Build Your Own Arcade Controls Forum
Main => Everything Else => Topic started by: BadMouth on January 11, 2014, 06:49:16 pm
-
Ever since I've had the Neo Geo Evolution build up and running, there has been a faint burning smell when it is turned on. :-[
The PC runs fine and nothing feels overly hot to the touch, so I chalked it up to MDF dust on the audio amps.
It hasn't faded, so I systematically went through the cab making sure nothing was overheating and sniffing various components. :P
The smell seems to be coming from the video card or the chipset heatsink right below it.
The fan on the video card was making some horrible noises and I couldn't find a replacement, so I replaced the entire gpu cooler with a fanless one from a spare card I had. The copper pipes and aluminum fins weren't hot at all today while testing, but do get pretty hot when running modern pc games.
The smell was there before replacing the heatsink though. I used arctic silver. The chipset heatsink felt pretty darn hot to the touch, but I don't know what is normal.
Anyone deal with something like this before and have advice for tracking it down?
Any recommendations for 3rd party software to track GPU temps? (It's an nvidia 9800gt)
What are normal gpu temps?
-
Take the card out and remove the fan and any excess thermal paste. Clean it right down and put a strong light on the back of the board. Look for any dark patches or brown marks on the chip.
Sounds like you have under cooled the chip and the smell could be the surrounding chips on the board melting from within.
Replace the GPU card with an alternative and store the offending card at room temperature. Try the card again in a weeks time with a heat sink and fan with minimal thermal paste.
I had this on a ATI 1800XT. The sucker got hot very quick when playing modern games. The fan was under performing, so I got an old 386/486 heatsink and fan and secured it. It worked well for several months and it was much quieter too. If you are lucky and have a thermal sensor on your PSU, you can add the fan to the sensor if you do not have a spare molex. The PSU I got had this feature so all my fans are monitored this way. There is also a small pot and you can override the cooling settings. Well worth the money IMO.
-
I agree with Ark. Thermal paste is great stuff, but it needs to be used sparingly. Too much and instead of helping conduct the heat away, it acts like a thermal insulator and prevents heat dissipation. If you are lucky, you just have some baked epoxy on those chips and it's still good.
-
Is it a wood smell or an electronics smell? If the pc is put in an akward position it could be heating up the cabinet itself... not enough to be dangerous, but enough to smell. I've had this happen on projects before... not mame, mostly audio stuff overheating.
-
if your heatsink is HOT, it's doing it's job. don't worry about it.
a cool cooler means you aren't getting that thermal transfer from the GPU.
the 9000 series do run fairly hot... (had an old 9500GT runs about 75 under load)
speedfan has GPU temp capabilities for some cards. (worked for my 9500GT but not my current HD7550)
http://www.almico.com/sfdownload.php (http://www.almico.com/sfdownload.php)
-
the 9000 series do run fairly hot... (had an old 9500GT runs about 75 under load)
I'm assuming you mean 75C? Cause 75F would be nice and cool to the touch.
-
(http://i.imgur.com/3ZidINK.jpg)
::)
-
Thanks. I appreciate the help.
Did not have time to breathe this weekend, but I'll put all your recommendations to use tomorrow evening. :cheers:
Is it a wood smell or an electronics smell? If the pc is put in an akward position it could be heating up the cabinet itself... not enough to be dangerous, but enough to smell. I've had this happen on projects before... not mame, mostly audio stuff overheating.
It's an electronics smell. I was in denial when I was telling myself it was just mdf dust on the amps.
-
I just went thru this with my PC at work. It ended up being the power supply. The fan wasn't working correctly and it was over heating. I replaced the power supply and haven't had any more burning smells.
-
In my experience a burning smell from a PC is most likely the PSU followed by the video card. If possible replace those components individually to see if it solves the problem.
-
In my experience a burning smell from a PC is most likely the PSU followed by the video card. If possible replace those components individually to see if it solves the problem.
This, or a fan...
-
Survey says.........
