Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

kirkbross

macrumors 6502a
Original poster
Mar 6, 2007
666
22
Los Angeles
I have a 2008 Mac Pro and and my system intermittently shows the full 8GB of RAM installed... other boots it's 7GB... sometimes 6GB or 4GB.

I've read other threads and done DIMM swapping and reseating and so forth and I have narrowed it down to (I believe) a faulty Riser A as the source of the problem.

Before I drop $100 for a new memory riser, any suggestions on how to clean the DIMM slots and/or the DIMMs themselves and whether that is likely to make a difference?
 
The memory riser cards (and any smaller, secondary boards within a system) are also referred to as "daughter cards" ... I think mine may have expired.
 
I have a 2008 Mac Pro and and my system intermittently shows the full 8GB of RAM installed... other boots it's 7GB... sometimes 6GB or 4GB.

I've read other threads and done DIMM swapping and reseating and so forth and I have narrowed it down to (I believe) a faulty Riser A as the source of the problem.

Before I drop $100 for a new memory riser, any suggestions on how to clean the DIMM slots and/or the DIMMs themselves and whether that is likely to make a difference?

The most perfect way to clean the edge connectors is to use cotton ear swabs and acetone. With this method you have to be careful no to load the cotton tips up too heavily. You do not want acetone on the plastic encasings and it's probably not a good idea to have much if any drip onto the PWB (motherboard). This means you have to load and bleed the cotton heads for every 3 or 4 pin pairs you clean.

A little safer way but not as perfect, is to use a toothbrush and rubbing alcohol in around a 60-70% solution with waiter. And then canned air to blow out any excess and repeat 2 or 3 times. In both cases wait after finishing about 20 or 30 minutes to make sure all is evaporated.

The above are for the female receptacle edge connectors. For the male card edges inspect them with a magnifying glass to see if there are even the mildest scratches. If there are (and there probably will be) use a pink pencil eraser and go over both sides of the gold plated contacts. Back and forth as if you were erasing pencil marks normally. When finished use the acetone applied to some tish (or better a linen cloth) and give both sides a few gentle swipes.

That's it. That should do it perfectly.

It's probably not dirt though. Those connections are pretty robust with lots of contact area. Still, I agree with the logic of cleaning before spending. :)
 
Last edited:
It's probably not dirt though. Those connections are pretty robust with lots of contact area. Still, I agree with the logic of cleaning before spending. :)
Thank you so much! I have some 99% alcohol... is that not good?

I agree it's probably not dirt... it would have to be severely dirty to actually block a connection.

I once baked my (dead) video card in the oven to re-flow the solder and it worked. I'm wondering if there is a similar phenomenon at play here -- that there are micro fissures in the solder of the riser card. Do you suppose baking it might be worth a try?
 
Thank you so much! I have some 99% alcohol... is that not good?

That's fine... just add ~30% water to it. :)

I agree it's probably not dirt... it would have to be severely dirty to actually block a connection.

I once baked my (dead) video card in the oven to re-flow the solder and it worked. I'm wondering if there is a similar phenomenon at play here -- that there are micro fissures in the solder of the riser card. Do you suppose baking it might be worth a try?

Highly doubtful and I dunno about the DIMM seats, they melt? The video card one involved removing all melatable parts first right? But if you wanna know for sure before attempting it buy a 6x loupe on ebay for $5 and check out the solder connections. It you see any cracks then you'll know.

If yo do see any tho I think I would just buy a replacement anyway. aren't they actually cheaper than $100? I would think closer to $30...

http://www.ebay.com/sch/i.html?_trk...TR0.TRC0&_nkw=macpro+riser&_sacat=0&_from=R40
 
Last edited:
Thank you so much! I have some 99% alcohol... is that not good?
This is fine.

DO NOT ADD WATER. Isopropyl alcohol, aka IPA, is typically used at 99.5% or better in the electronics industry as both a cleaner and solvent.

Another trick, is to use a white plastic draftsman eraser. Then clean with alcohol.

I agree it's probably not dirt... it would have to be severely dirty to actually block a connection.
You'd be surprised, particularly with high speed signals and small contacts & pads. Contact resistance can be a real problem.

I once baked my (dead) video card in the oven to re-flow the solder and it worked. I'm wondering if there is a similar phenomenon at play here -- that there are micro fissures in the solder of the riser card. Do you suppose baking it might be worth a try?
Possible (lead free solder alloys suck, and is more prone to bad joints). A good close-up pic would help.

