DI water corrosion of electronic components? 1

[jlinstrom]

We use ultrasonic humidifiers with DI water to maintain proper humidity to suppress ESD in an electronic assembly area.
We have fried 4 power supplies in 5 months on our SMT assembly machines. We have a very troublesome buildup of contamination in all pick&place machines that we think is causing the failures.
This contamination is isolated to electrical systems and looks like hoarfrost, for lack of a better name. In particular, it shows up on some wires, fuses and circuit boards but not at all on others. The contamination build-up is on a supply in use for less than 6 months. Please me if you have any ideas what could be causing it. FYI we do not see this contamination on any other equipment in the facility and the area is climate controlled and ESD safe.
I have read that DI can attack copper; we have copper-leaded components on assemblies, but also solder (lead) coated components.
The most puzzling part of this is that we see a thin white powder coating on the insulation of power supply wires.
Can any of this be explained by humidified DI water?
Thanks-
john

twodogs
SUNSPOT Svcs.

 

[MikeHalloran]

Deionization removes the natural pH buffers from water, so its pH is easily driven away from neutral. As a result, DI water is one of the most corrosive substances imaginable.



Mike Halloran
Pembroke Pines, FL, USA

 

[EdStainless]

It could be moisture drive, but I would guess that it is related to dust being attracted to the power supply leads by the surface charge, and then this material causing corrosion in the humid environment.
There is all of the particulate that you are putting into the air.
DI is water is not high purity, just ion neutral. the stuff that is left is all put into the air by your humidifiers. Try using distilled water. I bet that you are finding this residue on other surfaces also it is just not causing problems, yet.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[bimr]

DI water is corrosive only if oxygen is also present.

When water evaporates, the salts in the water do not evaporate, the salts stay in the humidifier.

It doesn't matter what water you are using. If you use water with salts in it, the water is just going to crud up your humidifier, not crud up your building or products. The salts are not going to leave the humidifier.

DI water is also more or less the same as distilled water.

Distilled water (or DI water) is used in humidifiers so that the humidifiers don't crup up (scale up) and require periodic cleaning.

No, the DI water has nothing to do with your problem.

I would suspect that you have too much humidity in the conditioned air. If you keep the humidity below 40%, then you will not have corrosion.

Some users actually install dehumidifiers in facilities to lower the humidity and prevent corrosion.

http://www.bry-air.com/applications/corrosion_prevention/default.html

I would suspect that the white material is probably dust that settles out on damp surfaces.

You will have to balance out the humidity between preventing corrosion and preventing static.

You probably should contact a HVAC engineer that is familiar with the construction of electronic manufacturing buildings.

 

[bimr]

Here is another link for you:

http://www.bry-air.com/files/app_updates/ElectricalIndustry.pdf

 

[jlinstrom]

I failed to add - the water goes thru RO (reverse osmosis) before the DI. We never saw this failure in 4 yrs. I've been here until humidifiers got fixed: no new food in area rules, no new floor wax, spill events, paint jobs. In other words, no new dust episodes, just a proximate humidifier that had broken U/S generators, recently fixed.
DI water and white dust deposited ONLY on high current areas, not uniformly sprayed around chassis areas...
Correlations?

thanks-

twodogs
SUNSPOT Svcs.

 

[bimr]

The only chemical that is evaporated into the air by the humidifier is pure water. It doesn't matter what the water source is. The humidifier will separate the water and dissolved salts and just put the water into the air by evaporation. The salts remain in the dehumidifier. The salts do not evaporate, they scale out in the dehumidifer.

The reason that people use pure water (DI, RO, Distilled, etc.) in a dehumidifier is that the salts that remain in the dehumidifier will be less volume than if you use a water with a high salt content.

It does not matter how the water vapor gets into the air because the salts do not travel with it. It is pure water in the air.

That is why, when it rains, you get pure water. Have you ever heard of it raining seawater? Of course not.

You just mentioned that you recently "fixed" the dehumidifier. That should ring a bell for you. You are now putting too much moisture into the air, which is causing your problems.

Suggest you try turning off the dehumidifiers and see if the problem goes away.

You really should consult with a HVAC engineer. Have him measure the humidity and give you a recommendation as to what the humidity should be.

 

[EdStainless]

No, any and all matter in the water ends up in the air. If you run ultrasonic humidifiers with high TDS water you will get a heavy layer of white 'dust' that is actually the solids from the water.

High current = high local field strength. Anything in the air with a charge will be attracted.

That said, BIMR is right, the problem is probably too much humidity. The water itself isn't corrosive, but it is keeping surface deposits damp and that is leading to corrosion.
Either lower the humidity, or reduce the airflow through these cabinets so that they run hotter and have less risk for absorbing moisture.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[jlinstrom]

I'm puzzled (twice).
Does humidifier water eject anything else besides pure water into the air (it's ultrasonic)? I understand about the residual crust (we have none in the humidifiers).
How different would things be if we used 'softened' water?

