Most interesting troubleshooting call I think I've ever had

[sfav8r]

OK, so a friend of mine calls and says he can't get his pond pump working even after installing a brand new pump. He tells me the steps he's taken and I can't really make sense of ot so I tell him I'll head over and take a look at it. I won't go into all the steps I took, but this is one of the strangest troubleshooting calls I've ever had. Here is the situation:

There is a GFCI protected outlet about 6' from the pond. It is fed from a dedicated circuit. When you plug the pump into the GFCI, it will run for about 2 seconds then shut-off. If you unplug it and plug it right back in, it will run for 2 seconds again. If you take the pump, put it in a bucket of water, plug it into any other outlet it will run all day. If you take anything BUT the pump and plug it into the GFI by the pond, it will work perfectly. (no really, I saw it with my own eyes). I monitored the voltage at the GFCI by the pond it it is OK at all times and the GFCI never trips.

Now here is what I found out and what I find to be really strange. Because of the weirdness described above, I decided to try putting the pump in a bucket and plugging it into the GFCI by the pond...violla, it works perfect...no shutting off. If you take the pump out of the bucket and put it in the pond, it immediately stops. Thinking there might be something odd with the water in the pond, I put the pump back in the bucket (it came back on). I then placed the bucket into the pond and the pump in the bucket immediately shut off even though it was still in the fresh water in the bucket. So, I think I am on track when I conclude that the issue only surfaces when the pump is grounded via the water in the pond. To try and confirm this, I put the bucket and pump on the walkway near the pump. It worked fine, then I ran a wire from the water in the bucket and touched it to the wet ground near the pond and again the pump shut off. As an experiment, I ran a ground wire from the electrical box to the water in the pump and the pump continues to run fine. So, my theory is that there is some stray voltage or something in the ground near the pond that makes the auto-shut-off in the pump decide it needs to turn off. I came up with the idea of isolating the circuit, so I used an isolation transformer between the GFCI and the pump and now it will work fine, even when submerged in the pond water. OK, I know all that is strange enough, but what I found out by accident is even more strange. I may make a youtube video of this because if I hadn't seen it myself, I'm not sure I would believe it.

When the pump is plugged into the GFCI outlet and submerged into the pond water, as I mentioned in the beginning it runs 2 seconds and then turns off. For reasons I cannot explain, if I take a 30' extension cord and plug it into the other outlet on the same GFCI. even though there is nothing plugged into that extension cord, the pump will start running. I tried a different extension cord and it will not work. Only one of the three extension cords I tried worked. The difference in the cords is that the one that worked is the type that has the wires twisted around each other and the ones that do not change anything are the type that are just individual wires that parallel each other. The only thing I can think of that even comes clsoe to making sense is that the cord with the spiraled wires when plugged into the other outlet (parallel to the pump) is somehow inducing a "coil" effect or something. I am really intrigued by this problem. as I mentioned, I have it "fixed" but I'd really like to understand what is going on here.

Any ideas?

 

[Dennis Alwon]

It is odd indeed. I do have a question. If you install an isolating transformer between the gfci and the pump will the gfci still function?

 

[Dennis Alwon]

I would be curious to know if changing the gfci was done or if anything happened and the other thought was that there is a break in the underground wire which causes the pump to labor. Now why does it work in one place and not the other is a bit of a mystery. There may be stray voltage but I don't understand why that is causing the pump to shut down.

 

[hillbilly1]

Sounds to me like a bad pump with the equipment ground not connected at the gfi or internally in the pump. When you have the pump in the bucket (I'm assuming a plastic bucket) it is effectively insulated from ground. That is why when you put the wire in the bucket to the concrete, it trips because you are removing the isolation from ground.

 

[sfav8r]

I would be curious to know if changing the gfci was done or if anything happened and the other thought was that there is a break in the underground wire which causes the pump to labor. Now why does it work in one place and not the other is a bit of a mystery. There may be stray voltage but I don't understand why that is causing the pump to shut down.

Hi Dennis, thanks for the reply. First, I mispoke in the first post. The transformer is just BEFORE the GFCI. I did change the GFCI and also tried a non-GFCI outlet just for grins. I cannot figure out why a stray voltage would make the pump stop either. I called the manufacturer to se how the auto shut-off feature works but apparently they arent as fascinated by the problem as I am ;-)

Any theory on why plugging an extension cord into the other half of the oulet could make any difference at all? FYI, I measered the voltage from the equipment ground to a temporary grond rod I put in near the pump and could not measure any voltage.

 

[hillbilly1]

Hi Dennis, thanks for the reply. First, I mispoke in the first post. The transformer is just BEFORE the GFCI. I did change the GFCI and also tried a non-GFCI outlet just for grins. I cannot figure out why a stray voltage would make the pump stop either. I called the manufacturer to se how the auto shut-off feature works but apparently they arent as fascinated by the problem as I am ;-)

Any theory on why plugging an extension cord into the other half of the oulet could make any difference at all? FYI, I measered the voltage from the equipment ground to a temporary grond rod I put in near the pump and could not measure any voltage.

Just re-read your post, the pump itself has an autoshut off? So the gfi is not tripping? Never seen a pump with an auto shutoff, what brand and model is it?

 

[sfav8r]

Sounds to me like a bad pump with the equipment ground not connected at the gfi or internally in the pump. When you have the pump in the bucket (I'm assuming a plastic bucket) it is effectively insulated from ground. That is why when you put the wire in the bucket to the concrete, it trips because you are removing the isolation from ground.

