IntellipH burned connector preemptive fix?

[JoyfulNoise]

Looking at all those pictures and …. WOW, the engineering team responsible for designing that did NOT do a good job. When dealing with high power electrical connections on a PCB, typically bladed connectors are used because large flat blades can dissipate heat better than rounded pin contacts. You see this a lot on refrigerator control boards where power for the compressor and the defrost heater needs to be switched. The connectors are all blade style and the PCB has very large area copper pads and traces to ensure heat spreading.

Me thinks they were too in a rush to get a product out the door …. But great DIY fixes! Keeping the current off the board will work well.

 

[bertschb]

I told my wife it was crazy that a multinational company like Pentair builds an electrical box that fails and the pool owners here actually know why the design is flawed and know how to fix it! It's amazing to me how companies drop the ball on their designs sometimes.

I'm a scuba diver and 20 years ago I purchased a canister light from a very respectable manufacturer of high end dive equipment. This was a very expensive light at the time (>$1,000). The first time I took it on a dive I was on the surface and went to flip the power switch to turn it on. No Bueno!!! There was no way to flip the switch with gloves on. I was diving a dry suit with dry gloves so removing my glove in the water to turn the light on was not an option. I had to swim back to the shore, remove my dry gloves, turn the light on, put the gloves back on and resume the dive. After that experience it was evident to me that NOBODY at that company had ever actually used this light while testing the design in the water. And yes, I could have tested the light myself with gloves on before using it but I'm not building and selling the lights so testing isn't something I would normally think to do. I ended up using my Dremel tool to modify the housing so you could turn the light on with gloves. I forwarded my fix to the manufacturer. They didn't reply but later designs looked just like what I did with my Dremel.

 

[JoyfulNoise]

I told my wife it was crazy that a multinational company like Pentair builds an electrical box that fails and the pool owners here actually know why the design is flawed and know how to fix it! It's amazing to me how companies drop the ball on their designs sometimes.

I'm a scuba diver and 20 years ago I purchased a canister light from a very respectable manufacturer of high end dive equipment. This was a very expensive light at the time (>$1,000). The first time I took it on a dive I was on the surface and went to flip the power switch to turn it on. No Bueno!!! There was no way to flip the switch with gloves on. I was diving a dry suit with dry gloves so removing my glove in the water to turn the light on was not an option. I had to swim back to the shore, remove my dry gloves, turn the light on, put the gloves back on and resume the dive. After that experience it was evident to me that NOBODY at that company had ever actually used this light while testing the design in the water. And yes, I could have tested the light myself with gloves on before using it but I'm not building and selling the lights so testing isn't something I would normally think to do. I ended up using my Dremel tool to modify the housing so you could turn the light on with gloves. I forwarded my fix to the manufacturer. They didn't reply but later designs looked just like what I did with my Dremel.

Field testing any equipment costs money. It’s a huge expense and often requires sinking a lot of money into a product before you actually sell any units. Good companies field test their products especially when massive liabilities are on the line (like cars, pharmaceuticals, medical devices, etc). Unfortunately, consumer product manufacturers won’t field test and just rely on field failures. Sadly, that makes the customer the guinea pig.

 

[bertschb]

I'm not an electrical engineer but this IntellipH control box looks simple even to me. I would think one of the most fundamental things you would do during design is make sure the wiring, connectors and PCB could handle the loads. And this device is specifically designed to be connected in between the main control panel and the SWCG. How difficult would it be to check the loads when they build all the equipment???? It's not like they are trying to design something that could be used by hundreds of other devices that they don't build and have no control over. I know virtually nothing about electricity and even I understand why the connector fails.

It's almost like a couple of rookie employees were sitting around the break room and said, "Hey, we could just put this new IntellpH controller in between the control panel and the SWCG! It works fine by itself so it should work fine connected to the SWCG. But Clevis, the SWCG draws a lot more power than the IntellipH. Shouldn't we test the wiring and connector to make sure they can carry the extra current? Naw, that would take a couple hours. I need to get home to watch the game."

 

[Turbo1Ton]

It's almost like a couple of rookie employees were sitting around the break room and said, "Hey, we could just put this new IntellpH controller in between the control panel and the SWCG! It works fine by itself so it should work fine connected to the SWCG. But Clevis, the SWCG draws a lot more power than the IntellipH. Shouldn't we test the wiring and connector to make sure they can carry the extra current? Naw, that would take a couple hours. I need to get home to watch the game."