We need to keep above 30%RH for ESD purposes and an infrequent oops is to leave heat off during a dark week-end, come in Monday AM, turn on the heat, and find HIGH (>70%RH) humidity. We are in Montana, humidity is low here and we need humidifiers to even reach 30%. Usually it is closer to 20%.

I'm thinking that high current as a static attractor si true here.
Thanks again.
john

twodogs
SUNSPOT Svcs.

 

[moltenmetal]

Softened water contains far more dissolved solids (TDS) than water treated by reverse osmosis followed by ion exchange deionization.

If your humidifier were purely evaporative, only water vapour would be entering the air. But if it is ultrasonic, it actually throws tiny droplets of water into the air where they evaporate, leaving any TDS as dust. The lower the TDS of the water, the less dust you'll have to worry about.

The high relative humidity is likely the problem. You need a better humidistat.

 

[bimr]

"Pure, particulate-free humidity. When using DI or RO water, adiabatic systems with high-grade stainless steel parts can produce humidification as pure as the supply water."

"Ultrasonic humidifiers have a submerged vibrating disk that creates a high-frequency oscillation, expelling small water droplets into air. Many of these systems use DI or RO water and are used in applications where particulate-free humidity is essential but where a large capacity is not required. The disks require replacement after approximately 10,000 operating hours, which can be a concern for applications that cannot tolerate off-line time or the expense of disk replacement. Water droplets are generally small and maximum capacity is approximately 80 lbs/hr per unit."

If you have a good quality dehumifier, it should not be putting dust into the air.

http://www.dristeem.com/driweb/rend...r_Adiabatic_Basics.pdf?action=vault&id=305909

 

[jlinstrom]

so, an ultrasonic throws up whatever is in the tank and after water evaporates, the other stuff is left as dust. If I use RO water, what might be in this dust? - calcium, magnesium? If I use water soften. what is replacement - sodium? Is this a worse dust on electronics? How about if I use distilled water - what is in it? All I want in the air is water!
And yes, we need a better humidistat to control both high & low RH.
thanks-

twodogs
SUNSPOT Svcs.

 

[moltenmetal]

If your RO and DI units are working properly, there shouldn't be much of anything in that water.

If you use distilled water, the only thing that'll be in there is what gets in there afterward, i.e. CO2 absorbed from the atmosphere etc.

If you use softened water instead of RO+DI water, the water will have sodium replacing the calcium/magnesium/iron that was in there to begin with. I don't know if this equates to "worse for electronics", but sodium carbonate is more soluble than calcium carbonate and hence more conductive in conditions which are not bone dry.

It sounds like you fixed a broken humidifier and now you're throwing more water into the air. The humidity is going up, but so is the dust load as a result. Best to control the humidity so you dump as little dust as possible into the air- or switch to an evaporative humidifier.

 

[Duwe6]

Little known fact outside of the Power Plant/Steam Boiler industry -- uninhibited RO and/or DI water is incredibly corrosive to any metals. Just like MikeHalloran said. And the thinner the film of water, the higher rate of corrosion. Oxygen plus water gives very severe corrosion.

The reason you normally don't see severe corrosion from water is that it leaves minerals behind. In any house over a couple of years old, the is a layer of "lime" inside the copper or galvanized steel pipes. [limi is in parentheses because the deposit is semi-soluable and insoluable minerals leached from the ground. Some is lime, bou there are a LOT of other minerals in pipe scale] The water will never touch the metal inside the pipes. And the "lime" is a self-healing coating. When it gets scraped off, it replates onto the ID of the pipe.

DI/RO water has no minerals. It is typically tested for conductivity, to insure that there are no free ions of any kind. Without an added Inhibiter added to boiler feed water, the steel pipes would rust out in about 5 years. And that is inside a system with very, very little oxygen. Out in the open, you can get years of 'normal' corrosion in days by applying a light mist of DI/RO water. Fact of life.

Suggest that you go with either sealed, watertight cabinets, or use motor-winding varnish to 'paint' all the bare portions of your terminals and terminations. Varnish the relay enclosure boxes shut.

 

[moltenmetal]

Duwe6: by your argument, rain or dew would be extremely corrosive. It contains very little dissolved mineral content.

Unless the OP's room's HVAC is totally out of control, I doubt they're seeing condensation of liquid water on the surfaces where the OP is seeing the "hoar frost". If there is any corrosion happening, it is happening as a result of exposure to excessively humidified air, not exposure to a film of (deionized) condensate. I seriously doubt that the "hoar frost" that the OP is seeing is a corrosion product.

