Pentair Intelliflo Pump causing Insteon Interference throughout house

[pgershon]

I have a rather large installation that has evolved from X10 to Insteon over then last 20 years or so.  About 75 Insteon modules now. My swimming pool equipment includes an Inteliflo pump installed in 2012. Over the last year or two it has increasingly caused trouble with my Insteon devices in various locations. I know the issue is the pool pump because Insteon works perfectly everywhere in the house when the pump is off. My question is how do I filter out the "noise" or other interference?

The pool equipment has its own panel (its a Jandy One Touch system).  Pool panel goes to a breaker in the pool house (I have some other lights and a fan that also go to that breaker). The pool breaker then feeds a main breaker in my basement, about 75 feet away. When the pool pump is on, three of my Insteon switches in the basement do not properly communicate (lights dont respond to off/on or sense).  Also, some outdoor lights (inline relay with sense) dont properly go on/off without sending signal multiple times.

I tried installing two XPF 20 Amp noise filters on the two legs of my pump load but it did not help.  I can add that the Intelliflo and another pump for my spa (spa pump not an issue) and a Heyward Aquarite Salt Chhlorinator all share the same 20 amp double pole breaker. The XPF was installed only onto the wires from the breaker to the Intelliflo directly.

Ideas?  I have found one other person online with a similar problem and no solution.  The pump functions perfectly for the pool, but the only solution I can think of for my Insteon would be to buy a new pump - not my preference.

Thanks in advance

[Goose66]

If the noise filters didn't work, I don't know what will. I use 20A X10 noise filters in my installation in several places, and they work fine for Insteon's frequency (131.65MHz). The noise could be bleeding over into unprotected circuits that run close to the pump. Maybe something larger installed back where the pool system subpanel comes off the main panel? My Jandy FloPro variable speed pump may interfere on wind up, but doesn't seem to cause problems normally throughout the day. Due to the nature of a variable speed pump that it runs so many more hours than a regular pump, I can see why interference could really be a problem.

[pgershon]

Might it help to run all three devices that plug into the 30 AMP breaker through the X-10 filters?  This would be the two pumps and the chlorinator.

[ELA]

On 9/11/2020 at 2:24 AM, pgershon said:

I know the issue is the pool pump because Insteon works perfectly everywhere in the house when the pump is off. My question is how do I filter out the "noise" or other interference?

The pool equipment has its own panel (its a Jandy One Touch system).  Pool panel goes to a breaker in the pool house (I have some other lights and a fan that also go to that breaker). The pool breaker then feeds a main breaker in my basement, about 75 feet away.

You say you know it is the pump.  You filtered it and the problem persists.

You say there are other loads in the pool house.  Any chance that those other loads are often on when the pump is on?

It is tough to find Insteon type filters larger than 20A so you may need to narrow it down to who actually is to blame.

Monitoring level 3 communications hops used can help you do that. 

If it is the pump then it is emitting a frequency that the Insteon filter is not designed to block.  A large EMI filter or Isolation transformer could help in that case.

[pgershon]

1 hour ago, ELA said:

You say there are other loads in the pool house.  Any chance that those other loads are often on when the pump is on?

 

I am able to test with pump on and all other loads off (interference persists) and with other loads on and pump off (no interference).  So I am pretty convinced.

I am thinking that the Leviton 6254 might help.  It could block the pool house breaker from the rest of the house - which is all I need to do.  I am not trying to communicate with pool equipment via Insteon / powerline - I do that through serial communications with the pool equipment (RS485).

[blueman2]

Pentair Intelliflo pumps are well known to create a lot of noise in the electrical system.  They are so bad,that you cannot even use a typical GFCI circuit breaker with these pumps.  They put out such noise that they will trip a typical GFCI.  Only 1 breaker that is resold by Pentair will work.  That breaker has an internal noise filter, which is basically an iron ferrite bead built in as a low pass filter.  So one idea is to just buy one of those iron donuts and run the pump's feed wires through them to see if that filters out the noise.  

[pgershon]

Thanks.  I did some research and ordered the Siemans breaker, which seems to be the same as the Pentair without the Pentair rebranding and for a lesser price.  My current breaker is a 20 AMP 2-pole breaker shared by the Intelliflo and my spa pump.  Would I be better off with a 30-amp breaker or staying with the 20 AMP?

[ELA]

Hello pgershon,

   I read an electricians forum on these pumps being notorious for noise just as blueman2 pointed out.