(http://forum.arcadecontrols.com/index.php?action=dlattach;topic=136713.0;attach=302824;image)
:o Holy ---steaming pile of meadow muffin--- that could have burned down the house!
Ok, so the burning smell wasn't coming from the PC.
In my defense, the power inlet is only a few inches from the video card.
What do I blame here? The cord? (seems to be the most damaged) The inlet? Poor connection between the two?
The pc only has a 430 watt supply, +125w of audio amps, +27" LCD screen (not LED), + a little 12v 2A power brick for the linear actuator.
Those are the only things in the cab drawing power.
-
is there a fuse in that fuse drawer? 100A? wrapped with tinfoil?
Looks like poor contact between the positive pin and cord arced and got hot.
-
is there a fuse in that fuse drawer? 100A? wrapped with tinfoil?
10A 250v
No tin foil.
EDIT: Power inlet says it's rated for 10A 250V, so I'm guessing either the power cable isn't rated for much or the connection at the pins was just bad and arcing.
Looking for a new cable, I see a lot of them are only 18awg. Hell, I think I used thicker than that for my speaker wire. Best I could find was a 3ft 14awg cable, so I ordered it.
The inlet could probably be salvaged, but it smells so I'll get a new one of those too.
-
I haven't seen this on a pc BUT I've seen it on my Xmas lights. When too much power is being pulled through a cord, any break in the connection (the ends, ect) will melt. It won't trip the breaker though because the fault isn't a short, the wires are just too thin.
Do you have a more heavy-duty power cord? I know some of the ones they include with PS these days are rather thin.
-
Do you have a more heavy-duty power cord? I know some of the ones they include with PS these days are rather thin.
Just ordered a 14awg one rated for 15A.
Still shopping for the power inlet.
As commented above, I see most are 18awg.
I didn't even think about the rating on the PC cable, just used one that was laying around.
There was probably more at work than just a shoddy cable, but I'm sure it didn't help.
The female connector in the end of the cable probably loosened up from moving the cab around and swinging down the access door that the inlet is mounted in.
The cable from the power strip in the cab is now running out through the hole where the inlet was.
No smell and nothing feels warm, so I'm sure nothing in the cab is drawing more current than it supposed to.
-
I'd check all your components to be extra safe.
-
(http://i.imgur.com/3ZidINK.jpg)
::)
Thanks for this, Gonna use it in a google thread where I got criticised for forgetting about Fahrenheit temps. They quoted -15 Celsius in the images and then showed a pic of a thermometer showing 2 degrees above 0 which to me made no sense.
-
Survey says.........
(http://forum.arcadecontrols.com/index.php?action=dlattach;topic=136713.0;attach=302824;image)
Yikes! Glad you caught that before anything major happened.
-
Judging by some of the melted areas being a bit further from the actual pin, I'd lean towards arcing.
-
I would replace the whole power supply...something 500 w or greater. No point in risking it. for $50
-
Maybe check for a "snug" fit on the new one anyways? I know you said it might have been lose from the door but it's possible that it was already a poor connection anyways.
-
(http://i.imgur.com/3ZidINK.jpg)
::)
Thanks for this, Gonna use it in a google thread where I got criticised for forgetting about Fahrenheit temps. They quoted -15 Celsius in the images and then showed a pic of a thermometer showing 2 degrees above 0 which to me made no sense.
That map isn't particularly accurate just for the record. It was made up for internet trolling. Tons of places still use the Fahrenheit temps. It's just Europe that doesn't, which is fine, but look how big the US is compared to Europe ....population-wise as well. My point is basically an entire continent of people (ignoring the smaller countries not shown on that map) use the F scale, which is a buttload of people, so it IS relevant FYI.
I will admit that the F scale is dumb, but so is the C scale. I mean yeah it's based around the properties of water.... so what? How is that really relevant? On the other hand the F scale is based upon how cold and hot the weather can typically get on planet earth (or at least the habitable spots on earth at the time of it's invention), so imho it makes more sense for outside temperatures at least. Think about it, in Europe and the US typically it doesn't really get any colder than 0 F in the winter or warmer than 100 F in the summer. There are exceptions but they are few and far between.