Your notion that deionized water is "extremely corrosive and hungry for ions" which you have repeated here is based on a misunderstanding of the underlying chemistry in my opinion.

It is true that deionized water will rapidly become contaminated with ions when passed through metallic piping systems, rendering it unfit for some uses- hence people use PFA or PVDF piping to keep the water pure in those applications. It is also true that a continuous supply of deionized water will, to the extent that it contains oxygen, cause corrosion of some metals, particularly iron and copper. It is also true that the addition of certain minerals and other additives can reduce the corrosivity of the water or aid in passivating the surface, rendering iron and copper adequately resistant despite the presence of oxygen.

It is quite another thing to conclude that there is a mechanism to support ongoing corrosion in a system merely because the first charge of water was deionized and hence (initially) "hungry for ions".

In a sealed system without oxygen, deionized water is not corrosive. In a system charged with oxygenated deionized water and then sealed, the oxygen will be rapidly depleted by corrosion, which will then STOP. In systems which are open, having a continuous feed such as your boiler feedwater example, the corrosion rate is related to the rate at which oxygen (or another electron acceptor in the case that the corrosion is occurring as a result of microorganisms) is fed and the rate at which the corrosion products are removed.

 

[Compositepro]

Ultrasonic humidifiers produce a mist of microscopic water droplets which then evaporate in the air. If there are any solids in the water it will become a dust particle in the air. These particles are almost always white and will be attracted to charged surfaces, which seems to be what you are observing.

Take a sample of water from your humidifier, let it evaporate to dryness in a clear glass, and look for any residue.

 

[moltenmetal]

D'oh! Sorry Duwe6- now I think I get what your point might have been. If these dehumidifiers are automatically fed, the lines carrying the RO/DI water might be corroding, producing the particles that the ultrasonic humidifiers are throwing in the air. That's a fairly likely scenario. The OP should replace any metallic components with plastic in those lines and see what happens.

 

[jlinstrom]

moltenmetal - good thought, but all pipes, fittings, etc. are plastic.
Also, what do you think my 'hoarfrost' is?
Is it corrosion of the solder on the pcbs (I don't know if they are RHOS)? Why would it preferrentially be on the low voltage DC side? The power supplies affected are from a few different mfgrs. so I doubt it is a design or mfgr. failure. I see nothing on the surfaces around the power supplies that looks like deposited solids, just on the pcb itself and -puzzle- the high current DC wires from these supplies up to the 'parts' they energize.
Thanks-

twodogs
SUNSPOT Svcs.

 

[bimr]

First off, if the humidifier is from a quality manufacturer and DI water is being used in the humidifier, there should be no dust in the water vapor or evaporated water. See what the humidifier manufacturers state:

"Pure, particulate-free humidity. When using DI or RO water, adiabatic systems with high-grade stainless steel parts can produce humidification as pure as the supply water."

"Ultrasonic humidifiers have a submerged vibrating disk that creates a high-frequency oscillation, expelling small water droplets into air. Many of these systems use DI or RO water and are used in applications where particulate-free humidity is essential but where a large capacity is not required. The disks require replacement after approximately 10,000 operating hours, which can be a concern for applications that cannot tolerate off-line time or the expense of disk replacement. Water droplets are generally small and maximum capacity is approximately 80 lbs/hr per unit."

If you have a good quality humifier, it should not be putting dust into the air.

http://www.dristeem.com/driweb/rend..._Adiabatic_Basics.pdf?action=vault&;id=305909

The second point is that one would expect that the white deposits are probably corrosion products on the metal because of high humidity.

I believe that the poster is saying that the white deposits are on metal not plastic components. That would indicated that the white deposits are from corrosion.

Try a test. Bag one component with a dessicant and leave another identical component out in the building air. After a few weeks, compare the components to see if the building air is causing the problem.

If the white deposits were on everything, the white material could be dust from the humidifier. If the white material is not over everything, then the white material is not dust.

A third point. If you are in fact making electronic components in a somewhat clean environment, why would you shut off the HVAC system that is conditioning the air on the weekends? If the building temperature drops over the weekend, it will cause the humidity to increase, resulting in condensation.

Consult with a HVAC engineering specialist. Have him measure the humidity and give you a recommendation as to how to operate your HVAC system for best results.

 

[jlinstrom]

we monitor the temp and humidity , samples every 10 minutes or so. During the P/S failures, temps were 63-77 and RH was 58-33.
HVAC is 24/7. I checked and humidifier water is R.O. not D.I.

White dust is NOT on everything, just the DC side BUT this includes the insulation on the DC wires. Areas inside/around the pick&place machines have grey linty dust but none of this white crusty powder.
I'm working on a time-line, maybe that will give a clue.

john

twodogs
SUNSPOT Svcs.