Lots of problems with GFCIs tripping with the alternate (siemens) breaker being  the solution.   From what I read it sounds like any filter inside the breaker would be designed to protect the breaker and not necessarily upstream Insteon loads.   Designed to prevent GFCI trips.

    If you try the breaker and it does not help purchase a quality high current EMI filter or isolation transformer.    Apparently these pumps are driven by a VFD ( variable frequency drive).  These produce high frequency noise.  The trick is determining what frequencies are produced vs. those which Insteon is susceptible to.

     I had one instance where a filterlinc failed to filter out an offending noise source.  That was noise at about 40Khz -  much less than the Insteon freq.  ... so the filterlinc did not isolate it well.

I instead added an isolation transformer that fixed the issue.  That load was just a few amps and a small transformer sufficed.

Problem is that Isolation transformers at higher currents are expensive and bulky.    EMI filters are  much smaller but can still be spendy.  The EMi filter should be installed as close to the VFD as possible.

 

 

[pgershon]

I tried the breaker today (20 AMP Siemans).  The result was interesting.  I connected just the Intelliflo to the breaker, removing much of the other equipment that was connected to the old breaker.  To my surprise, there was no interference as my pool house lights could be switched on and off by Insteon with the Intelliflo running. I then connected the other equipment to the old 30 AMP breaker and the interference was back.  I took the other equipment back offline but the interference would not go away.

It seemed like the Siemans breaker helped/filtered for a short period of time, but the effect was fleeting and it went away. I have the Intelliflo connected to the new breaker, but zero improvement in the equipment interference.

How do I get an idea of what frequency is being produced?

[Goose66]

13 hours ago, pgershon said:

How do I get an idea of what frequency is being produced?

Oscilloscope (or PC-based software equivalent). Has to be able to handle 240V (preferably 400-600V). You'll want to run traces in between neutral and each phase as well as across the two phases at the panel where the breaker for the pool equipment sub-panel is. Each trace should be run while you turn on and off various equipment and lights from the subpanel. Keep tracking of the timing of the on/off of each piece of equivalent and then let the oscilloscope do a frequency component analysis (Fourier transform) for the period of operation of each device on each trace. 

A lot of work, and I don't know that the answers you will get will point to a solution. You know that you have to filter out interference around the 131.65MHz band. The question is where do you put the filter, and what type of filter must you buy to put it in that spot. Again, my suggestion is a 60-100A bi-pole filter at the main panel that feeds the pool equipment sub-panel. But that will also block your Inseton controlled lights in the pool house, so you will probably want a separate 20A feed to the pool house for lights and such that will work with Insteon.

[ELA]

6 hours ago, Goose66 said:

You know that you have to filter out interference around the 131.65MHz band. The question is where do you put the filter, and what type of filter must you buy to put it in that spot. Again, my suggestion is a 60-100A bi-pole filter at the main panel that feeds the pool equipment sub-panel. But that will also block your Inseton controlled lights in the pool house, so you will probably want a separate 20A feed to the pool house for lights and such that will work with Insteon.

How do you know he has to filter out interference around 131.65Mhz?   A filterlinc filters out interference around 131.65Khz.

      As I pointed out ... I experienced that a  40Khz noise source  can upset the communications at 131.65Khz.    The filterlinc failed to isolate that Freq....    Isolation transformer did.

Without an O'scope or Spec Analyzer it is difficult to know the offending Freq.   That is why the use of a quality EMI filter ( specs with attenuation extending to lower frequencies) or Isolation transformer is suggested as a broader band approach.

To most effectively attenuate a noise source you always want to place the filter as close to the offending source as possible. 

 

 

 

[Goose66]

@ELA Just trying to be helpful - not looking to argue with you. 131.65 MHz is the band for PLC for Insteon.

@pgershon  You said something about controlling the lights in your pool house through Insteon. Is that the pump house, and is that Insteon controller dual-band? Perhaps it's actually RF interference from the pump that's causing problems in the dual-band switch or module.

[pgershon]

5 hours ago, Goose66 said:

@ELA

@pgershon  You said something about controlling the lights in your pool house through Insteon. Is that the pump house, and is that Insteon controller dual-band? Perhaps it's actually RF interference from the pump that's causing problems in the dual-band switch or module.

