3 phase AC unit troubleshooting

Status
Not open for further replies.
goldstar

goldstar

Senior Member
Location
New Jersey
Occupation
Electrical Contractor
I had a situation last week (with the record breaking heat) where a roof-top AC unit was blowing one of its 30 amp time delay fuses in the disconnect and shutting down the unit. My first reaction was to replace the fuse and clip an amp probe on to see what was happening. The unit had a 3-phase, 208V input. According to the nameplate info the max. draw on any leg should have been 19.6 amps. I had an analog meter with me showing two of the legs drawing approx 24 amps and one drawing approx 18 amps. That didn't seem too far out of line so I closed up the disconnect, waited around for a half hour or so to make sure it kept running OK and left. However, a half hour after I left it shut down again.

I went back the following day with the AC guy (who claimed it was an electrical problem and not an AC problem) and re-opened the disconnect, replaced the fuse and powered up the unit. I clipped my analog amp probe on the problem leg and he clipped on his digital amp probe. In a matter of five minutes or so I saw my analog meter reading go from 24 amps to 40 amps for about 5 seconds and then drop down but, his digital amp probe display didn't move. I also tested for a change in voltage but all phases to ground remained the same at approx 120 VAC and phase to phase approx 208 VAC. We then proceeded to check every connection point from the disconnect, to the sub-panel to the feeder distribution panel and all connections were tight. I would not have even gone down to the distribution panel except for the fact that the AC guy claimed that just this season three out of the eight units on the roof were experiencing the same problems.

Now, it is my thinking that a fuse blows because something on the load side is causing an increase in amperage. Could a loose connection on the distribution side cause the same scenario ? Any thoughts or opinions would be appreciated. Also, is it common for digital amp probes to not react as quickly as analog ones do ?

Thanks in advance,

Phil
Gold Star Electric,
New Jersey
 
goldstar said:
I had a situation last week (with the record breaking heat) where a roof-top AC unit was blowing one of its 30 amp time delay fuses in the disconnect and shutting down the unit. My first reaction was to replace the fuse and clip an amp probe on to see what was happening. The unit had a 3-phase, 208V input. According to the nameplate info the max. draw on any leg should have been 19.6 amps. I had an analog meter with me showing two of the legs drawing approx 24 amps and one drawing approx 18 amps. That didn't seem too far out of line so I closed up the disconnect, waited around for a half hour or so to make sure it kept running OK and left. However, a half hour after I left it shut down again.
Click to expand...

The nameplate says 19.6 amps, you are fused for 30 amps (slowblow), and read 24 amps on two legs? I think the 24 amps on the two legs is telling you something. What is your wire size?

Are the compressors hermetic? If so, are they HOT to the touch? Are the refrigerant pressures in normal range?

Yes, digital meters react slower. Ever check a capacitor with a digital ohm meter?
 
Minuteman said:
...

Yes, digital meters react slower...
Click to expand...
True, but there's something wrong with one or the other meters (or both) if the analog one measures a 16A increase for 5 seconds and the digital one doesn't measure any change at all.
 
Mike,

I didn't check to see if the refrigerant pressure was normal or if the compressors were hermetic. While I would agree that having that knowledge would help me in this situation I draw the line at checking voltage and amperage.

Wire size is # 10 AWG. The other wierd thing I forgot to mention was that while the 3 phase disconnect was properly fused at 30 amps the circuit breaker feeding the unit was sized at 60 amps. I changed that out real quick but God knows how long it's been that way.
 
I thought maybe you had a conversation from the HVAC guy about the equipment. A hermetic compressors motor windings are cooled by the refrigerant. A A/C unit with overly high discharge pressure will cause the motor to run HOT. (Lack of maintenance causes high discharge pressure.)

The fact that the amp draws are high on 2 legs and normal (or low) on one suggest to me equipment problems, not problems in line voltage.
 
Phil, the AC is pulling around 20 amps while running. It has #10 wire on a 30 amp breaker. Can't you put a 40 amp breaker and maybe change the disconnect to a non-fused if you want to save on costs or fusible and fuse at 40 amps? Do you think maybe the start up amperage is getting to you? Does the nameplate have like a 30 minimum and 40 maximum on the fuse size? I don't know but being a 3 phase it sounds like 20 amps is ok for that unit.

We always try to bump up the wire size on an AC unit because of the start up amperage.

Also, you have not mentioned how far the unit is from the breaker box. Is it far enough away to cause a good amount of voltage loss?

