Electric System Problems

[3LeggedDog]

Background: I bought a used 2-vessel electrical system a couple of years ago from someone who had largely assembled the control panel themselves.

It had worked great for me for 20+ batches but I noticed that it was failing to heat for a brew session a few months back. I noticed that the component labeled the AC Contactor had overheated and begun melting. I was able to find a replacement online, require, and get things back up and going.

This was good for another 10 or so batches, but yesterday, it failed to kick the heating element on, and I noticed that the same AC contractor had melted, as had the Solid State Relay and the wire coming from it.

Disclaimer: I am so far from electrically competent it’s not even funny. I know how to use this system to make beer, and I can use a screwdriver, but that’s about it!

My question: I can certainly replace these components and get things back up and running again. But I was curious as to whether there might be other more significant problems that you all can identify from the pictures. One thing that I wondered about was if the rust on some of the screws might be causing issues? Otherwise I’m at a loss for what might be causing these components to fail.

Would value any help or resources you all can provide. Happy to answer any questions that might clarify things!

Thanks a bunch!

 

[sibelman]

Wow. The most typical causes of overheating like this are poor connections and under-sized wire.

It's also possible that the components (contactor, SSR) are under-sized. Can't see the photos you may have attached (Google error).

Rebuilding with appropriate wire and well-terminated connections would likely fix your problem. However, the evident fire risk and your lack of electrical knowledge/skills make continuing with this unknown panel rather dicey.

 

[stamandster]

First off, most of your images aren't loading. Instead of linking to google, just attach them to the post. Second, depending on the element wattage, the wiring at first glance looks severely under sized to/from the SSR (only image that loaded) for anything for than about 2kw. When a wire is undersized for the resistive load it will basically overheat the wire (especially un-insulated junctions). It looks like the new SSR is 40a which _should_ be enough for whatever element you are running in a typical home brewery.

 

[3LeggedDog]

Background: I bought a used 2-vessel electrical system a couple of years ago from someone who had largely assembled the control panel themselves.

It had worked great for me for 20+ batches but I noticed that it was failing to heat for a brew session a few months back. I noticed that the component labeled the AC Contactor had overheated and begun melting. I was able to find a replacement online, require, and get things back up and going.

This was good for another 10 or so batches, but yesterday, it failed to kick the heating element on, and I noticed that the same AC contractor had melted, as had the Solid State Relay and the wire coming from it.

Disclaimer: I am so far from electrically competent it’s not even funny. I know how to use this system to make beer, and I can use a screwdriver, but that’s about it!

My question: I can certainly replace these components and get things back up and running again. But I was curious as to whether there might be other more significant problems that you all can identify from the pictures. One thing that I wondered about was if the rust on some of the screws might be causing issues? Otherwise I’m at a loss for what might be causing these components to fail.

Would value any help or resources you all can provide. Happy to answer any questions that might clarify things!

Thanks a bunch!

Apologies about the photos embedding. They're attached here.

 

[doug293cz]

overheat the wire (especially un-insulated junctions)

Never heard this. Can you elaborate?

Brew on

 

[stamandster]

Do you have a full shot of the entire interior? Also, looks like your running without a ground? What size is your heating element? What wire size are you using for the high wattage side?

 

[stamandster]

Never heard this. Can you elaborate?

Brew on

I simply meant that the connection, if improperly done, can cause issues that can lead to degradation. If you've ever been in a panel that's smoking, it normally not the wire itself but the connection at the breaker.

 

[doug293cz]

I simply meant that the connection, if improperly done, can cause issues that can lead to degradation. If you've ever been in a panel that's smoking, it normally not the wire itself but the connection at the breaker. But... I feel like your just trying to call me out on something that's been miscommunicated.

Yes, I am very familiar with the heat damage due to high resistance connections, usually caused by loose or corroded connections. And, that is definitely what is going on in OP's panel.

It was the "un-insulated" comment that confused me.