-
I will admit that the F scale is dumb, but so is the C scale. I mean yeah it's based around the properties of water.... so what? How is that really relevant? On the other hand the F scale is based upon how cold and hot the weather can typically get on planet earth (or at least the habitable spots on earth at the time of it's invention), so imho it makes more sense for outside temperatures at least. Think about it, in Europe and the US typically it doesn't really get any colder than 0 F in the winter or warmer than 100 F in the summer. There are exceptions but they are few and far between.
true dat.
-
Fahrenheit scale (as well as all the other forms of measure the US uses) is absolutely ridiculous and random.
Fahrenheit. arbitrary start point. arbitrary end point. based on...lowest attainable temperature of...saltwater?. okay. It's something.
inches. based on being 1/12th of the random size chosen for the foot. eeegh, this looks good. :dunno
Pound. based on 6992 troy grains. a grain being the mass of a single seed of a cereal. hopefully the seeds are all the same size and weight. NOT.
metric is all based on the same thing and is universally reproducible.
a gram of water is exactly 1cmx1cmx1cm in size. the freezing point of pure water is exactly 0c. and it boils (at sea level) at 100c.
so a kilogram is 1000 1x1x1 cubes of water in volume (a kilogram.)
a meter is 1000 1x1x1cubes of water all lined up in a row.
-
No it's not ridiculous and random. I literally just explained to you in my last point what it's based upon. The F scale is based upon outside weather temperatures, which let's face it is the only temperature that the average Joe really needs to deal with on a day to day basis. And in that respect it gives a far more sensible reading than C. I mean the scale was invented for thermometers. How can you possibly say that a thermometer where 0 is the coldest it will probably ever get outside and 100 is as hot as it will probably get doesn't make any sense?
The imperial system's weight is based upon commerce... in other words it's useful. Look on a box of food sometime. Typically it'll come in imperial sizes (1 pound, 1/4 pound ect ect) but if you look at the grams it rarely comes out even. If the measurement system comes out even for common sizes of things then it's well calibrated, if not then... well it isn't suited for the task.
Likewise in length measurement. It was based around carpentry, particularly home and ship building. You'll find that for these types of measurements, which again, are the only ones a regular person is going to deal with on a day to day basis, the imperial system is best. The 16 segments in a unit thing is particularly important. If I remember correctly it's based upon the thickness of wood grain you can consistently shave off via a chisel.
Miles are a totally different scale. This is because they were calibrated to a totally different task. I'm not as familiar with their origins so maybe someone else can fill it in.
The metric system was invented for scientific research, because as you pointed out, it scales well. And if you are a scientist in the US you will of course be using the metric system.
It isn't well suited for our daily lives though... I'm not sure how the rest of the world got conned into the metric system when we'd been using the imperial system for centuries to be honest.
-
Think about it, in Europe and the US typically it doesn't really get any colder than 0 F in the winter or warmer than 100 F in the summer. There are exceptions but they are few and far between.
You need to experience some UK weather Howard. :lol
-
Seriously. Come to Michigan. Instead of the 'Winter Wonderland' state I believe we should be called the 'Why is it 55 degrees, snowing, and the sun is shining?' state.
-
No it's not ridiculous and random. I literally just explained to you in my last point what it's based upon. The F scale is based upon outside weather temperatures, which let's face it is the only temperature that the average Joe really needs to deal with on a day to day basis. And in that respect it gives a far more sensible reading than C. I mean the scale was invented for thermometers. How can you possibly say that a thermometer where 0 is the coldest it will probably ever get outside and 100 is as hot as it will probably get doesn't make any sense?
huuugh. :-\
http://en.wikipedia.org/wiki/Fahrenheit (http://en.wikipedia.org/wiki/Fahrenheit)
Fahrenheit (symbol °F) is a temperature scale based on one proposed in 1724 by the physicist Daniel Gabriel Fahrenheit (1686–1736), after whom the scale is named.[1] On Fahrenheit's original scale the lower defining point was the lowest temperature which he could reproducibly obtain using brine (defining zero degrees), while the highest was the best estimate of the average human body temperature (defining 96 degrees).