It may also be bad contact pins in the riser board connectors (relaxed over time, and not making a gas tight connection, aka cold weld).
 
Great info nanofrog... I haven't done the cleaning yet but I booted up recently and now it shows 7 GB... with two of the 1GB DIMMs only registering 512MB... very odd. I thought DIMMs were either alive or dead. Period. Not half alive.
 
Great info nanofrog... I haven't done the cleaning yet but I booted up recently and now it shows 7 GB... with two of the 1GB DIMMs only registering 512MB... very odd. I thought DIMMs were either alive or dead. Period. Not half alive.

this is a protection method from the bios/efi. one of them is likely problematic - is it showing a red LED on the riser? i've had that happen with 1, 2, and 4 gig modules. very frustrating.
 
Clean both the riser contacts as well as each DIMM, then test the problematic sticks out one at a time. Assuming they come up OK (ideally, I recommend testing every single stick individually), install all sticks back in the riser, and see if a fault LED lights up or not, and if the memory capacity is showing what's installed.
 
This is fine.

DO NOT ADD WATER. Isopropyl alcohol, aka IPA, is typically used at 99.5% or better in the electronics industry as both a cleaner and solvent.

And in recent times they also use good old soap and water. Especially when cleaning is a prescribed step in the manufacturing / fabrication process. Tests show it works just as well, doesn't need expensive waste disposal, and it makes the greenies happier. :)

Water can dissolve or remove some debris which highly concentrated alcohol alone can not. ;) By mixing the two you're creating a super cleaner of sorts.

Additionally unless you're using a pressure spray technique alone (like they do in the manufacturing line) 99% alcohol may actually be a bad thing. High concentrations of alcohol rapidly removes moisture which in turn will rapidly break down the microstructure of many materials (cloth, tish, plastics, some anodizations, etc.) which end up as debris on the very parts you're attempting to clean. You can sometimes ever see the affects of this on hard materials such as aluminum, some paints, and many kinds of plastics - where the hard material turns white. What's happening is that the microstructure of the material came apart and created a film of broken microstructures across the surface. By adding 30% to 40% water this process is counteracted.

For the last couple of years I've been cleaning and refurbishing antique and legacy camera lenses for a living. These little facts become very apparent when working with clear glass under high magnifications - as I do.

Another trick, is to use a white plastic draftsman eraser. Then clean with alcohol.

This is a different thing than the pink eraser tho. Pink erasers have a very fine grit which serves to remove micro-scratches in the gold pads as well as baked on hardened greases and oils that may be present. White erasers will only remove (and smudge) softer oils and greases.
 
Last edited:
And in recent times they also use good old soap and water. Especially when cleaning is a prescribed step in the manufacturing / fabrication process. Tests show it works just as well, doesn't need expensive waste disposal, and it makes the greenies happier. :)
I wasn't indicating that there aren't instances where water was used, but in this specific case, you want the rapid evaporation (no water in the connector contacts).

And as per water itself, it needs to be deionized water (usually obtained by distillation), not tap water. The reason is so no minerals, aka ions, can be left behind which cause problems.

Where it's used primarily, along with a saponifier (soap), is for cleaning water soluble flux off of rapid soldered boards (reflow oven or wave process). WS flux is nasty stuff (highly corrosive), so it must be cleaned off, and quickly once the soldering process is completed.

Water can dissolve or remove some debris which highly concentrated alcohol alone can not. ;) By mixing the two you're creating a super cleaner of sorts.
Water is a good solvent, but not in this case, and it's not the most commonly used in electronics. IPA and other alcohols are more prevalent in electronics manufacture however, due to the chemistry being removed (non water soluble oils and greases, oxidation prevention, or as a thinner in non water soluble chemicals, such as rosin based flux).

Mixing alcohol with xylene makes a super cleaner (80/20% split respectively). ;) Give it a try if you don't believe me (better than straight IPA, but not as aggressive as acetone, so safer for plastics). :D

Additionally unless you're using a pressure spray technique alone (like they do in the manufacturing line) 99% alcohol may actually be a bad thing. High concentrations of alcohol rapidly removes moisture which in turn will rapidly break down the microstructure of many materials (cloth, tish, plastics, some anodizations, etc.) which end up as debris on the very parts you're attempting to clean. You can sometimes ever see the affects of this on hard materials such as aluminum, some paints, and many kinds of plastics - where the hard material turns white. What's happening is that the microstructure of the material came apart and created a film of broken microstructures across the surface. By adding 30% to 40% water this process is counteracted.

