Low leg on 3 Phase 480V
I was called out to a residential midrise building adjacent to Pacific Coast Highway to investigate fluctuating voltage that feeds an LG Multi V series 4 roof mounted A/C . The A/C units keep frying their circuit boards up on the roof and the manufacturer is claiming it is due to voltage fluctuations.
LG Multi V-IV: Variable Refrigerant Flow Cooling System | HVAC | LG India Business
I started in the basement on the 480V 400 amp dedicated feeders and found two legs at 290 and the center leg at 260. This circuit is feed from it's own 500 KvA Delta-Y transformer. The voltages were constant but the current was constantly fluctuating between 25 and 80 amps and my meter would flash 62.5HZ every so often. Inspecting the distribution panel and the roof disconnects it became obvious why the current was fluctuating, when these units cycle on and off they ramp up and down slowly. I've never worked on a VFD type system before other than mounting one of those Big Ass fans with the 12' blades. These units were amazingly quiet and smooth as the current ramps up and down. There was no voltage variance as the units cycled though. All the readings from basement and roof were consistent so there was nothing abnormal other than the center leg being low.
I've read through all the posts on this site I can find about voltage fluctuations and there was some great input about if one of the phase legs is 5% low it can cause a 20% increase in current. I can't remember the exact numbers that were posted but the gist of it was that you need to derate three phase motors by 50% because of one low phase voltage.
So my question is wouldn't the motors burn out and not the control boards due to a low phase voltage? Or are VFD drives very finicky? My hunch is that it is the salt air from being so close to the ocean.
The Utility primary has a voltage monitor and it shows it as supplying 4888 on two legs and 4964 on the 3rd leg on a 4800V system. So the Utility feed has one high leg. There are 3 other transformers connected to the primary feeding adjacent buildings. If those transformers are out of balance what is the chance it will cause a low leg on the transformer feeding the A/C units? It seems I will need to go back and measure for a low leg on the output of the other 3 transformers to isolate whether it is a utility problem or not.
LG Multi V-IV: Variable Refrigerant Flow Cooling System | HVAC | LG India Business
I started in the basement on the 480V 400 amp dedicated feeders and found two legs at 290 and the center leg at 260. This circuit is feed from it's own 500 KvA Delta-Y transformer. The voltages were constant but the current was constantly fluctuating between 25 and 80 amps and my meter would flash 62.5HZ every so often. Inspecting the distribution panel and the roof disconnects it became obvious why the current was fluctuating, when these units cycle on and off they ramp up and down slowly. I've never worked on a VFD type system before other than mounting one of those Big Ass fans with the 12' blades. These units were amazingly quiet and smooth as the current ramps up and down. There was no voltage variance as the units cycled though. All the readings from basement and roof were consistent so there was nothing abnormal other than the center leg being low.
I've read through all the posts on this site I can find about voltage fluctuations and there was some great input about if one of the phase legs is 5% low it can cause a 20% increase in current. I can't remember the exact numbers that were posted but the gist of it was that you need to derate three phase motors by 50% because of one low phase voltage.
So my question is wouldn't the motors burn out and not the control boards due to a low phase voltage? Or are VFD drives very finicky? My hunch is that it is the salt air from being so close to the ocean.
The Utility primary has a voltage monitor and it shows it as supplying 4888 on two legs and 4964 on the 3rd leg on a 4800V system. So the Utility feed has one high leg. There are 3 other transformers connected to the primary feeding adjacent buildings. If those transformers are out of balance what is the chance it will cause a low leg on the transformer feeding the A/C units? It seems I will need to go back and measure for a low leg on the output of the other 3 transformers to isolate whether it is a utility problem or not.
1 - 20 of 21 Posts
Electrical contractor 37 years. Electrical inspector 2 years
Joined
·
4,141 Posts
You might be better seeking an HVAC specialist. A lot of these ductless and variable refrigerant systems are great and energy efficient but the electronics are tempremental and sometimes require additional TVSS . My brother is always sending me information on the new products and how to protect them but I am still old school and not into the new equipment.
That is what the building engineer is planning on doing. He has two in mind that protects from under voltage, A PMD series from Macromatic and the ICM450A from ICM. These units will protect the system against undervoltage. Unfortunately if the voltage is always out of tolerance then the A/C units will never run.
vfd's don't really care about the incoming voltage. The output side of the vfd will be balanced so the motors going to be happy.
