Pentair Pool Pump GFCI trips

Status
Not open for further replies.
FionaZuppa said:
i dont have a tripping issue.

but perhaps just those that snap-clamp on. $1.33 at allied
70065354.jpg
Click to expand...
Right on.

Sent from my XT1650 using Tapatalk
 
i do not think its the RF itself causing gfi to trip. X10 controllers inject 120kHz onto the mains and they do not cause gfi or afi's to trip.
definitely seems to be some "fault" current on the egc.
 
GoldDigger said:
The X10 standard carefully limits the energy and amplitude of its carrier transmission.

Sent from my XT1585 using Tapatalk
Click to expand...

so you think the RF being imposed onto pool pump frame is creating amplitude higher than the X10?

i still say the easiest thing to try is ferrites and a choke inline with the egc. you can also try wrapping the EGC wire in/around a ferrite ring 7-8 turns.
 
FionaZuppa said:
why not amp clamp the egc (take off gfi), see how their stuff is bad, then you can send a nasty-gram to pentair.
Click to expand...
GFCI is going to trip at 4-6 mA, can you accurately measure that level with your amp clamp?

Jraef said:
The issue is, these newer Pentair pool pumps have a built-in VFD to a 3 phase motor (the VFD accepts single phase, puts out 3 phase). VFDs produce a lot of high frequency (above 150kHz) common mode noise if not installed exactly perfectly with regard to grounding between the drive and motor. So in some GFCIs, that HF CM noise gets detected and acted upon. The Siemens GFCI breakers apparently discriminate against that HF CM noise and don't trip as easily. So you can either replace the GFCI with a Siemens version, or you can add a shielded isolation transformer between the GFCI and the pump.

I have HEARD (but not had the chance to try) that increasing the size of the ground connection inside of the pump between the drive and the motor can help, because the HF CM noise travels on the surface of conductors, not deep into them. By increasing the conductor size you are increasing the surface area. That's one of the reasons to use Shielded VFD cable, it's an alternate path for CM noise to travel. In this case because the VFD is close coupled to the pump motor, that doesn't apply here.
Click to expand...

I was going to suggest that a ground fault in the motor won't trip a GFCI on the supply side of a VFD. Then past couple threads bring up that the VFD and motor are all part of same assembly, so I guess it still makes sense to GFCI protect it for pool applications.
 
kwired said:
GFCI is going to trip at 4-6 mA, can you accurately measure that level with your amp clamp? ...
Click to expand...
Probably not, but a 50 mV meter with an appropriate shunt resistor can, some DMMs can, and you can always make a coil with ten or a hundred turns around the amp clamp jaw to increase its sensitivity.
 
kwired said:
GFCI is going to trip at 4-6 mA, can you accurately measure that level with your amp clamp?
Click to expand...
1) the General GTC CM100 can
2) my General CM660 is 0-600A with 10mA resolution, so if its 10mA or more, i can read it
3) so yes, some amp clamps can read gfci trip mA levels
4) need the right amp clamp for the job ;)
5) my Fluke DMM is 10A capable. so perhaps just inline a DMM with the EGC
6) why not just disconnect the EGC from pool pump to see if that wire is a carrier of mA (no gfi trip)?
 
FionaZuppa said:
1) the General GTC CM100 can
2) my General CM660 is 0-600A with 10mA resolution, so if its 10mA or more, i can read it The GFCI is supposed to trip at about half that level
3) so yes, some amp clamps can read gfci trip mA levels I figured there were some out there that can, most of us don't have them or seldom would even need them.
4) need the right amp clamp for the job ;)
5) my Fluke DMM is 10A capable. so perhaps just inline a DMM with the EGC Pool and associated components may very well allow leakage via other paths as has been mentioned clamping all the circuit conductors would be the thing to do, besides this is what the GFCI itself does.
6) why not just disconnect the EGC from pool pump to see if that wire is a carrier of mA (no gfi trip)? my reply to 5) applies here as well.
Click to expand...
 
FionaZuppa said:
1) the General GTC CM100 can
2) my General CM660 is 0-600A with 10mA resolution, so if its 10mA or more, i can read it
3) so yes, some amp clamps can read gfci trip mA levels
4) need the right amp clamp for the job ;)
5) my Fluke DMM is 10A capable. so perhaps just inline a DMM with the EGC
6) why not just disconnect the EGC from pool pump to see if that wire is a carrier of mA (no gfi trip)?
Click to expand...
The tripping is very intermmitent. I've been back to at least 50 jobs in the last 5 years to troubleshoot VS pumps tripping GFCI's and less than 5 have actually had the GFCI trip while is was on site. After a thorough look for any deficincies with the installtion, I usually replace the breaker with a brand new one, re-box the old and note the job and date. The breaker goes back into service when the client calls back to say the breaker tripped again. I tell them its a problem with the pump and to call the builder to deal with Pentair warranty.

This is the manufacturers problem and they need to own up to it.

