GFCI tripping anomolies

I thought I had a reasonable understancing of GFCI devices.
My first problems were when I was using GFCI for protection in a
hydroponic greenhouse. There was a lot of plant nutrient which was essentially salt water. I had immersible pumps in the solution that were on a timer. They tended to trip when the timer turned the pumps off. These used pumps with a separate ground wire.
I rationalized that the inductive kick somehow led to leakage to ground. By placing a capacitor between hot and neutral, was able to get rid of the unwanted trips.
More recently, I was using a Water Pik in the bathroom. It used a two wire connection with no separate grounding conductor. GFCI should still work well. There never was tripping when I turned the device off. When I remove the plug from the wall socket with device switched off was when I would get occasional trips. I rationalized that in the bathroom environment there may have been just enough surface moisture to connect to my hand.
Today was different. I turned off the Water Pik. My hands were relatively dry. I grabbed the plug as far away from the socket as I could while getting a firm grip on the plug structure so as not to put a strain on the wires. When I disconnected the plug, I got a trip AFTER the plug was disconnected. I guess that the time delay between the time I saw the plug's prongs and the click was about 1/5 of a second.
Does anyone here have reasonable scenarios for what happened?
--

Sam

Conservatives are against Darwinism but for natural selection.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Sounds like either your GFCIs are old or just plain cheap ones.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

I had several fans that tripped a GFCI when I changed the speed. They were several different brands, from about 1980-2000 when I moved from the place where I needed a fan. They were about US$20 to US$30, so they were not super high quality.
Since they were in the bathroom, I also attributed the problem first to water and, when that was eliminated, to some capacitor issue. I forgot to check if there actually was a capacitor in the thing.

Does the GFCI trip when you reset it and the Water Pik is not plugged in?
Does the self test button correctly trip the GFCI?
Does a GFCI Outlet Tester
(http://www.acehardware.com/product/index.jsp?productId &65480&CAWELAID9389922 or similar) show any problems?
(I don't have any suggestions for a reason for the tripping yet, but the above 3 things would help in narrowing the search. My bathroom had no detected issues, so I consider the problem open. However I never checked for the presence of a capacitor in the fans.)

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

The self test works.
I had no reason to suspect the GFCI wiring was incorrect. Nevertheless it was easy to check, so I did so. It was wired correctly.
I have run capacitors from hot to ground some time ago. IIRC GFCI tripped at 3.5?F but not 0.22?F at 120VAC. For minimizing trips when inductive loads are broken, the capacitor is placed between hot and neutral, not to protective ground. In principle, that capacitance can be very large without an imbalance getting detected. That, of course, is not the case for a neutral to ground connection.
--

Sam

Conservatives are against Darwinism but for natural selection.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Fri, 23 Dec 2011 16:00:59 -0800, Salmon Egg

As I said, I don't know if the fan that tripped things had a capacitor, but it only had two wires, hot and neutral; no ground wire.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

On my Mac computer, the "?" was entered as a Greek "mu". I remember the tripping value was consistent tripping current specification. What is possible is that enough moisture leaked in a submersible pump that the leakage was close to the tripping value. That conductivity is likely to be temperature dependent. One pump had enough leakage so that I could not use it on a GFCI protected circuit. It was small enough that the pump worked well on a grounded circuit without GFCI protection. Nevertheless, I did not want to use it in that environment with highly conductive solutions everywhere.
--

Sam

Conservatives are against Darwinism but for natural selection.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.