A lot of meters struggle with anything on the vfd output side so im not sure i would trust any reading on that side unless you have a fluke meter.
As for the site power just call poco and tell them you have a voltage imbalance. They may bitch and tell you its with-in spec but there probably going to check it out just to keep you happy. Sounds like they have a regulator clocked back (Doesn't sound like a poor connection as you are stable at different loads). Now depending on why they have the regulator clocked they may play nice and fix the issue (unless they have it that way to fix someone else's issue).
Honestly if im looking at bad boards im looking at multiply levels of surge suppression as low voltage may mess with the chips and cause a program glitch that can be recovered by rebooting, high voltage spikes will fry the boards.
A lot of meters struggle with anything on the vfd output side so im not sure i would trust any reading on that side unless you have a fluke meter.
As for the site power just call poco and tell them you have a voltage imbalance. They may bitch and tell you its with-in spec but there probably going to check it out just to keep you happy. Sounds like they have a regulator clocked back (Doesn't sound like a poor connection as you are stable at different loads). Now depending on why they have the regulator clocked they may play nice and fix the issue (unless they have it that way to fix someone else's issue).
Honestly if im looking at bad boards im looking at multiply levels of surge suppression as low voltage may mess with the chips and cause a program glitch that can be recovered by rebooting, high voltage spikes will fry the boards.
That is a huge phase to phase imbalance. Definitely need to get that figured out as it will cause overheating issues. Are the loads balanced on the transformer secondary?. What are you voltages with the secondary disconnect open? What is your voltage feeding the transformer?
I can't check the primary voltage at the transformer but the Utility feeder has a meter and the voltages are 4888 for A & B and 4964 for C on a 4800 Volt system. Their feeders split off to power 3 separate transformers. The one I'm dealing with is the newest at 500KvA, the other two are so old I couldn't find the name plate. Based on their size I'd say they're 115 KvA. All the loads on the A/C transformer are 3 phase so I don't see how they could be imbalanced. I was checking for that but with the units cycling on and off and the current going up and down there is no way to get an accurate reading without having a meter that will clamp all 3 legs at once.
What do you think the chances are that an imbalance on the other transformer are causing the low leg?
What do you think the chances are that an imbalance on the other transformer are causing the low leg?
Poco uses automatic voltage regulators (buck and boost transformers). These are independent so you can buck or boost any one of the phases.
These could be out of calibration or one could have a problem so its no longer adjusting in auto. poco normally insists on a 7% out of balance as a emergency call or 5% for a urgent call but as your not calling them out in the middle of the night 3.6% will get someone interested in the next few days.
I have never been charged by poco for coming to take a look after reporting a imbalance (there meter is showing a 3.6% imbalance which needs to be corrected before trying to fix the voltage on your side). Most of the time its a voltage regulator that accidentally got left in manual after being calibrated.
Apparently the circuit boards in these LG units are pricey, the management told me they spending $15,000 per visit to have one or two boards replaced and it's happening every few months. Also they assessed each homeowner $50,000 to upgrade the HVAC system and the boards take a month to get so there is a lot of cold cranky old people in the building. Even if they had to pay the City to fix the supply, if it solved the problem it would be well worth it. Thanks everyone for the input.
Once the imbalance is fixed you need to install tvss in multiply areas. I live in florida and we either install surge suppressors or replace very expensive parts on a regular basis. Anything over a 17 seer ac unit can not handle a voltage spike especially the blow fan motor control board.
Joined
·
14,398 Posts
I don't have any kind of educated guess here but I am thinking the electronics in the units have to be ahead of the VFD. I'd think the VFDs can use fairly crappy power to make good power for the motors but the electronics don't get any help from that.
With such expensive boards, surge protection is a must, but surge protection doesn't do anything for imbalances. The clamping voltage before they start to operate is WAY over 3% overvoltage. I'd install it anyway but I'd really like to hear more why the manufacturer's engineer said the problem is being caused by voltage imbalances. I am not saying I doubt it, I just don't understand how it would happen. If for example there's a power supply inside the unit making say 24V for the controls, and the voltage imbalance on the source is putting out 25V or 23V, that wouldn't fry the electronics.