They finished that idea on a Friday afternoon. No time to test. Forgot where they were on Monday morning.

Guy 1 - That IntellpH was good, right?
Guy 2 - I dunno, you know how much I drank over the weekend?
Guy 1 - Well, it's due today. We've gotta have something.
Guy 2 - Send it.

--Jeff

 

[JoyfulNoise]

I think the failure mode is non-obvious on a bench test because the connectors are likely spec’ed correctly and it’s a situation that requires accelerated wear testing to develop these types of failures. The wiring, connectors and board are probably within the limits of the devices connected. So only through actual field testing can problems be seen. As was alluded to in all the posts, this type of failure took more than a year to show up on an IC40 or even longer but showed up quicker on an IC60. And even that probably only revealed itself with heavy use of the IC60.

Bevis and Butthead in the product engineering department did their job … they built what the product and marketing mangers demanded they build. In fact, B & B probably saw the failure from a mile away but, because management is full of ding-dongs that barely graduated from business school and can’t spell “connector” without Word auto-correct helping them, they decided to do exactly as they were told to do and give those knuckleheads EXACTLY what they demanded. In fact, they likely warned the product team about the unusual nature of requesting that the IC power be routed through the IpH board instead of separate connections but the guys in marketing thought their idea was so cool and the IntelliCenter team didn’t want to hear about anymore modification demands to their product designs. So the higher ups get the failing grade for not listening to the engineers (cuz nobody wants to hear the words “we need more money for testing”) and not budgeting more money for field testing.

 

[bertschb]

So the higher ups get the failing grade for not listening to the engineers

I agree Matt. This is a much more likely scenario. I've been retired long enough that I forgot about inept upper management.

 

[Dirk]

I checked on the IntellipH warranty and it falls under the TradeGrade warranty (Pool Product Warranty Information). Meaning that if it was purchased online it is only a 60 day warranty. I don't know where Brian purchased his, so his might have had a 1 year warranty, if purchased from a brick and mortar store. If so, sorry about any confusion on that. It is worth noting that we seldom see these failures within 1 year, when running an IC40. The guys running IC60's are another story. I think they are failing within 6 months.

--Jeff

That seems unfair, as it is not at all like installing an IntelliCenter or a pump, but Pentair's usual MO is to neglect to contact me before making such decisions...

Thanks for the clarification.

 

[Dirk]

I think the failure mode is non-obvious on a bench test because the connectors are likely spec’ed correctly and it’s a situation that requires accelerated wear testing to develop these types of failures. The wiring, connectors and board are probably within the limits of the devices connected. So only through actual field testing can problems be seen. As was alluded to in all the posts, this type of failure took more than a year to show up on an IC40 or even longer but showed up quicker on an IC60. And even that probably only revealed itself with heavy use of the IC60.

Bevis and Butthead in the product engineering department did their job … they built what the product and marketing mangers demanded they build. In fact, B & B probably saw the failure from a mile away but, because management is full of ding-dongs that barely graduated from business school and can’t spell “connector” without Word auto-correct helping them, they decided to do exactly as they were told to do and give those knuckleheads EXACTLY what they demanded. In fact, they likely warned the product team about the unusual nature of requesting that the IC power be routed through the IpH board instead of separate connections but the guys in marketing thought their idea was so cool and the IntelliCenter team didn’t want to hear about anymore modification demands to their product designs. So the higher ups get the failing grade for not listening to the engineers (cuz nobody wants to hear the words “we need more money for testing”) and not budgeting more money for field testing.

And I think some, or even most, of the blame can be laid at the feet of the manufacturer of the connector. Cheap metal of some sort, tinned with something else of questionable quality, just enough to pass their own criteria and get the stamp for its capability (under pristine conditions). But once any sort of negative external force in introduced (like outdoor air conditions) and their spec crumbles. So even if the engineers spec'd the correct need, Pentair's buyer found the cheapest part with the right numbers from some Chinese catalog, and that one got the nod.

Then Pentair took advantage of the quantity discount and had who-knows-how-many boards built with it, all in one run at further discount, and now they're going to sell them until they're gone, and eat the loss of replacement, because it's still cheaper than recalling and tossing existing inventory. If it ever gets fixed, it'll be after the warehouse full of existing inventory is cleaned out.

