A sample text widget

Etiam pulvinar consectetur dolor sed malesuada. Ut convallis euismod dolor nec pretium. Nunc ut tristique massa.

Nam sodales mi vitae dolor ullamcorper et vulputate enim accumsan. Morbi orci magna, tincidunt vitae molestie nec, molestie at mi. Nulla nulla lorem, suscipit in posuere in, interdum non magna.

Encapsulating the problem… Waterproofing the solder joints.

Coatings I tried to use to insulate the solder joints.

Coatings I tried to use to insulate the solder joints.

Almost everything in design is a compromise. Even something as simple as insulating wire connections that may be exposed to water.   It always comes back to: “What are you trying to accomplish”, and even then often there are compromises.  In the case of the thermistor board I may want to use the thermistors to determine water temperature as well as air temperature.  However, there is a competing issue. The thermistor does not have a very fast response to changing temperatures.   In the post “Our First Look at Thermistor Details“, the specification for the thermistor thermal time constant is 20 seconds.  That means it will take approximately 100 seconds, over a minute and a half, for the thermistor to settle correctly to read a large temperature changes.

I cannot really imagine that response to be too long for anything I hope to do with the thermistors, but I really don’t want to increase the time if I can avoid it.   A simple answer to the problem would be to dip the thermistor and wire connections in something like the plastic coating for tool handles and be done with it.  Another option would be to completely encase the whole thing in silicon rubber.   However, doing all of those would put extra thermal insulation around the thermistor and increase the thermal time constant.  I would also want to check each of those to determine that it does not affect the thermistor itself.

What I chose to do was to attempt to just coat the solder joints and a little way up the wire insulation and where the device and wire meets to ensure a waterproof connection.   I wanted to try some kind of epoxy paint, but I could not find any locally.  I did some looking on the net and found out that fingernail polish is basically automobile paint in little bottles so that was one of the choices.  (For any male out there who gets lectures from his wife about his choices of food and beverage being unhealthy and his wife uses nail polish, here is a little ammo for you.  Nail Polish Ingredients.  Just don’t tell her you got it from me.)  I chose a clear coat because most of the colors use some metallic oxides and that might be conductive.

My second choice was a spray called “Spray Liquid Tape – Electrical Spray-on Insulation Coating”.  It sounded like it was exactly what I wanted.   I should have thought some more.   The final coating tested was some glue I have use for similar things in the past.  This is “Welder contact adhesive – permanently seals, bonds & repairs”.  It sounded like it would work and from past experience, I knew it was a lot more viscous than the nail polish.

To test this I first made up twenty resistor pig-tail assemblies.   I used 1MΩ resistors for two reasons.  First, I have lots of 1M resistors laying around and I don’t have any planned use for them.  Second, and really more importantly what I am worried about is a parallel resistance path and with the reference resistance at 1M I should easily be able to determine if I have a parallel path of even several MΩ.

I divided the resistor-pig-tail assemblies into 4 groups with five members each.  Each of the pigtails was tagged so I could keep track of each individual unit.   There are so many because I thought it might be necessary to do some statistical analysis.  That turned out to be not necessary.  The resistance of each of the 20 was measured and place in a spreadsheet.

I then treated each group with the insulation treatment and left one group untreated.   One of the questions I was concerned about was if my Ohmmeter was consistent day-to-day.  The untreated “control group” enabled me to determine it was.  I then took the measurements again and subtracted the second reading from the first on each resistor and then averaged the change for each group.  The summary data is:

Treatment Difference (M Ohms)
Nail Polish -0.0002
Sealing Glue 0
Insulation Spray -0.0006
Control Group -0.0014

Since my meter only displays 3 1/2 digits this means essentially no change.  One of the control group resistors did change some, but what I think happened is I misread the last digit.  It should have been a 7 and I must have read a 1.  (Your’s truly is the randomness in the process.)    Next I dunked the working end of all of the assemblies in ice water to see how they would hold up.   I took measurements while these were under the water and subtracted from the initial data as described above.  The summary is:

Treatment Difference (M Ohms)
Nail Polish 0.0096
Sealing Glue -0.0102
Insulation Spray 0.1436
Control Group N.A.
The Spray Insulation on the Resistor.

The Spray Insulation on the Resistor.

Obviously, there were big problems with the insulating spray. I looked closely at why this failed. Because I was worried about build up of the insulation on the resistor I did not spray enough on the resistor.   This would not be my method of choice in any case, because the spray application method makes it impossible to read the component values.

Now I was ready for the final test.   I dunked the groups of resistors in to hot salt water.  In this test I am doing several things.  First, of course, I am checking the effects of a higher temperature.  The salt was added to make the water more conductive to make it an even worse case.  The final thing I wanted to find out is how hard it is to maintain a constant temperature because that will be necessary for the calibration of the thermistor circuit board.

Things did not go so good.   First I managed to also dunk the control group so there will be some data for them also.  Second, I managed to dunk my electronic thermometer.  My guess is it is destroyed, but it was not an expensive one.    The final bad news is all of them failed as seen in this summary data.

Treatment Difference (M Ohms)
Nail Polish 0.6342
Sealing Glue 0.4786
Insulation Spray 0.927
Control Group 0.9492

The insulation spray was almost as bad as the bare wires of the control group.  Both the Nail polish and Sealing Glue failed badly also.  Time to go back to the drawing board!   I looked a little more closely at the data and one of the sealing glue resistors changed very little; it changed from 0.983 to 0.968 MΩ or 0.015 MΩ change.   The question then became: “What did I do different with that particular resistor compared to the rest of that group?”  Being an old man, it was necessary to get out a magnifying glass and a bright light.  I also used my Ohmmeter on the continuity scale to see if I could find the problem.

The Sealing Glue Coating of the connections.

The Sealing Glue Coating of the connections.

What I found was the sealing glue contracts as it cures.   As it contracts it pulls from the end where there is nothing to adhere to.  In other words it pulls back and leaves the very end exposed.

This may not be a problem with the actual thermistors because those  have a radial lead arrangement and not axial leads like the resistor.  But more importantly, now that I know of the trap I will be much more careful the next time I use it.

Altogether I would call this test inconclusive.  However, because one passed, I then took all of the resistors and rinsed off the salt water.   I cut off the spray on insulation resistors and installed 5 new resistors on those pig tails.   I then coated all 20 resistors including the control group with more sealing glue.  One group will have a nail polish base coat covered with a sealing glue outer coat.  I visually inspected each resistor and six of them are getting another application.   The next test will only be the room temperature data and the hot salt salt water test.

Hopefully I whoop up on Murphy on this next one.   Murphy is famous for his infamous law:  “Anything that can go wrong, will go wrong”.

Gary

If you found this blog post enjoyable and educational, please consider subscribing to this blog.  You can subscribe to e-mail updates by using the e-mail subscription form at the bottom of each page and on the home page.   You can subscribe via rss updates or join the facebook or twitter updates by using the icons on the home page.   If you have any questions or concerns or preferences on where I guide this blog, please e-mail me or join the facebook page.


 

Print Friendly

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>