Sent from my SM-G928V using Tapatalk
 
texie said:
This is a known issue with these pumps. I think there are even other threads about this. I seem to recall that Pentair claims they do not have these issues with Siemans GFCI. I have a novel idea, maybe someone needs to take Pentair to task and force/shame them into investigate and fix what the issue is here. Just taking the coward way out and telling people that they must use a certain brand would not be acceptable to me. What does one do if Siemans is not an option for the panel that is installed? OK, add a subpanel for the Siemans breaker. Who's paying for this?
Click to expand...

Any solid ideas on how to force or shame Pentair to the task of investigating this?

I'm doing what I can through my network of pool builders and sharing the issues I've had on other builders jobs. I have a few seriously considering making the switch and paying more for Jandy stuff.

I stay in touch with the east coast service manager and rep but there clueless and try to push the Pentair version of the Siemens breaker which is rated for only in their control system load centers.

Any suggestions are appreciated.

Sent from my SM-G928V using Tapatalk
 
bclumen said:
Any solid ideas on how to force or shame Pentair to the task of investigating this?

I'm doing what I can through my network of pool builders and sharing the issues I've had on other builders jobs. I have a few seriously considering making the switch and paying more for Jandy stuff.

I stay in touch with the east coast service manager and rep but there clueless and try to push the Pentair version of the Siemens breaker which is rated for only in their control system load centers.

Any suggestions are appreciated.

Sent from my SM-G928V using Tapatalk
Click to expand...

I would suggest reaching out to UL (as they are the listing NRTL for this pump) and NEMA (hopefully Pentair is a member). I'm not familiar with who the reps are for UL or NEMA in your area as I am in the west. This can be a trying task to get everyone's attention in cases like this, but I feel strongly that manufacturers should be held accountable for these things. We see way to many of these problems in today's installations, especially as the code moves toward far more situations where AFCI and/or GFCI are required. It has always amazed me how all these major manufacturers just continue ignore obvious issues and just pawn it off on the installer. I have been involved in many situations like this and as a rule I will go to great lengths to do my best to make them come up with a solution as a matter of policy.
Start here: http://www.ul.com/code-authorities/keeping-in-touch/contact-us/
https://www.nema.org/About/Pages/Staff-Directory.aspx
 
texie said:
I would suggest reaching out to UL (as they are the listing NRTL for this pump) and NEMA (hopefully Pentair is a member). I'm not familiar with who the reps are for UL or NEMA in your area as I am in the west. This can be a trying task to get everyone's attention in cases like this, but I feel strongly that manufacturers should be held accountable for these things. We see way to many of these problems in today's installations, especially as the code moves toward far more situations where AFCI and/or GFCI are required. It has always amazed me how all these major manufacturers just continue ignore obvious issues and just pawn it off on the installer. I have been involved in many situations like this and as a rule I will go to great lengths to do my best to make them come up with a solution as a matter of policy.
Start here: http://www.ul.com/code-authorities/keeping-in-touch/contact-us/
https://www.nema.org/About/Pages/Staff-Directory.aspx
Click to expand...
UL will probably just say it already meets listing requirements, which it probably is already listed. Such listing requirements also are more likely focused on safety than reliability.
 
kwired said:
?...
I was going to suggest that a ground fault in the motor won't trip a GFCI on the supply side of a VFD. Then past couple threads bring up that the VFD and motor are all part of same assembly, so I guess it still makes sense to GFCI protect it for pool applications.
Click to expand...
Most modern VFDs, theirs included, will themselves detect and trip on a GF in the motor but the VFD will not be listed as a Class A GFCI (4-6mA), which is what the Code requires, because the CM noise current will exceed that. The issue here is that because the FEED to the VFD requires that Class A GFCI, the CM noise that doesn't get immediately back into the VFD is getting picked up by that external GFCI in some cases. I don't think the pump mfrs consider it consistent enough to warrant further investment in their design, since from their perspective, there appears to be a solution; using specifically tested GFCIs that don't react to their units. I'm not condoning that position, I'm just postulating on why they may not be subject to "shaming" from us. WE are not their target market...

I know a guy in New Zealand who experimented extensively with ferrite cores on the ground connections to try to deal with this issue in dairies, where the CM noise affects cows connected to milking machines and decreases their output. It had some effect, but was not a cure. He wasn't doing this for the purpose of trying to prevent GFCI tripping though. It would be an interesting experiment.
 
Here is an image taken from this website that actually does a decent job of showing the phenomenon. This is a "before and after" shot showing the effects of adding a filter to the output of a VFD. But you can see in this that while the "before" image shows 1.72A (not mA) of ground current noise, even after the filter is installed it's still 6.92mA. Far far less, but still above that 6mA threshold, if your GFCI is not discriminating out high frequency current (this is not shown, but is likely in the 150MHz range).

Fig14.png
I'd say that if the GFCIs are not CONSTANTLY tripping, the mfr likely has already employed filters, but there is still some CM noise that gets through.
 
Status
Not open for further replies.
Top