 

[Flying Tivo]

Getting back to the Red>Black melting. @JoyfulNoise Matt, do you think the auto-reversing polarity of the cell makes the current flow more on the + than the -? Would that infer that the red melted ones dont reverse polarity as much as the ones that are equally melted?

 

[JoyfulNoise]

Getting back to the Red>Black melting. @JoyfulNoise Matt, do you think the auto-reversing polarity of the cell makes the current flow more on the + than the -? Would that infer that the red melted ones dont reverse polarity as much as the ones that are equally melted?

I would need to see a circuit diagram for both but I don’t think the polarity reversals plays into it. The IpH is supplying power to the IC. The internal IC electronics built on to the cell is controlling the switching of polarity. That should not affect the input power polarity at all.

 

[Dirk]

I would need to see a circuit diagram for both but I don’t think the polarity reversals plays into it. The IpH is supplying power to the IC. The internal IC electronics built on to the cell is controlling the switching of polarity. That should not affect the input power polarity at all.

That's correct. The red and black that routes through the IpH is not switching polarity, that's done internally in the IntelliChlor. Remember, the IpH circuit board is using the same DC power source, as is the control circuitry inside the IntelliChlor, so the polarity from the power supply is constant. But the current flow isn't. The IntelliChlor turns on and off its electrolysis every five minutes, and switches polarity to its cell plates just as often (I think), so the current is not at all steady throughout its daily run. Felipe, if you were asking if during one cell-polarity cycle the IC is using more current than the other polarity cycle, I suppose that's possible, but if it is different, it wouldn't be my much. If anything, it's all the "on-and-off-ing," rather than the polarity switching, that might be contributing to the conductor failure, kinda like how an incandescent light bulb's filament gets the brunt of its wear during the initial power up.

 

[Flying Tivo]

The red and black that routes through the IpH is not switching polarity, that's done internally in the IntelliChlor.

Probably me thinking in spanish and translating is not working very well. What i meant is the current draw! I understand that is does not switch. Current (amps)would flow on the red when in normal polarity and more through the black when in reversed polarity. Remember an MD trying to understand electricity!

 

[Dirk]

Probably me thinking in spanish and translating is not working very well. What i meant is the current draw! I understand that is does not switch. Current (amps)would flow on the red when in normal polarity and more through the black when in reversed polarity. Remember an MD trying to understand electricity!

Gotcha. No, it wouldn't work that way. The current that is traveling through the IpH would likely be identical regardless of the polarity on the IntelliChlor plates. Regardless, even if the IC draws slightly more current when the IC plates are in "+" mode or "–" mode, the current through the IpH's wires would be identical. Put another way, whether the IC is on or off, or producing chlorine or not, or reversing polarity this way or that, whatever amount of current running though any of the four power leads in the IpH (the two reds and two blacks), whether it's zero amps, or 10 or 9.9, whatever, is also running in the other three power leads. The amount of current does vary, depending on what the IC is doing, but all four wires each carry the same amount at any given point in time.

Think of it like the pipes in your pool's plumbing. Flow (like electrical current) is constant throughout it's path. Whether the pump is on high, or low, or medium, the amount of water running through the pipe feeding your pump is the same amount coming out of the pump, or out of the filter, or through your SWG. At whatever point you measured flow, it would be the same. If it's 5 GPM going into the pump, there will be 5 GPM coming out. And 5 GPM coming out of your filter, 5 GPM flowing back to the pool. If the pump RPMs increase, and the pressure increases, so that the pump is now drawing 10 GPM out of the pool, you'll be getting 10 GPM everywhere else, too.

Electricity works the same way. Even if the electrons are routing this way and that within a circuit board, whatever amount goes in through the black wire, comes out through the red.* Even if there is energy loss within the board, like heat dissipation from a resistor, or an LED emitting its energy via light and heat, the current going into the board would be measuring the same as the current coming out.

*Just for giggles: what's actually happening is the molecules in the wire are exchanging electrons. They jump from molecule to molecule, along the wire. If you push on one electron at one end of a wire, it'll jump to the next molecule, and push one electron out of that molecule to its neighbor. And so on, all the way to the other end of the wire. The first one that got pushed doesn't speed through wire and pop out the other end at the same time, that'd be a different electron. Eventually it'll make its way out, but not immediately. Like if you push a cup of water into a pipe full of water. A cup will come out the other end, but not that original cup. It'll get there, if you keep adding more cups of water, but it'll be some time later.

