3 wire 240v GFCI question

[thelorax121]

So my control panel has 2 separate inputs each with its own inline gfci, one 120v running to the power in and neutral busses, then a 240v that runs through a breaker, then a contactor (coil powered by the power bus and controlled with a 2 POS on/off switch) to the element. When I tried to auto tune the PID today, as soon as I switched the element to the on position, both GFCIs tripped, killing the power. When I did a little more research, I saw that a ground/neutral fault can trip the GFCI and since the 3-wire 240v outlet has the ground and neutral tied together, and that is in turn grounded with everything else (including the 120v inlet and receptacles), could thi have caused both GFCIs to trip, and if so, how would I wire that ground to avoid this issue? If this does not seem like the genesis of the problem, plese let me know and I can try to put up a more comprehensive diagram to get this bug figured out. Thanks for the help!

 

[P-J]

I sort of have a mental picture of your setup. I do believe you problem is as you describe. You cannot tie ground and neutral together. Also you cannot tie neutral fron 2 sources together. However, I think it would help a lot if you posted a diagram to really get to the root of the issue and for anyone to offer suggestions..

 

[DeafSmith]

Some (most?) GFCI's have a circuit to detect neutral to ground faults, so if you have a circuit loop where one side of the loop goes through the GFCI and the other side doesn't, the GFCI will detect a fault. Here's a description of how that works:

http://www.codecheck.com/cc/gfci_principal.htm

 

[thelorax121]

Here is a crude sketch of the power in, with all the grounds mounted to the same post in the control box.



Uploaded with ImageShack.us

DeafSmith
That's what I am thinking since both of the GFCIs tripped as soon as the the 240v input drew current. So if that is the case, how would I go about grounding the neutral/ground from the 240v input safely yet separately from the other grounds in the panel

 

[DeafSmith]

I can't see an obvious problem from your diagram - nothing that would cause the GFCI to trip when turn on the element. One other possibility to look for is a connection between the 240 volt hot line and the 120 volt hot line - this would have the same effect as a grounded neutral. I didn't see one in your diagram, but double check your wiring.

 

[P-J]

Also chech the neutral between the 120V source and ground or the ground from the 240V supply.

 

[samc]

I don't like to mix power sources. I am not an electrician but I can't see how it is safe to have the neutral from 120v source connected to the 240.

Personally I would have rewired the 240 to include neutral and not installed the 120v line. Seems safer and less chance for error.

 

[Walker]

I really can't reconcile the description and the diagram to understand what is connected to what here.

You are using the two hot lines from your incoming 240V cable for the element. That I can see and understand.

You are using the hot and neutral from your incoming 120V cable for pretty much everything else.

What are you doing with the third wire from that 240V cable and the ground wire from 120V cable?

 

[Wberry]

Perhaps I am miss reading this but have you installed a GFCI on the 240V circuit?

 

[Walker]

Perhaps I am miss reading this but have you installed a GFCI on the 240V circuit?

I believe he's got one of those 240v cords with an inline GFCI on it.

First post mentions that both the 240V feed and the 120V feed have GFCI.

 

[Wberry]

Yeah without some real information about all the involved devices, I think your tag line speaks worlds ... I would love to recommend that he meter the potentials between the different grounds and neutrals to ensure no ground loops/floating grounds/improperly bonded neutrals; but, it would not reveal anything with a complete picture of the circuit from the mains panel to his rig.

 

[rhoop]

I'm wondering the same as Walker, where's the neutral wire on the 240v? You said the neutral and ground were tied together on the 240v. Did you splice them all together inside your control panel? If so, I'd agree, that's probably your problem. If you aren't using the neutral on the 240v for anything, just put a marrett on it and don't use it. The only time the neutral and ground should ever be together is in your main panel where the power comes into the house. Anywhere else and it "can" be dangerous.

If that's not your issue...

On your schematic, you show the element having a neutral. Is this correct? Most (by most I mean I've never seen one with a neutral) straight up heat elements don't have one. If it does, it can't be hooked up onto the neutral bus. It would have to be hooked up to a separate neutral with the 240v line (not connected with the neutral from your 120v line at all).

Hope this helps.

I haven't done any brewing with electrical rigs, but I am an electrician... If I missed what your saying or can help out in any other way, let me know.

 

[rhoop]

Exactly Wberry, can't agree more.

 

[thelorax121]

I did not splice together the 240v ground and neutral, the outlet that I have available is a 3-wire NEMA 10-30R, so by spliced together I was referring to my understanding that in this older system, the ground and neutral are wired together as the same wire (this is the green wire coming out of my 240v inlet). This is the reason that I have a separate 120v inlet and what I think may be causing the issue (at least one of the factors at least) because I am running this combined wire to the ground post in my control panel where the 120v inlet and receptacles are also grounded.