Edit: Should have trimmed the quote in my previous post better, as I can see how my question could be misinterpreted as asking about loose connections overheating wires.

Brew on

 

[stamandster]

I mean, the connection is technically "un-insulated" In any event, I agree, based on the images there looks to be a lot of connection corrosion.

 

[3LeggedDog]

Do you have a full shot of the entire interior? Also, looks like your running without a ground? What size is your heating element? What wire size are you using for the high wattage side?

Appreciate all of your feedback on this front. It's been super helpful.

Two pictures of the full interior are attached here. The heating element is 5500W.

Looks like 10 gauge wire is what they used on the wattage side of things.

 

[stamandster]

Thanks! Trying to follow from the heating element plug, where to the white and black wires connect into?

 

[3LeggedDog]

Thanks! Trying to follow from the heating element plug, where to the white and black wires connect into?

It’s tough to follow in the picture. Bit of a neat down there. I believe that those are both flowing into the left AC Contactor.

 

[whoaru99]

10ga should be fine for that level of current. You'll have confirm the contactor and other parts ratings.

As others have alluded to/said, just from a visual standpoint it looks like there were loose and or dirty connections, and quite possibly poor crimp of the terminal onto the wire with that extent of localized heating/melting.

Those segment(s) of wire need to be replaced using proper size terminal and good crimps, in addition of course, to all the other burnt/melted stuff. Any poor connection not addressed/re-used is just another event in the making.

Might be worth looking into wire ferrules too for any stranded connections that go directly under a screw&clamp type connection.

 

[sibelman]

https://www.homedepot.com/p/Gardner...ade-Terminals-Yellow-15-Pack-15-116/202522898

 

[stamandster]

@whoaru99 is referring to these ferrule connectors -- https://www.amazon.com/Terminals-Connectors-Assortment-Connector-Insulated/dp/B0CJ2PX84M
Just an opinion, but I would also suggest stepping up to 30a contactor relays for the element.

 

[sbro]

Looks like some corrosion on the wires at the connectors and some rust on a few screws on junction blocks indicating a humid environment.

check the screws are torqued correctly all the way from the heating receptacle -> contactor ->ssr -> terminal block and if you find any that were loose that could have contributed to the problem. Next remove that same wiring that you just checked the screws and look for signs of corrosion on the copper. You should replace any wires that have signs of overheating or corrosion.
I agree with whoaru99 about using ferrules on the stranded wire connection points where appropriate.

 

[whoaru99]

For signal type/low current connections the spade terminals, stacking, and the terminal legs spreading like above probably isn't causing any significant issues, although it doesn't look good.

But, for high current connections, I'd go with ring terminals I opine they may provide more contact area and are not subject to the stacking misalignment and leg spread as exhibited in the photo in previous post.

 

[Whisky River]

Does anyone here use non-conductive dialectric grease on the connections when assembling their controllers to try and mitigate corrosion in the hot damp brewing environment?

 

[Control Freak]

"Just an opinion, but I would also suggest stepping up to 30a contactor relays for the element."
Agree with this part of the diagnosis. I checked my panel and the contactors are rated at 63 Amps and that's for a 30amp supply and 5500w elements.
I'm assuming you need to upsize the contactors because of the frequency of the pulsing of the PID controller.

 

[whoaru99]

I'm assuming you need to upsize the contactors because of the frequency of the pulsing of the PID controller.

I wouldn't think so off top of my head. 100% duty cycle/full on seems like it would be worst case for the contactor and that the pulsing would be on average less current (presuming SSR is what's actually doing the pulsing and the contactor is just on/off master / safety cut out type of thing).

 

[stamandster]

I may or may not trust a "BAOMAIN" relay to be rated at what it says. In the end, I would trust 30a rating more than the 25a rating for better margin of error when running 5500w/240v=22.9a for an extended period of time. He has to replace it anyways, I think the cost is negligible, and it could help reduce troubleshooting issues in the future to take it out of the equation as a non-EE or similar.