That effect, of pushing on an electron at one end to have another pop out at the other, happens at the speed of light (or thereabouts). That speed is constant. How much pushing is going on is like water pressure. That's voltage. How many electrons move from the pushing is current, like water flow.

Contrary to intuition, the push starts at the negative end. The electrons move from the negative side to the positive end.

That's all greatly simplified, of course, but that's the gist of it...

 

[bertschb]

@Dirk You lost me at "gotcha"

JK! I appreciate you taking the time to share your thoughts. It's helpful to those of us who lack a better understanding of how electricity works.

 

[JoyfulNoise]

Gotcha. No, it wouldn't work that way. The current that is traveling through the IpH would likely be identical regardless of the polarity on the IntelliChlor plates. Regardless, even if the IC draws slightly more current when the IC plates are in "+" mode or "–" mode, the current through the IpH's wires would be identical. Put another way, whether the IC is on or off, or producing chlorine or not, or reversing polarity this way or that, whatever amount of current running though any of the four power leads in the IpH (the two reds and two blacks), whether it's zero amps, or 10 or 9.9, whatever, is also running in the other three power leads. The amount of current does vary, depending on what the IC is doing, but all four wires each carry the same amount at any given point in time.

Think of it like the pipes in your pool's plumbing. Flow (like electrical current) is constant throughout it's path. Whether the pump is on high, or low, or medium, the amount of water running through the pipe feeding your pump is the same amount coming out of the pump, or out of the filter, or through your SWG. At whatever point you measured flow, it would be the same. If it's 5 GPM going into the pump, there will be 5 GPM coming out. And 5 GPM coming out of your filter, 5 GPM flowing back to the pool. If the pump RPMs increase, and the pressure increases, so that the pump is now drawing 10 GPM out of the pool, you'll be getting 10 GPM everywhere else, too.

Electricity works the same way. Even if the electrons are routing this way and that within a circuit board, whatever amount goes in through the black wire, comes out through the red.* Even if there is energy loss within the board, like heat dissipation from a resistor, or an LED emitting its energy via light and heat, the current going into the board would be measuring the same as the current coming out.

*Just for giggles: what's actually happening is the molecules in the wire are exchanging electrons. They jump from molecule to molecule, along the wire. If you push on one electron at one end of a wire, it'll jump to the next molecule, and push one electron out of that molecule to its neighbor. And so on, all the way to the other end of the wire. The first one that got pushed doesn't speed through wire and pop out the other end at the same time, that'd be a different electron. Eventually it'll make its way out, but not immediately. Like if you push a cup of water into a pipe full of water. A cup will come out the other end, but not that original cup. It'll get there, if you keep adding more cups of water, but it'll be some time later.

That effect, of pushing on an electron at one end to have another pop out at the other, happens at the speed of light (or thereabouts). That speed is constant. How much pushing is going on is like water pressure. That's voltage. How many electrons move from the pushing is current, like water flow.

Contrary to intuition, the push starts at the negative end. The electrons move from the negative side to the positive end.

That's all greatly simplified, of course, but that's the gist of it...

@Dirk You lost me at "gotcha"

JK! I appreciate you taking the time to share your thoughts. It's helpful to those of us who lack a better understanding of how electricity works.

So not to belabor the point, but the electrochemistry works a little bit differently but not much different. There is the simplistic/idealization of an electrochemical cell as nothing more than two metal plates with a conductive fluid between them and that fluid acts like an ideal resistor. Apply a voltage across the fluid and you get a fixed current....