As far as the element is concerned, there is no neutral running to it. I have a neutral wire from the neutral bus (from the 120v) and a hot line from its respective bus through the 2 pos NO switch that both run to the 120v coils of the 240v 40A element contactor. The load side of the contactor is then supplied by one hot line directly from the inlet and the other hot line runs from the inlet through the SSR driven by the PID then to the contactor.

I don't know if this clarifies anything, but let me know if you can glean anything from it. I am going to fiddle with it some today based on your suggestions, and if I can't get it still I will try to post some more detalied info

 

[samc]

You have a neutral from your 120 distribution block going to the coil for the 220, I don't think that is kosher.

 

[thelorax121]

How else would I get a neutral line to the coil? The contactor itself is rated to 240v, but the coil to power it is 120v

 

[rhoop]

Haha, my bad. I took the ground coming off the element in your diagram as a neutral. I guess I've got a worse cold than I thought. Let's try again. Powering the 240v coil from a separate source (the 120v line) is fine, so that shouldn't be an issue. If I'm understanding you right, you can start the control circuit up fine, so the 120v contact pulls in and stays in, and the power light is on right? But then when you flip/push the element switch, that's when both GFCI's trip?

If I understand this correctly, have you double checked the wiring to make sure that the 120v line and neutral don't connect anywhere? Especially the connection points on the 240v relay. It's easy to get them mixed up. If all the connections are good. Try turning off the 240v circuit, and then try turning everything (including the element) on. Yes, the element won't actually turn on, but the 240v contact should still pull in. Then try it the other way around (only if you feel comfortable working live!!!) Turn off the 120v circuit and turn on the 240v. You will have to manually push in the contacts on the 240v coil. See if the 240v gfci trips. Let me know what happens.

Work safe! Make sure you wear safety glasses when doing this! I'm not trying to talk down to you at all, but it only takes a quick slip and a 30a 240v explosion can blind you! Seriously, if you don't feel comfortable pushing in the coil, don't do it. There's other ways we can troubleshoot it. You might be perfectly fine doing it and that's great, but I don't want to give advice that gets somebody hurt, you know what I mean?

 

[thelorax121]

Ok, so I went with me gut reaction first when I went to figure it out and tried firing everything up after removing the ground/neutral line from the 240v inlet from the mutual ground screw in the panel and sure enough, everything fired up without an issue! Now my issue is what to do with that extra line in my panel, and how to go about grounding the inlet safely yet separately; any ideas?

 

[Wberry]

My 2 cents would be to simply cap the ground/neutral from the 240V and ensure that the ground from the 120V service is beefy enough to handle the amperage of a ground out of all the supplied amperage (both the 240 & 120 circuits).

IN THEORY - if your element went bad you would case ground to your rig and the only available path would be back through the 120V ground. Coming out of the industrial automation world ... I am paranoid about grounding on multiple levels. I would highly suggest ensuring that your WHOLE rig has a great ground of solid not stranded copper bonding everything - don't necessarily trust bolted and hinged connections to be current carrying capable. Great resources are looking at local code specs for hot tubs and above ground pools.

One last consideration, overall, is isolating the ground for your PID controller. To assist in lowering erroneous PID compensation, thus wearing our your element and amplifier, you will want your controller to have access to both a current ground and signal ground ... current grounds want to be solid/low stranded copper and signal grounds want to be highly stranded for more surface area.

Examples of this are as follows: Sensors, pump drives/amplifier control circuits, valve feedback and controller cards want signal grounds so they they share a common ground potential/reference. Items like pumps, heaters, frames, cases, panels all want current grounds rated for the max potential that can develop at that point in the system. The problem I see often in the threads posted here and other forums about people having a hard time tuning their control cards is a lack of understanding of types of grounds and what they are used for. Integrated electronics need a different approach then raw consumption devices - pumps, fans, etc.

Back to your problem though - I would meter the ground/neutral from your 240V supply to a KNOWN properly wired outlet in your house - using continuity on your meter to see if it is really a ground or a neutral. Please report back so that we can make a better recommendation and ensure you are safely wired. Unfortunately, electricity and proper wiring approach for projects like this are very dependent on how your house is wired. Let us know!

 

[samc]

How would he go about grounding a PID that does not have a ground? Not sure what OP is using but for example Auber PID that I have does not have a ground terminal and the case is plastic.

 

[Wberry]

First, thelorax121 - sorry for half hijacking your thread here; but, this is a good topic of discussion for those interested in a bit of theory. Additionally, if there are any TV/Radio Station Electricians that want to chime in - Please Do! They have the bigger problems with grounding and isolation then you can possibly imagine!

How would he go about grounding a PID that does not have a ground? Not sure what OP is using but for example Auber PID that I have does not have a ground terminal and the case is plastic.