 

[doug293cz]

I wouldn't think so off top of my head. 100% duty cycle/full on seems like it would be worst case for the contactor and that the pulsing would be on average less current (presuming SSR is what's actually doing the pulsing and the contactor is just on/off master / safety cut out type of thing).

What usually causes contactor failure is contact element erosion from arcing over many thousands of make and break cycles. Make and break at higher currents causes more damage than at low currents. So, a marginally rated contactor would likely fail sooner than an over rated contactor at the same currents.

I don't see any evidence of contactor or SSR failure in OP's pics - just connections that overheated due to being loose/corroded. It is the heat generated at the high resistance connection points that caused damage to the plastic device packages, not failure of the devices themselves.

Brew on

 

[Snowkiwi]

From your photos, the part that has melted down is a Solid State relay, the Data sheet for that particular part number gives the load rating at 40 Amps. If you are supplying a purely resistive load of 5500 Watts at a supply voltage of 120 volts a.c. (I am guessing your supply voltage as I am not in your country), your load current would be somewhere around 45 Amps. The SSR will not like switching that load too often before it malfunctions. If you can check by measuring the load current that will confirm the actual current is about right for the wattage of your element. You may need to try an SSR with a higher current rating as a trial.

 

[whoaru99]

From your photos, the part that has melted down is a Solid State relay, the Data sheet for that particular part number gives the load rating at 40 Amps. If you are supplying a purely resistive load of 5500 Watts at a supply voltage of 120 volts a.c. (I am guessing your supply voltage as I am not in your country), your load current would be somewhere around 45 Amps.

USA has most commonly 120/240 split phase power for residential service. A 5500W element typically would be using 240V unless running at lower voltage to get reduced output.

 

[stamandster]

For those that aren't familiar, in most residential, 120/240 split (single) phase in the US is 2 x 120v lines that are 180 degree's out of phase from each other that can supply 240v by using both phases together. This is why you can run 120v in the same control panel as the 240v parts. This is different than EU which supplies 240v on a single hot, returns on a single neutral. However, in the OPs control panel there's no wire for EGC, which does bring up some safety concerns (can't have a GFCI).

 

[whoaru99]

That's a common description but for me that's sorta backwards from the actual physical (technical?) nature of the system.

It's inherently a 240V transformer but with a center tap that facilitates deriving two 120V legs from the 240V transformer, not really two 120V legs giving 240.

Some may say the point is semantics but that is the physical nature of the system.

 

[stamandster]

Yes, absolutely accurate! I was simply describing what the end-user "usage experience" is, and not from the panel to the pole delivery. Both are accurate

 

[doug293cz]

From your photos, the part that has melted down is a Solid State relay, the Data sheet for that particular part number gives the load rating at 40 Amps. If you are supplying a purely resistive load of 5500 Watts at a supply voltage of 120 volts a.c. (I am guessing your supply voltage as I am not in your country), your load current would be somewhere around 45 Amps. The SSR will not like switching that load too often before it malfunctions. If you can check by measuring the load current that will confirm the actual current is about right for the wattage of your element. You may need to try an SSR with a higher current rating as a trial.

All 5500W elements are intended for 240V. The max common current capability of 120V residential circuits is 20A. A 5500W element at 120V will only put out 1375W. A 5500W element has a resistance of 240^2 / 5500 = 10.47 ohms. At 120V, a 10.47 ohm load will only draw 120 / 10.47 = 11.46A.

Also, if you look closely at the picture of the melted SSR package, the melting is located at one of the terminal screws, which means that's where the heat was generated. If the excess heat was coming from the TRIAC (the active power switching device in the SSR) then the melting would be at a more central position of the SSR package. Thermal failures at connection points are almost always due to excessively high contact resistance in the connection.