I wish that was the case.  The pool house breakers has a bunch of lights and it feeds 70 AMPs to the JANDY breaker box (with the pumps).  The pump must be causing wireline interference because the devices in my basement are affected too - and these are single band Insteon (pre-dual band).  It's 100% wireline noise.  The basement units worked fine when I had XPNR units in the circuit - but the pool house lights were blocked.  The XPNR did help with issues away from the pool house, but it also totally blocked everything at the pool house. The XPF filters did nothing/

[ELA]

Hello ,

pgershon

Interesting to hear that XPNR had a somewhat positive affect.

Quality filters are expensive.

I would hope one could get an EMI filter recommendation from the pump manufacturer/supplier.   If not I would recommend this one.

For installation as  close to the pump/variable drive as possible.  I did not compare to the cost of a shielded/Isolation transformer.

https://www.schaffner.com/product-storage/datasheets/fn-352-z/

Good luck

[pgershon]

That is a $300+ price-point.  It might be fine if I had confidence it would work, but its hard to know and I am not sure I want to just gamble with that kind of money.  Unfortunately the pump manufacturer has been less than helpful, suggesting I ask a local electrician to look at the equipment,

[larryllix]

Some confusion here. 132MHz would not make it 20 feet down house power wiring.

Insteon uses 132KHz for powerline, and 915MHz for rf comms.  Powerline comms at 132KHz would be the frequency to be concerned with.

https://www.insteon.com/technology#ourtechnology

[larryllix]

33 minutes ago, pgershon said:

That is a $300+ price-point.  It might be fine if I had confidence it would work, but its hard to know and I am not sure I want to just gamble with that kind of money.  Unfortunately the pump manufacturer has been less than helpful, suggesting I ask a local electrician to look at the equipment,

You may have developed a bad GFCI breaker and it is creating the noise. Filtering the pump may do nothing then but unloading the circuit may, just as your symptoms suggest. Have you tried to flip the breaker on/off a few times to wipe the contacts? They cost a few bucks so it wouldn't be cheap to replace it just to test the theory. The breaker could have developed bad contacts to the feeder bus in the panel also. Stab-Loc breaker types were notorious for that.

I doubt an isolation transformer would do anything. Insteon, with it's low KHz frequency is designed to be passed through transformer cores. The isolated core blocks high voltage capacitive  induction from jumping between windings, not magnetic induction.

[pgershon]

Thanks for your thoughts - your knowledge here dwarfs mine.  I have replaced the GFCI breaker.  In fact I installed a 20 Amp Siemans GFCI breaker to replace the regular 30 Amp breakers that had been there.  Installation of the GFCI was interesting in that the pump ran for about 1-5 mins (I did not time it) without interference initially.  I have not been able to duplicate.  Just to summarize, I have tried the following:

1) XPF filters (2 of them) - no effect

2) XPNR filters (2 of them) - made insteon signals fail even when pump was off

3) GFCI breaker

4) PZZ01/ Leviton 6254 - no effect

Additional observations:  My Insteon receiver receives signals from the insteon devices when switches depressed manually.  The interference occurs with sending commands to the insteon devices (which do not receive, acknowledge or take the desired action).  It definitely seems to be a one-way problem.  Also, when the pump goes off, it may take a few minutes for the interference to fully dissipate.  And when it goes on initially after being off for a prolonged period, the interference takes a few minutes before it becomes problematic.

[Brian H]

If the 40KHz noise source has harmonics in its waveform.

The third harmonic of 40KHz is the X10 120KHz power line signal frequency. It probably has no problem getting into an Insteon receiver. As X10 signaling was part of the original Insteon features.

If the noise is closer to 44KHz, That is a third harmonic of 132KHz. Right on top of an Insteon 131.65KHz command

[epfunke]

On 9/11/2020 at 4:24 AM, pgershon said:

I tried installing two XPF 20 Amp noise filters on the two legs of my pump load but it did not help.  I can add that the Intelliflo and another pump for my spa (spa pump not an issue) and a Heyward Aquarite Salt Chhlorinator all share the same 20 amp double pole breaker. The XPF was installed only onto the wires from the breaker to the Intelliflo directly

I have been following this thread and as an electrical engineer that has been using x10 and now insteon for years have been interested  in this thread.  After your last summary earlier today i re-read your original post to make sure i understood the topology.  In re-reading the post the (sentence in bold above) i would want to confirm  that the white wires from the filters were attached to a neutral and not just left disconnected.  That filter has two functions,  it blocks insteon / noise from passing from the black to the red leads, and it also shunts noise / insteon frequencies from the red to the neutral( white)side of this filter.   This does not absorb the Insteon signals from the supply side (black) as much as the XPNR  which basically is a shunt that does not care what direction the signal is coming from (supply vs load)  and as you stated earlier it is not surprising that it killed insteon signals.  That filter needs to be installed as possible to what is generating the noise preferably inside the offending equipment or in the very first junction box.

