkill.
> 
>  
> 


I believe that honour goes to running wubi under wine on an Ubuntu guest
VM in VBox/VMWare/KVM, running on an Ubuntu host.

What do you get when you make it through that labyrinth? Why, it
installs Ubuntu of course!

-- 
Alan McKinnon
alan.mckinnon@gmail.com