Brew on

 

[doug293cz]

However, in the OPs control panel there's no wire for EGC, which does bring up some safety concerns (can't have a GFCI).

On the contrary, there is in fact a ground (earth) wire from the supply that is bonded to the enclosure and carried thru to the element and pump outlets.

Brew on

 

[whoaru99]

And, even if there was no equipment grounding wire, one could/can still have functional GFI.

 

[doug293cz]

And, even if there was no equipment grounding wire, one could/can still have functional GFI.

But, the "Test" button probably won't work.

Brew on

 

[stamandster]

On the contrary, there is in fact a ground (earth) wire from the supply that is bonded to the enclosure and carried thru to the element and pump outlets.

Brew on

Indeed, and this is a good infographic. And I thought a test button would still work as it uses a resistor to leak a small amount of current to create an imbalance? Also, I see the ground in the panel now in the images.

 

[whoaru99]

But, the "Test" button probably won't work.

Brew on

Whether it's been a GFI receptacle I've installed in old work situations where there was no equipment grounding wire, or a GFI in a panel where there is no equipment grounding connected to the GFI, the test buttons worked.

What the GFI did have though, in those cases, was a neutral connection which of course is grounded; but it's not the grounding wire of topic in this point.

The imbalance detection is using whatever wires/connections pass though the current sense circuits of the GFI, that's line and neutral or lines and neutral, but grounding wire doesn't.

 

[doug293cz]

Indeed, and this is a good infographic. And I thought a test button would still work as it uses a resistor to leak a small amount of current to create an imbalance? Also, I see the ground in the panel now in the images.

View attachment 851752

Whether it's been a GFI receptacle I've installed in old work situations where there was no equipment grounding wire, or a GFI in a panel where there is no equipment grounding connected to the GFI, the test buttons worked.

What the GFI did have though, in those cases, was a neutral connection which of course is grounded; but it's not the grounding wire of topic in this point.

The imbalance detection is using whatever wires/connections pass though the current sense circuits of the GFI, that's line and neutral or lines and neutral, but grounding wire doesn't.

My understanding was that the "leak" resistor was connected from a hot to ground (since the ground does not pass thru the current sensing coil.) Thus when there is no ground, you can't create the leakage that bypasses the sensing coil. To work without a ground, the test switch and resistor would have to connect line to line, or line to neutral, with this connection being routed to bypass the current sensing coil. Is this how GFCI Test circuits are actually wired?

Brew on

 

[whoaru99]

...

 

[whoaru99]

My understanding was that the "leak" resistor was connected from a hot to ground (since the ground does not pass thru the current sensing coil.) Thus when there is no ground, you can't create the leakage that bypasses the sensing coil. To work without a ground, the test switch and resistor would have to connect line to line, or line to neutral, with this connection being routed to bypass the current sensing coil. Is this how GFCI Test circuits are actually wired?



Brew on

Perhaps more convincing than my personal anecdotes and experiences saying it, here's info from Leviton showing it.

The press to test button and test current limiting resistor connections are between line and neutral.

 

[KyBeer]

10ga should be fine for that level of current. You'll have confirm the contactor and other parts ratings.

As others have alluded to/said, just from a visual standpoint it looks like there were loose and or dirty connections, and quite possibly poor crimp of the terminal onto the wire with that extent of localized heating/melting.

Those segment(s) of wire need to be replaced using proper size terminal and good crimps, in addition of course, to all the other burnt/melted stuff. Any poor connection not addressed/re-used is just another event in the making.

Might be worth looking into wire ferrules too for any stranded connections that go directly under a screw&clamp type connection.

I used 8AWG for my 40 amp 240 volt circuits in my electric brew system I built. 10 AWG is good for loads lasting less them 7 seconds,( see section 7 of the national electrical code book), but for longer load times a larger gauge wire can handle the long load times. I would verify all connections are tight. Arcing can result in higher resistance conections. ( Ohm's law states that if the voltage remains constant and the resistance increases the current will follow, thus P=I x E, the result is heat!).