OK so on the theoretical level and this is probably WAY too deep for this discussion - and to that end I think I went a bit off the deep end with my previous post; but, for the interested here is the control theory...

When utilizing PID controllers without a case, line, or sensor ground, you want the thermocouple (your temp sensor) to be electrically isolated from your entire system. Isolated/Ungrounded thermocouples have a much slower response time but an increased safety and control factor. This isolated state allows the controller to set its own ground/zero potential plane where ever it so chooses - +5V, -5V, +20V, 8002V, does not matter.

When utilizing PID controllers WITH a case ground and non-isolated sensor is when things become a bit tricky. You might want the high response with a steam system which has the capability to rapidly heat fluids. Hypothetically ... suppose your SSR, pump, or other piece of equipment is leaking current via its case and has raised the ground plane from 0V to 5V - meaning that 5 volts of potential is leaked back to the ground line through your entire system and your total potential between Line and Ground is now 115VAC as opposed to 120VAC. In order for the controller to not to have skewed input data, its ground plane at the controller must additionally be raised by 5V, there by putting the sensor and PID controller at the same reference potential.

This is why you have to match the style of thermocouple to the style of controller and additionally shows the importance of proper grounding NOT JUST FOR CURRENT but additionally for equal references for your electronics.

Sorry for the tangent but hopefully someone will find some resolution to their problems here.

Peace

 

[rhoop]

Hmm. Very interesting! So your suggesting that a separate ground be run for the PID if possible, and if not, making sure the thermocouple is isolated? Makes sense! I still wonder what caused both GFCI's to trip? There must be some sort of potential difference somewhere in the ground/neutral system to cause it. If it was one or the other, that would be easy, but both at the same time...

 

[thelorax121]

Yea I am still stumped about that as well, but as soon as I removed the g/n wire coming off the inlet from the gound bar and capped it, everything worked beautifully. As for grounding the PIDs, any idea on how to go about this, or how would you isolate the RTD sensors? Also, you metion that isolating may slow down the response time, by what degree do you think?

 

[P-J]

...
As for grounding the PIDs, any idea on how to go about this, or how would you isolate the RTD sensors? ...

The PIDs do not require grounding and have no provision for doing that either. As far as isolating the RTD goes, it is electrically isolated from the get go. The probe shell does not have any electrical connection with the sensor wiring. The probe wiring is isolated by the PID circuits - no worries.

 

[Wberry]

Sorry for the late reply here ...

GFCIs are a tricky subject. I have (2) possible guesses:

First, your pump(s) are leaking a bit (something over 5ma) of current across their windings, which is dissipated through the path of least resistance ... your ground on the 240V connection. In this case you would develop an imbalance in the GFCI supplying the motors and you have a false trip.

Secondly, your heating element has been saturated with moisture and upon firing it up there is a significant leakage to ground. This would additionally provide an imbalance between the GFCI's lines causing a trip.

The NEC is relatively particular about the utilization of GFCIs where ever the chance for water and electricity could meet; but, your situation is slightly complex with the mixture of 120V & 240V needs. In this instance, I would think the code correct way to wire your system would be as follows:

Have a 50A 4 wire receptacle installed. Then run your horse c**k of an extension cord (LOL) to your rig with a hard-wired connection at your control box. In the rig's NEMA 12 enclosure, I would mount a 50A mains breaker that then branch feeds several VERY tight curve breakers. You could then utilize an in-line GFCIs to protect your receptacles for your equipment. OR, you could switch the windings on your motor and see if your electronics will take 220 and power it all off the same circuit.

The only reason why I bring this up is that technically you are feeding a single device from (2) power sources and there is no single way (breaker/cord/gfci/etc) to remove power from the rig. Enough on that topic and back to the PID stuff ...

The short version of what I posted earlier is that there are (in simple terms) (2) predominant styles of thermocouples - ones without a ground and ones with. You need to make sure that, which ever style thermocouple you use, it is paired correctly with the same style of controller. Ungrounded thermocouples are very easy to install and relatively noise immune - their downside is their slow response time. Very fast responding thermocouples are grounded and require a significant amount of design energy in searching for possible electrical interference that can alter the reference plane.

I would bet that your sensors are isolated and that any shielding they have for electrical interference is taken to case ground where they meet your HLT/MLT/Kettle. Sorry, If I took you off track looking for other possible issues that likely do not exist.

Peace

 

[rhoop]

The motors could easily trip it, and the moisture in the element, but it would be a pretty big coincidence if it happened at the same time! Or does the controller do the pumps through control contacts not on the diagram as well as control the heat element? Are both GFCI's inline? Maybe they monitor ground as well as the common. Then it would be something as simple as some ground current, which we've already established. Do the GFCI's in the states monitor ground? I don't know of any consumer grade ones that do here in Canada, but then again, electricity works completely different once you cross the border.