Different L-G voltages on a 240v L-L PSU fed from a 3 Phase High Leg Delta secondary.

[MrMoe]

Working on installing some electronics supplied by a three-phase four-wire high-leg (center-tapped) delta secondary and am wondering whether a switching PSU at 240V, 2 hots and a ground (C13 plug) could run into problems when the line to ground voltages vary between the two hots going into the PSU. There is a possibility to switch to an ungrounded 240V delta secondary (120V supplied from elsewhere) or a 208V wye configuration so feel free to comment on that if you wish.

Details:

1. PSU is powered from a single phase panel that is completely wired with double-pole breakers in order to provide 240V L-L voltage (AC 60Hz).
2. One of the panel's hot line is the high-leg of the delta secondary, which means that the line-ground voltages from the hot legs into the PSU are 208V and 120V.
3. I don't have the schematics/diagram for the PSU but it's a generic 1800w server PSU; 1800w 200-240V AC to 12V DC (two V+ & 2 V- terminals). Internally the ground from the C14 plug is directly terminated to one the the PCB's mounting screws.
4. Powered the unit up to roughly 50% load without a problem, yet (to either PSU or equipment).

Would really like to know if this setup may present problems down the line, especially when connected to other equipment powered by the other phases (equivalent L-G voltage legs), be it via data cables, direct physical contact of cases, mounting racks, etc. Thanks!

 

[Little Bill]

If there are no line to neutral loads the high leg won't be a problem. The voltage that it would see is 240V L-L.

 

[MrMoe]

If there are no line to neutral loads the high leg won't be a problem. The voltage that it would see is 240V L-L.

Gotcha... Thanks for the quick reply!

Edit: Just wanted to add that there are no L-N loads on the high-leg and roughly 2% of center tapped transformer nameplate L-N load on the 120V circuits.

 

[texie]

Gotcha... Thanks for the quick reply!

Edit: Just wanted to add that there are no L-N loads on the high-leg and roughly 2% of center tapped transformer nameplate L-N load on the 120V circuits.

Keep in mind that if you use the high leg and another phase for a 240 single phase load you will need a non-slash rated breaker to be compliant.

 

[synchro]

If the switching supply is also rated for 240V 50 Hz then it can accept 240V RMS line to ground, and so the 208V L-G of the high leg will definitely be a non-issue.

 

[MrMoe]

Keep in mind that if you use the high leg and another phase for a 240 single phase load you will need a non-slash rated breaker to be compliant.

Great call.... We're aware of that and, luckily, we're working with overhauled military equipment (i.e. old stuff with new components) and they didn't skimp on this. All of the two/three pole BQ type breakers (is Siemens the only manufacturer making these?) that came along with it are 240V straight rated and tested good, which has been a massive blessing as those are neither cheap nor easy to source. I'm attaching a rather fuzzy pic (cropped picture of an entire panel).

 

[gar]

201019-2330 EDT

MrMoe:

Your first post describes a strange usage of a of a single phase panel. Are wires from the center tapped secondary not present? Not clear in your description.

.

 

[zbang]

OK... another high-leg question

Is this high-leg system made up or two or three transformers? I've only seen it with two, where the first (A-C) is larger and gets both the single and three-phase loads and the second makes it an open-delta between one side of the first (A) and the third phase (B). Like the diagram without the B-C transformer.



One of the things this implies is that power drawn only from the B and C legs is being supplied by both transformers; a nice recipe for overloads. If one is careful doing only AC and AB loads that won't be much of a problem. Might confuse people in the future, through. (If there is a third transformer, ignore all that.)

Previous opinion here, which I agree with, is that the high-leg should only be used to power actual 3-phase loads

To answer the original question- if a device isn't expecting a grounded conductor (neutral), then the voltage to that doesn't matter as long as it stays below the max for the device.

All that aside, if you can get a 3-phase wye feed, that'll be better all around. IMNSHO and all that.

 

[MrMoe]

201019-2330 EDT

MrMoe:

Your first post describes a strange usage of a of a single phase panel. Are wires from the center tapped secondary not present? Not clear in your description.

.