 

[doug293cz]

Ohm's law states that if the voltage remains constant and the resistance increases the current will follow, thus P=I x E, the result is heat!

This sounds like you are saying that at constant voltage, if the resistance increases, then the current will also increase. This is wrong. Ohm's Law is: V = I * R, where V is voltage, I is current, and R is the resistance thru which the current flows. Ohm's Law can be rearranged as: I = V / R, which clearly shows that at constant voltage, if the resistance increases, the current decreases.

What you want to look at when it comes to localized heating is this formula: P = R * I^2, where P is power. Since the current flowing in any branch of a circuit is constant everywhere, you can break the circuit branch into individual pieces, each with their own resistance value. The heat generated in each of the pieces is then determined by the current and the resistance of the particular piece.

Let's look at a simple example:

The fixed voltage is 240V, and the branch of interest is two connections with a nominal resistance of 0.005 ohm (5 mohm) each and a heating element with a resistance of 10.30 ohms. The total resistance of the branch is 10.31 ohms, and the current flowing will be 240 / 10.31 = 23.28 A. Now, if one of the connections goes bad, and the connection resistance increases to 1.0 ohm, then the branch resistance increases to 11.305 ohms, and the current drops to 240 / 11.305 = 21.23 A. For the original connection, the heat generated in the connection will be 0.005 * 23.28^2 = 2.71 W. For the degraded connection, the heat generated will be 1.0 * 21.23^2 = 450.7 W, which is more than enough to overheat a small volume and burn the insulation.

Brew on

 

[Broken Crow]

I used 8AWG for my 40 amp 240 volt circuits in my electric brew system I built. 10 AWG is good for loads lasting less them 7 seconds,( see section 7 of the national electrical code book), but for longer load times a larger gauge wire can handle the long load times. I would verify all connections are tight. Arcing can result in higher resistance conections. ( Ohm's law states that if the voltage remains constant and the resistance increases the current will follow, thus P=I x E, the result is heat!).

The OPEC oil-crunch happened when I was a kid. Did you know that if you got in a head-on collision while driving a 64-72 Mustang at 55MPH you stood a good chance of walking away, but if you were in a '74 or later the engine would rip through the firewall and tear your lower torso off? (I've towed the wrecks away from both)
For my entire life I've seen the ongoing battle between experienced engineers and the bean-counters armed with wet-behind-the-ears 'engineering graduates' engaged in life or death battles on behalf of the consumers.... From materials to tooling, everything is consantly cheaping out so normally I appluad over-rating of components but at a certain point it becomes overkill. Somewhere in Detroit about 18 years ago is a warehouse that held 1/4 million popular engines that coudn't be installed in vehicles because there was a risk of the connecting rod (sintered metal) fracturing and blasting right through the block....it happened because of interns who looked at the cycle life of cutting tools wanting lick boots to save pennies citing the textbook number of cycles the tool was supposed to be good for....when the cutting tools dulls, it generates heat and causes the metal to swell and makes a deformed high-friction connection....just sayin.
Our electric codes tell us what is required and sure you can exceed that and as anyone in this game knows you can buy 6500W or even 8000W heating elements, but we rely on the defacto 'standards' of the homebrewing community and don't want to burn our wort so we optimise with 5500W elements with the greatest surface area we can get.
I guess this is my long-winded and drunken way of saying that 10-gauge wire and a max 5500W element with properly rated SSR and clean secure connections is more than adequate for brewing purposes...equipping for anything more is just wasteful... unless you're using a very long extension cord and boiling for many hours.

 

[sibelman]

10-gauge wire and a max 5500W element with properly rated SSR and clean secure connections is more than adequate

Yes... and 10ga. is rather easier to work with than 8ga. Heavy enough wire is crucial, but issues at the connection points seem to make the most trouble for electric brewers.