Also what about the starting contacts, are they pitted or is a contact loose? I have almost the same problem on a rooftop AC here and I am going to look for loose wires, bad contacts, loose fuse holders, etc.

Keep us posted on you problem. I am curious as to what comes about to try and save myself some trouble if the same problem comes to me down the road.
 
minuteman said:
The fact that the amp draws are high on 2 legs and normal (or low) on one suggest to me equipment problems, not problems in line voltage.
Click to expand...
That had been my position from the start but I'm starting to change my opinion. BTW, I wasn't the original contractor on this job. Whoever did this installation was a shoemaker. The run is abour 40' from breaker panel to the roof-top unit with one splice point just before it penetrates the roof. Yes, I was with the AC contractor at the time I inspected the unit but we didn't talk about refrigerant or hermedicaly sealed AC compressors.

DD, I did just what you suggested. Upon the recommendation and approval of the AC contractor I changed the disconnect on the first unit to a 60 amp and fused the unit at 40 amps. It seems to be holding up fine but I believe that is a band-aid cure at best. There is an inherent problem somewhere in the building but finding it is going to be a nightmare.

I got a call today from the same office building that another tennant is experiencing the same problem with their AC unit. Somewhere along the line I have the feeling that there's a loose connection, a main breaker on its way out or an annealing problem (didn't know what that was until I looked it up in the dictionary). The problem is occurring in only one half of the building. I've checked connections right down to the main switchgear and found nothing out of the ordinary. After I get there tomorrow I may decide to call POCO in to have them determine if there are any loose crimps or connections at the transformers. I'll keep you posted.

Phil
 
That's a really great idea but the problem is that I'm working for a customer that is air tight and believes that I will one day go out there, wave a magic wand, correct the problem and not bill him for anything. I keep going out on these emergencies and I'll keep billing him for my time. If I don't find the problem soon or when he's had enough billings from my exploratory work he'll stop calling me and hire another poor, unsuspecting soul and start all over again. This is the type of person that invests the absolute minimum into the maintainence of his building and figures that once something is installed it will last forever !!! To give an example, a year and a half ago I submitted a price to replace ballasts and bulbs in all his hallway light fixtures. More than half of them were out. As of today the only ones I've changed were in the stairwell fixtures and that's because his tennants were griping that they couldn't see and it was a safety problem. In reality, I don't have high hopes of finding this problem but who knows, maybe today I'll get lucky.
 
Smart $ said:
..there's something wrong with one or the other meters (or both) if the analog one measures a 16A increase for 5 seconds and the digital one doesn't measure any change at all.
Click to expand...
If one meter is measuring true RMS, then harmonics would be included, while averaging meters miss harmonic current frequencies.

24A / 40A is a 60% difference, not small, and if not compressor related, a simple frequency meter/function only sees amperage spikes as harmonic, if that (+ or -) sequence is the predominant magnitude. Zero sequence (tripplens) are suposed to be unique to neutrals.

A periodic 16 amps of harmonic current on top of 24 amps of fundamental current may find your freqency meter jumping around a bit, from high to low, but will not give you a traceable signature to follow back to the source. Minuteman's harmonic analyzer may be the best tool for this, 43B rentals are about $200. a week. My local Ford Electronics shop carries EXTECH, and also offers a try B4 buy policy.

If this 16A jump was harmonic, rather than compressor related, it would require a large nearby harmonic load. VSD's generate similar harmonics on phases, but other intermittend loads could also leak similar frequencies into the system; the more common culprits are SCR's, lighting balasts, printers, copiers, & computers.

--'It is better to be thought a fool than to open your mouth and remove all doubt.'
-- Mrs. Bora, 2nd Grade, Spring 1972
 
Last edited:
Good info Roger. Would harmonics generating a 40 amp pulse on one phase cause a 30 amp time delay fuse to blow ?

BTW, niether meter had true RMS features.
 
I'd say, if the fuse is thermal it should'nt matter what frequency the overcurrent was, as long as the time-delay / overcurrent duration was exceeded.

However, if neither meter measured RMS, then an intermittent undervoltage (v-drop) may be equally culpalble of causing overcurrent with a large enough load pulling on the same phase.

If your going to monitor that phase with a frequency meter for harmonics, I'd also monitor it with a volt meter or logging meter. Then crunch the power figures to see how much v-drop is needed to reach the extra 16A on that phase.
 
Last edited:
Status
Not open for further replies.
Top