It's a single-phase panel with two hot lines coming in and a ground bar. Since the secondary is center-tapped, the L-G will read high on the high-leg hot coming in than on the other hot line coming in. There are no neutrals here, it's simply due to the fact that the center tap coming out of the transformer is grounded to the system electrical ground, which I'm assuming is why L-G is equal to L-N voltages if there was a neutral included in the mix.

There is another panel in the system that is also single-phase and dedicated to 120V loads. That one does have single pole breakers, common neutral busbar, and ground bar; only serves 120V. Then there are a few other 3 phase panels dedicated solely to 3 phase loads.

 

[MrMoe]

OK... another high-leg question

Is this high-leg system made up or two or three transformers? I've only seen it with two, where the first (A-C) is larger and gets both the single and three-phase loads and the second makes it an open-delta between one side of the first (A) and the third phase (B). Like the diagram without the B-C transformer.

View attachment 2554046

One of the things this implies is that power drawn only from the B and C legs is being supplied by both transformers; a nice recipe for overloads. If one is careful doing only AC and AB loads that won't be much of a problem. Might confuse people in the future, through. (If there is a third transformer, ignore all that.)

Previous opinion here, which I agree with, is that the high-leg should only be used to power actual 3-phase loads

To answer the original question- if a device isn't expecting a grounded conductor (neutral), then the voltage to that doesn't matter as long as it stays below the max for the device.

All that aside, if you can get a 3-phase wye feed, that'll be better all around. IMNSHO and all that.

Bingo! It's indeed an open delta, two-transformer setup. During troubleshooting one of the transformers was removed yet the engineer ordered a replacement and, as far as I can tell, is looking to keep the closed-delta center tapped configuration. I have to work given the assumption that I'll have 240V L-L from both setups so switching to a 208V WYE secondary would be a PITA (basically recalc loads, breaker types, etc) but I agree with you in that I'm not in love with the wild-leg. I'd really rather just have an untapped secondary, equal L-G across all lines, and derive the miniscule amount of 120V L-N loads from another available secondary service.

 

[gar]

202020-0919 EDT

MrMoe:

Did you actually make voltage measurements in said single phase panel to know that it has been wired in this strange, un-recommended, way? Light loading of a wild leg to neutral will work but should not be done.

.

 

[kwired]

Bingo! It's indeed an open delta, two-transformer setup. During troubleshooting one of the transformers was removed yet the engineer ordered a replacement and, as far as I can tell, is looking to keep the closed-delta center tapped configuration. I have to work given the assumption that I'll have 240V L-L from both setups so switching to a 208V WYE secondary would be a PITA (basically recalc loads, breaker types, etc) but I agree with you in that I'm not in love with the wild-leg. I'd really rather just have an untapped secondary, equal L-G across all lines, and derive the miniscule amount of 120V L-N loads from another available secondary service.

A major reason to go with open delta is to lessen cost when three phase is needed but the three phase load is minimal. Another reason is also to lessen cost on primary side when long distances are involved to a somewhat limited load and/or isolated premises (three wires vs 4 wires over that distance).

If all three primary phases are present and the load is significant enough that it would be best to balance across all three phases then full delta is generally preferred. Still is a choice of corner grounding it or mid point grounding one phase - mid point on a 240 system does give you 120 volts without needing to separately derive it. If the majority of load is 120 volts then a wye system would be more preferred when it comes to balancing the load.

 

[MrMoe]

202020-0919 EDT

MrMoe:

Did you actually make voltage measurements in said single phase panel to know that it has been wired in this strange, un-recommended, way? Light loading of a wild leg to neutral will work but should not be done.

.

I'd like to clarify that the load, assuming normal operating conditions, is pulling 240V Line-Line. However, given that one of the hot lines in this L-L configuration is the delta's high-leg, the line-ground voltages between the two lines are different. My concern is based on that.

I understand that ground and neutral are sometimes used interchangeably but I wanted to restate that, under normal operating conditions, the equipment is not pulling L-N loads.... Nothing in that panel pulls L-N; it's all L-L. There is a separate panel with 120V from the center tap to both lines and that does indeed use q hot and neutral, with both of the panel's hot wires being 120V L-N (or L-G for that matter)

 

[kwired]

I'd like to clarify that the load, assuming normal operating conditions, is pulling 240V Line-Line. However, given that one of the hot lines in this L-L configuration is the delta's high-leg, the line-ground voltages between the two lines are different. My concern is based on that.