The imperial system's weight is based upon commerce... in other words it's useful. Look on a box of food sometime. Typically it'll come in imperial sizes (1 pound, 1/4 pound ect ect) but if you look at the grams it rarely comes out even. If the measurement system comes out even for common sizes of things then it's well calibrated, if not then... well it isn't suited for the task.
I live in Canada, all my food has grams/liters on them. No ounces. No pounds. No Spanish. (French though)
yes for lots of things we still use imperial. i cut crap in feet and inches. :dunno i cook stuff in the oven at 350F. :dunno i buy a 4 foot by 8 foot sheet of drywall. :dunno studs are 8 feet long. 16 inch centers... but seriously, it's only because that's how it's supplied and that's how it's still done.
studs are 243.84 centimeters. :banghead: and drywall is 121.92 x 243.84 centimeters :banghead: see, it sounds retarded because that's how it comes and it's based on imperial measurements.
if studs came 250 centimeters long they would be 8 feet 2 and 27/64th inches long. now who sounds retarded? :angry:
it doesn't come out "even" as you say because it's not the same thing. it CAN'T come out even.
Likewise in length measurement. It was based around carpentry, particularly home and ship building. You'll find that for these types of measurements, which again, are the only ones a regular person is going to deal with on a day to day basis, the imperial system is best. The 16 segments in a unit thing is particularly important. If I remember correctly it's based upon the thickness of wood grain you can consistently shave off via a chisel.
HURK! :-[
http://en.wikipedia.org/wiki/Foot_(length) (http://en.wikipedia.org/wiki/Foot_(length))
Historically the foot, which was used in Ancient Greece, Ancient Rome, England, Scotland and many Continental European countries and which varied from country to country and in some cases from city to city, was part local systems of units. Its length was usually between 250 mm and 335 mm and was generally, but not always, subdivided into 12 inches or 16 digits. Historically the human body has been used to provide the basis for units of length.[34] The foot of a Caucasian male is typically about 15.3% of his height,[35] giving a person of 160 cm (5 ft 3 in) a foot of 245 mm and one of 180 cm (5 ft 11 in) a foot of 275 mm. These figures are less than the foot used in most cities over time, suggesting that the "foot" was actually a synonym for a "shoe".
Miles are a totally different scale. This is because they were calibrated to a totally different task. I'm not as familiar with their origins so maybe someone else can fill it in.
seriously? it's the same as the foot. they just picked a random amount of feet...egh, that's a mile.
http://en.wikipedia.org/wiki/Mile (http://en.wikipedia.org/wiki/Mile)
The metric system was invented for scientific research, because as you pointed out, it scales well. And if you are a scientist in the US you will of course be using the metric system.
It isn't well suited for our daily lives though... I'm not sure how the rest of the world got conned into the metric system when we'd been using the imperial system for centuries to be honest.
it's only impractical to you and your "daily life" because you don't use it. if you use it, it's fine. This is probably why Americans think we live in igloos and drive dogsleds when we talk of -10 weather outside. ---steaming pile of meadow muffin---, -10 is t-shirt weather for some of us.
if you read through the wikipedia articles i linked you will notice one VERY important commonality between them all. They aren't standard. Depending on country...depending on state...depending on city...depending even on just the person, a mile is a different length...a pound is different weight...an ounce is different volume. When it comes to trading fairly, close enough is not close enough. A pound of flour must be the same pound everywhere. But it isn't.
a liter is always a liter no matter who you are or where you are. 1000 1x1x1cm cubes of water will always be 1 liter whether you sell milk, vodka, blood, or gas. that very same liter of water weighs 1 kilogram whether you are in antarctica, russia, albania, or rome. those 1000 1x1x1cm cubes of water lined up will always be a meter whether you build a house, car, or submarine sandwich.
it's consistent, repeatable, and based on the same tangible thing.
now, all i need to do is convince everyone to metricise time. (100 seconds to a minute...100 minutes in an hour...etc.) :dunno