... Unfortunately that's not the reality. The dynamic electrical properties of an electrochemical cell are quite a bit more complex. Not to get into the mind-numbing details, but an electrochemical cell actually looks like a short circuit (or very low resistance) initially for a very brief period of time and then it acts like a resistor in series with an impedance that can be modeled as a pure capacitor and a resistor in parallel on each plate. The details are not as important as the implications and one of the major implications of this in the operation of any electrochemical cell is the concept of an "in-rush" current. When you first turn a cell "ON", there can be a huge rush of current into the cell until the solution equilibrates by forming a capacitive double-layer near the electrodes. The in-rush current can be quite large but transient in nature, perhaps lasting less than a second. But, it is still a real current nonetheless and so anything that supplies power to the circuit has to be able to handle this large current load. This is the origin of why Hayward SWG controllers would constantly blow out a varistor on their power supply boards - the NTC varistor is part of the power supply circuit to effectively look like a really large resistance when the circuit first turns on. That way it acts to limit the amount of in-flowing current. Once the cell is running, and the varistor fully charges up and gets a little bit warm, it looks like a low resistance (effectively you can ignore it) in the circuit. Essentially the varistor was there to take the gut-punch of the in-rush current. The down side is that, after many cycles of doing this, the varistor will eventually wear out and not work anymore causing the power supply to fail. But, it's a $2 part and that, plus 10mins with a soldering iron, will get you back to working again. Incidentally, this is also why my wife's commercial treadmill kept popping the breaker to our bedroom randomly when she would turn it on - all bedroom circuits nowadays use AFI circuit breakers and, with the commercial treadmill pulling in a huge amount of in-rush current when the motor would engage (inductive loads suck in huge amps at first), the AFI CB thought it was an "arc event" and would pop. My fix, other than paying an electrician a huge sum of money to run a dedicated 20amp regular CB circuit into the bedroom, was to run a small extension cord (properly rated of course) to the GFI wall outlet right outside our bedroom door to power the treadmill. Worked perfectly after that.

My guess is that with the IpH, there is not only the steady state high current being demanded by the cell and running through the connector but also huge start up currents that constantly whack the connector on the board every time the cell turns ON from it's OFF state (so duty cycle matters). Add that all up and you get your very own accelerated wear testing ... and an eventual fried connector.

 

[Dirk]

So not to belabor the point, but the electrochemistry works a little bit differently but not much different. There is the simplistic/idealization of an electrochemical cell as nothing more than two metal plates with a conductive fluid between them and that fluid acts like an ideal resistor. Apply a voltage across the fluid and you get a fixed current....

... Unfortunately that's not the reality. The dynamic electrical properties of an electrochemical cell are quite a bit more complex. Not to get into the mind-numbing details, but an electrochemical cell actually looks like a short circuit (or very low resistance) initially for a very brief period of time and then it acts like a resistor in series with an impedance that can be modeled as a pure capacitor and a resistor in parallel on each plate. The details are not as important as the implications and one of the major implications of this in the operation of any electrochemical cell is the concept of an "in-rush" current. When you first turn a cell "ON", there can be a huge rush of current into the cell until the solution equilibrates by forming a capacitive double-layer near the electrodes. The in-rush current can be quite large but transient in nature, perhaps lasting less than a second. But, it is still a real current nonetheless and so anything that supplies power to the circuit has to be able to handle this large current load. This is the origin of why Hayward SWG controllers would constantly blow out a varistor on their power supply boards - the NTC varistor is part of the power supply circuit to effectively look like a really large resistance when the circuit first turns on. That way it acts to limit the amount of in-flowing current. Once the cell is running, and the varistor fully charges up and gets a little bit warm, it looks like a low resistance (effectively you can ignore it) in the circuit. Essentially the varistor was there to take the gut-punch of the in-rush current. The down side is that, after many cycles of doing this, the varistor will eventually wear out and not work anymore causing the power supply to fail. But, it's a $2 part and that, plus 10mins with a soldering iron, will get you back to working again. Incidentally, this is also why my wife's commercial treadmill kept popping the breaker to our bedroom randomly when she would turn it on - all bedroom circuits nowadays use AFI circuit breakers and, with the commercial treadmill pulling in a huge amount of in-rush current when the motor would engage (inductive loads suck in huge amps at first), the AFI CB thought it was an "arc event" and would pop. My fix, other than paying an electrician a huge sum of money to run a dedicated 20amp regular CB circuit into the bedroom, was to run a small extension cord (properly rated of course) to the GFI wall outlet right outside our bedroom door to power the treadmill. Worked perfectly after that.

My guess is that with the IpH, there is not only the steady state high current being demanded by the cell and running through the connector but also huge start up currents that constantly whack the connector on the board every time the cell turns ON from it's OFF state (so duty cycle matters). Add that all up and you get your very own accelerated wear testing ... and an eventual fried connector.

Great explanation! Better than my "light bulb filament when first turned on" analogy, but I think I was in the ballpark. #32

 

[bertschb]

The more you guys post here, the stupider I feel.

 

[Dirk]

Well, we're just tryin' to make ourselves sound smart!