If these units really are sensitive to that little bit of voltage imbalance, unless the utility can correct it, I don't think there are a lot of practical options. Conditioning this much power would cost a fortune and take up a lot of space. Maybe the transformer taps can be altered to adjust? I'd also want some kind of light or other monitor in case the utility ever corrects the imbalance and the taps are now creating an imbalance in balanced input power rather than correcting an imbalance in unbalanced input power.
With such expensive boards, surge protection is a must, but surge protection doesn't do anything for imbalances. The clamping voltage before they start to operate is WAY over 3% overvoltage. I'd install it anyway but I'd really like to hear more why the manufacturer's engineer said the problem is being caused by voltage imbalances. I am not saying I doubt it, I just don't understand how it would happen. If for example there's a power supply inside the unit making say 24V for the controls, and the voltage imbalance on the source is putting out 25V or 23V, that wouldn't fry the electronics.
If these units really are sensitive to that little bit of voltage imbalance, unless the utility can correct it, I don't think there are a lot of practical options. Conditioning this much power would cost a fortune and take up a lot of space. Maybe the transformer taps can be altered to adjust? I'd also want some kind of light or other monitor in case the utility ever corrects the imbalance and the taps are now creating an imbalance in balanced input power rather than correcting an imbalance in unbalanced input power.
@gpop does your Poco take power quality trouble calls? I ask as the one I deal with up here has told us to call in a power quality trouble call when we suspect it may be a bad supply issue. We've done it a few times and it gets them to drive the line and check their gear. Maybe it's the same thing, just different terminology?
Joined
·
5,750 Posts
VFDs are not really all that susceptible to voltage imbalances, all they are doing is rectifying the AC to DC. A severe imbalance can affect the lifespan of the internal DC bus capacitors because they end up working harder to smooth out the resultant ripple on the DC bus.
But if the PC boards that are frying are not the boards inside of the VFDs, then this has nothing to do with the VFDs. PC boards never run off of line voltage so there are power supplies somewhere providing them with their operating voltage. If the boards are failing because of line problems, those would have to be the CRAPPIEST power supplies on the market! Most modern power supplies are now built to handle a really wide range of input voltage. The ones I use for 480V supplies will accept anything from 260 to 550VAC input.
One thing that would be more likely to me is that the control boards are picking up stray voltages on their analog and/or digital input wiring and they have inadequate filtering to protect against that. This can happen especially when someone ran control wiring with unshielded cables in parallel with power cables. That may have been the HVAC contractor or it may even be an internal wiring issue in their units.
So most likely this is a QC issue with their equipment or installation instructions and they are throwing out this “voltage imbalance” as a red herring to distract their customers from the real issue.
But if the PC boards that are frying are not the boards inside of the VFDs, then this has nothing to do with the VFDs. PC boards never run off of line voltage so there are power supplies somewhere providing them with their operating voltage. If the boards are failing because of line problems, those would have to be the CRAPPIEST power supplies on the market! Most modern power supplies are now built to handle a really wide range of input voltage. The ones I use for 480V supplies will accept anything from 260 to 550VAC input.
One thing that would be more likely to me is that the control boards are picking up stray voltages on their analog and/or digital input wiring and they have inadequate filtering to protect against that. This can happen especially when someone ran control wiring with unshielded cables in parallel with power cables. That may have been the HVAC contractor or it may even be an internal wiring issue in their units.
So most likely this is a QC issue with their equipment or installation instructions and they are throwing out this “voltage imbalance” as a red herring to distract their customers from the real issue.
Let’s back up.
@RAD COM what are the phase to phase voltage readings, these are the only ones that mater for a 480volt load.
something don’t add up tap wise from your readings.
you say 4800 volt primary but say it is running high.
your phase to ground readings should be over 277 if primary is high but you-read low.
it looks like the poco is running high for other transformer loads but no one bumped up taps for output for this 500kva one.
@RAD COM what are the phase to phase voltage readings, these are the only ones that mater for a 480volt load.
something don’t add up tap wise from your readings.
you say 4800 volt primary but say it is running high.
your phase to ground readings should be over 277 if primary is high but you-read low.
it looks like the poco is running high for other transformer loads but no one bumped up taps for output for this 500kva one.
1 - 20 of 21 Posts