For the last couple of years I've been cleaning and refurbishing antique and legacy camera lenses for a living. These little facts become very apparent when working with clear glass under high magnifications - as I do.
There shouldn't be any moisture in the PCB or contacts at all to cause this problem.

I use IPA all of the time, and though I'm aware of what you speak of, it's not that common a problem for electronics if the exposure time is limited to a short duration (rapid evaporation; do not let it soak). Heavier deposits have to be run multiple times, assisted with mechanical abrasion (ESD plastic brush for example), until it's removed rather than soaking in IPA. It's also a good idea to perform a test run to be sure the component packages (plastic usually), can handle the processes and specific chemistries being used.

All of this is rather overkill for what the OP needs though. ;) :D

This is a different thing than the pink eraser tho. Pink erasers have a very fine grit which serves to remove micro-scratches in the gold pads as well as baked on hardened greases and oils that may be present. White erasers will only remove (and smudge) softer oils and greases.
On heavier oxidation, such as if the contacts were exposed copper, this is what I'd have recommended. But those contacts should be gold plated. Combine the fact that gold is soft, and the plating is probably on the thinner side, an abrasive isn't the first thing I'd go for in order to keep it from being scratched off.
 

OK. But just a note about moisture. If you were to remove all of the moisture (100%) from the PWB it would turn to dust. Most things would actually. I'm not positive about gold but probably that too. :) There's some moisture in pretty much everything - or it dustifies. :)

Alcohol in the high ninety percents is strong enough to remove enough moisture (oils, water, whatever other liquids) to cause problems with a lot of different materials. Adding 30% water still allows the water alcohol solution to rapidly evaporate. At around 50/50 it's as you describe tho. ;)
 
OK. But just a note about moisture. If you were to remove all of the moisture (100%) from the PWB it would turn to dust. Most things would actually. I'm not positive about gold but probably that too. :) There's some moisture in pretty much everything - or it dustifies. :)
:confused: Not quite sure what you're getting at science wise, so a couple of sources (or at least some hints) would be helpful. :confused:

I can see it with a lot of organic materials (absolutely regarding biological life forms), but not all, and not as much with inorganic materials, such as metals.

Take bricks for example. The kiln process changes the chemical makeup so that the bonds will hold it together (not water). Water is actually driven out, though I wouldn't say at 100% (i.e. concrete too, where it can actually take decades before it's fully cured).

But water can, and does, cause damage after the brick or concrete has been formed, especially if there's a sub freezing to above freezing temperature cycling going on (ice forming in cracks, and causing damage, liquid water then further penetrates the brick, freezes ... wash, rinse, repeat until the brick fails). Thus water incursion into brick and concrete are generally regarded as bad due to the destructive nature, even though it was necessary as part of the formation process.

Alcohol in the high ninety percents is strong enough to remove enough moisture (oils, water, whatever other liquids) to cause problems with a lot of different materials. Adding 30% water still allows the water alcohol solution to rapidly evaporate. At around 50/50 it's as you describe tho. ;)
It still evaporates quickly, but not necessarily quickly enough.

Production rates are based on high speed (i.e. shorter cleaning and drying cycles = increased production and lower costs), and there are also instances where water cannot be allowed to come in contact with the components (some are hygroscopic, and it can damage them; even to the point where heating during the soldering process causes them to rupture).

There's also potential issues with corrosion from say electrolytic processes due to water present beneath a part, so if water is being used in the manufacturing, the manufacturing engineers have to be sure the components can in fact be immersed in water, and for how long if they are (ideally, these figures are included in the specifications on the datasheets; unfortunately, not always).

Since alcohol is also hygroscopic, the higher concentrations have a higher capacity for removing any liquid or moisture from the board and components. This capacity is one of the primary reasons it's used as well, besides being a good performing, low cost, environmentally friendly solvent. :D
 
I am OP... I just decided to soak my entire Mac Pro in a tub of Aqua regia for a week and it's fine now.
So you dissolved it into oblivion ... yeah, that's one way to *solve* the problem :eek: (aqua regia = nitric acid + hydrochloric acid in a 1:3 mix, and is extremely corrosive; it's even used to extract gold and refine it into ultra high purity @ 99.999%). :D
 
So you dissolved it into oblivion ... yeah, that's one way to *solve* the problem :eek: (aqua regia = nitric acid + hydrochloric acid in a 1:3 mix, and is extremely corrosive; it's even used to extract gold and refine it into ultra high purity @ 99.999%). :D

Now if he feeds it to his next monster machine we could make a movie and call it "Solvent Green". :)
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.