I understand that ground and neutral are sometimes used interchangeably but I wanted to restate that, under normal operating conditions, the equipment is not pulling L-N loads.... Nothing in that panel pulls L-N; it's all L-L. There is a separate panel with 120V from the center tap to both lines and that does indeed use q hot and neutral, with both of the panel's hot wires being 120V L-N (or L-G for that matter)

A concern can be if there is surge protection components that are only designed for 120 nominal to ground then they may be continuously trying to clamp that 208 to ground on the high leg and will likely burn out that surge protection component with a continuous overvoltage applied to it.

 

[GoldDigger]

Capicitative noise filters on the input may be (unlikely but...) wired from line to ground. In that case the capicItve currents will not cancel and this could be a problem with a GFCI.

Sent from my Pixel 4a using Tapatalk

 

[MrMoe]

202020-0919 EDT

MrMoe:

Did you actually make voltage measurements in said single phase panel to know that it has been wired in this strange, un-recommended, way? Light loading of a wild leg to neutral will work but should not be done.

.

I obviously don't have the final word on how the secondary is and/or will be provided. The EE must have chosen it for a reason. However, your comment pretty much echoes most of what I've been able to find via internet searches and conversations with industry pros. Basically, most seem opposed to high-leg delta configurations.

I'm neither for nor against it. However, I do see some logic to using it when 97%+ of the load will be 3 phase and you have some single phase loads that will either be 120V or 240V. In this scenario, in order to balance the loads as best as possible, you pretty much have to place some of the L-L single (split?) phase loads on the high leg in order to balance out the other single-phase loads across all phases.

This goes back to my original question. Why is the high-leg center-tapped delta configuration unsuitable in this scenario and what are the negative consequences related to it? Why would a delta-wye config be a better option? I'm genuinely asking and definitely not looking to justify the current setup. I just want to understand why this configuration is "un-recommended", "antiquated", etc.

 

[gar]

202020-2246 EDT

MrMoe:

In my area I generally see two pole transformers. One is larger, and is the for the single phase center tapped loads. This is less expensive for the power company. If the three phase load increases they put up a larger wild transformer, or add a third transformer. Balancing of load is done by the way the primaries of different customers load the primary side.

.

 

[kwired]

I obviously don't have the final word on how the secondary is and/or will be provided. The EE must have chosen it for a reason. However, your comment pretty much echoes most of what I've been able to find via internet searches and conversations with industry pros. Basically, most seem opposed to high-leg delta configurations.

I'm neither for nor against it. However, I do see some logic to using it when 97%+ of the load will be 3 phase and you have some single phase loads that will either be 120V or 240V. In this scenario, in order to balance the loads as best as possible, you pretty much have to place some of the L-L single (split?) phase loads on the high leg in order to balance out the other single-phase loads across all phases.

This goes back to my original question. Why is the high-leg center-tapped delta configuration unsuitable in this scenario and what are the negative consequences related to it? Why would a delta-wye config be a better option? I'm genuinely asking and definitely not looking to justify the current setup. I just want to understand why this configuration is "un-recommended", "antiquated", etc.

Every individual installation has it's own set of circumstances that may or may not make this system favorable, and you need to factor into those circumstances whether it is a full or open delta and if conditions on primary side are factor in this, particularly the open delta option. If primary situation is somewhat limiting to only using open delta you can increase transformer sizes if the load would dictate so.

Basically there is no simple answer. If all three primary phase lines are available and the majority of loading is 120 volt loads then that is about the only time a wye secondary is pretty much a must.

 

[oldsparky52]

This goes back to my original question. Why is the high-leg center-tapped delta configuration unsuitable in this scenario and what are the negative consequences related to it?

Basically because the equipment manufacturers say so. If they say okay and something does go wrong it's on them. They will not make any money on agreeing that it's okay and they accept all liability if they say okay. Personally, I don't blame them. Why should they take on this liability?

 

[oldsparky52]

Did you ask the manufacturer?

You wrote

I don't have the schematics/diagram for the PSU but it's a generic 1800w server PSU; 1800w 200-240V AC to 12V DC (two V+ & 2 V- terminals). Internally the ground from the C14 plug is directly terminated to one the the PCB's mounting screws.

That seems to indicate everything is okay to do it your way.