 

The other question that is raised is why insteon interference would be delayed?    and to clarify, insteon signals sent from pump house to house are received during interference, but things in pumphouse cant hear signals from the house?

[pgershon]

7 hours ago, Brian H said:

If the 40KHz noise source has harmonics in its waveform.

The third harmonic of 40KHz is the X10 120KHz power line signal frequency. It probably has no problem getting into an Insteon receiver. As X10 signaling was part of the original Insteon features.

If the noise is closer to 44KHz, That is a third harmonic of 132KHz. Right on top of an Insteon 131.65KHz command

So what would you suggest?

[pgershon]

5 hours ago, epfunke said:

I have been following this thread and as an electrical engineer that has been using x10 and now insteon for years have been interested  in this thread.  After your last summary earlier today i re-read your original post to make sure i understood the topology.  In re-reading the post the (sentence in bold above) i would want to confirm  that the white wires from the filters were attached to a neutral and not just left disconnected.  That filter has two functions,  it blocks insteon / noise from passing from the black to the red leads, and it also shunts noise / insteon frequencies from the red to the neutral( white)side of this filter.   This does not absorb the Insteon signals from the supply side (black) as much as the XPNR  which basically is a shunt that does not care what direction the signal is coming from (supply vs load)  and as you stated earlier it is not surprising that it killed insteon signals.  That filter needs to be installed as possible to what is generating the noise preferably inside the offending equipment or in the very first junction box.

 

The other question that is raised is why insteon interference would be delayed?    and to clarify, insteon signals sent from pump house to house are received during interference, but things in pumphouse cant hear signals from the house?

1) I can confirm that the XPF was also wired to the neutral in the junction box - I am going to edit prior post to reflect that.

2) The delay, I now believe, was artifact.  I had an old CR234 in my basement that was leftover from X10 days - it was causing periodic interference on top of everything else. I now have removed it and things work consistently with pump off and poorly with pump on.

3) Clarification is correct - signals from pool house travel to house no problem and get received with accurate reporting. Signals from house to pool house devices are the problem. I use Indigo and it reports errors when trying top communicate with the Insteon devices ing pool house (when pump is running) - these include ceiling fan and multiple lights (dual band).

4) XPNR was placed where the pump wires first come into Jandy box.  I cannot get closer to the device than that. The Jandy box connects to the pool house through a 70 amp breaker.  The XPNR blocked all signal from pool house (next breaker over) from communicating.  I will reinstall tomorrow to see if it affects signals to main house or just from it.

5) At the suggestion of a person on a different board, I am going to try a Uxcell AC 115/250V 20A CW4L2-20A-S Noise Suppressor Power EMI Filter as well.

[epfunke]

just so I am understanding things.  (after a google search on Jandy box)  that is the pump controller? so no insteon needs to get to it?  and it is on a 70A Breaker? so the xpf cant protect that circuit.  Is that the same as the intelliflow you had mentioned having the xpf on in the original post.

 

Any chance of a quick sketch showing the layout.  with approximate distances?

like   Main house:------ 100 ft ----- :pool house panel: 20breaker---10 ft? -----Insteon device

                                                            :pool house panel:  70a breaker- 15ft? -----  jandy box

 

Some other thoughts...  this noise is probably broadbanded with lots of harmonics,  you can try a an AM radio tuned as low as it will go and see if you hear any interference when the pump is on vs off?  try and use it as a sniffer?  Is it the pump motor itself?  or the controller? (my bet at this point) this may shed some light.  Also if you have a receiver probe for a tone tracer, that may also pick up the noise and help locate it

Or is the jandy box some other control and the intelliflow is the pump controller.

Knowing more about the topology may shed some light on this     

 

[epfunke]

That filter may help.  one key would be to keep that ground wire as short as possible 

[ELA]

Have you tried the XPNR, as close to the pump as possible, along with the XPF but closer to the pump breaker?

The idea being that the XPNR shunts the noise outside of the Insteon band near to the pump and the XPF isolates the Insteon devices from the noise source, at 132Khz and  closer to the Insteon devices.

Can we get an attenuation specification link to the Uxcell AC 115/250V 20A CW4L2-20A  filter?