Have tried 3 different pumps with the same result. I think you have the right idea with the grounding, but it must be something other than a "bad" pump. I wish I knew how the auto shut-off worked because it may offer a clue. Any thoughts on why plugging the extension cord into the other half of the GFI would make any difference? That one really has me puzzled.

 

[charlie b]

I will echo hillbilly1's question. What do you mean by saying that the pump runs for 2 seconds, and then shuts off? Does it stop because the GFCI has tripped? Does it stop because there is some control feature you haven't told us about yet? I gather that there is no separate motor controller, and that the only way to start or stop the pump is by plugging it in and unplugging it. But is there something built into the housing of the pump that has an overload trip or other on/off control mechanism?

 

[sfav8r]

I will echo hillbilly1's question. What do you mean by saying that the pump runs for 2 seconds, and then shuts off? Does it stop because the GFCI has tripped? Does it stop because there is some control feature you haven't told us about yet? I gather that there is no separate motor controller, and that the only way to start or stop the pump is by plugging it in and unplugging it. But is there something built into the housing of the pump that has an overload trip or other on/off control mechanism?

As mentioned in the original post "I monitored the voltage at the GFCI by the pond it it is OK at all times and the GFCI never trips." The pump has an auto shut off build in to protect it if the pond goes dry. Unfortunately, I don't know how the auto shut-off works. I think that might offer a clue. It certainly appears that the issue is when the water being pumped is exposed to ground, the pump stops. However, teh GFI never trips so it doesn't apper to be a leakage problem. As I have mentioned, the really odd part to me is the part about plugging the extension cord into the other receptacle. How that changes things really has me stumped.

 

[hillbilly1]

It may be some sort of electronic level sensor, if it does, it probably has a pair of electrodes that when water is across them, it triggers the run signal. Do you have a model number and manufacture?

 

[hurk27]

the only thing that would make sense to me is when you place the pump in the pond the inlet gets blocked and the sensor shuts the pump off, nothing else make any sense?:?

 

[Dennis Alwon]

the only thing that would make sense to me is when you place the pump in the pond the inlet gets blocked and the sensor shuts the pump off, nothing else make any sense?:?

That's what I thought but I believe he put the pump in the bucket that hd fresh water and then put the bucket in the pond and it went off immediately- 2 seconds

 

[hillbilly1]

It may have a control similar to this:
http://www.google.com/patents/US7243540?printsec=drawing&dq=low+water+cut+off&ei=K88NUNycE4re0QGrmtH3Aw#v=onepage&q=low%20water%20cut%20off&f=false

 

[hurk27]

That's what I thought but I believe he put the pump in the bucket that hd fresh water and then put the bucket in the pond and it went off immediately- 2 seconds

ya it doesn't make much sense? unless it has built in GFCI protection also some how? but then the plugging in the extension cord makes no sense either? or I would say to look for an air balance hose in the cord at the plug as some built in float switch's will use this and maybe a change of elevation may be playing a part?

 

[ptonsparky]

The company my brother works for makes a system that recycles irrigation runoff at the end of fields. They discovered that rainwater runoff would not always be sensed by the electronic level control they were using. The rainwater did not contain enough "contaminants" to allow the switch to function. They would work in water buckets with clean water or typical irrigation runoff, just not rainwater.

 

[hillbilly1]

The company my brother works for makes a system that recycles irrigation runoff at the end of fields. They discovered that rainwater runoff would not always be sensed by the electronic level control they were using. The rainwater did not contain enough "contaminants" to allow the switch to function. They would work in water buckets with clean water or typical irrigation runoff, just not rainwater.

The link that I attached shown an adjustment for water quality, but I doubt that pump has one.

 

[PetrosA]

I would stick a coil of bare copper in the pond and take readings to the hot, neutral and ground of the GFI receptacle. I suspect there is some stray voltage (possibly 240V from the other leg leaking from another UF somewhere) returning from the pond to the GFI through the pump that either the GFI doesn't know what to do with, or the electronics in the pump don't know what to do with but which causes something to shut down or disengage.

 

[elohr46]

The company my brother works for makes a system that recycles irrigation runoff at the end of fields. They discovered that rainwater runoff would not always be sensed by the electronic level control they were using. The rainwater did not contain enough "contaminants" to allow the switch to function. They would work in water buckets with clean water or typical irrigation runoff, just not rainwater.

This is most likely your answer, I ran across this same exact situation last year on a pool cover pump. Sometimes the sensor works and sometimes it doesnt depnding on the water quality.

 

[steve66]

This is most likely your answer, I ran across this same exact situation last year on a pool cover pump. Sometimes the sensor works and sometimes it doesnt depnding on the water quality.

I agree, although this comment from the OP makes me wonder:

I then placed the bucket into the pond and the pump in the bucket immediately shut off even though it was still in the fresh water in the bucket.

The bucket of fresh water would probably have chlorine in it. The chlorine ions would make the water a very good conductor for a liquid level sensing circuit. Rainwater - not so much.

I didn't see where the OP said what he did to fix it:

I mentioned, I have it "fixed" but I'd really like to understand what is going on here.

 

[kwired]

When you used an isolating transformer did you ground the derived system?

I think you should measure for voltage from the pond to the grounded service conductor both loaded and unloaded. Could be voltage coming from someplace else messing with the operation of this apparent water sensing circuit. If you did not ground the separately derived system you may have treated a symptom but did not make the real problem go away.