3 phase power

IEEE 519 says plus 5% to minus 10% of nominal voltage..... except for short periods of time..... needless to say this was written by the utilities.

wye systems should have the same voltage phase to phase in all respects delta systems have what is called (here abouts) a wild/stinger or high leg.

What were the readings phase to ground? averaging or RMS meter? peak or average sampling?

Your readings at first glance look to me like an delta system that might be a tad low on the voltage.

DVM (rms) The problem is they have been in this building for years.

I go in to service the cnc machine tools when required, this morning one of the cnc machines was being knocked off line very time a welder lit an arc or another machine was started.

The guy says the welder dont have enough current even for a good arc, I checked the voltages just phase to phase and saw that. I know for a fact it used to be 240 +- 5 volts on all 3 phases.

They called the power company (edison) guy shows up says its ok, so I have to go back in the morning to be present when he shows up.

Oh one other thing while i was standing there conventional lath popped the overload on the main motor starter, every machine in the shop is somehow effected.

They have an adjoining building serviced by another transformer and all the voltages appear correct and that was what made me call the power company out.

Thanks for the response

Regards

Daveb

| 3 phase commerical power, mfg plants etc. either 480 or 240 when | reading line voltage between phase should be very close correct? | What tolerence would you expect? | | I just left a place 2 legs are 241 and one leg is 203 measured between | phases, whats up with that

Something sounds mixed up there. Either the wiring is, our your tests of it were.

You should carry out several voltage tests on all the wires with carefully recorded results. First, try to determine what type of system you appear to have.

If there are 5 wires, you have wye or tapped delta, with ground. If there are 4 wires, you may have wye or tapped delta, without ground.

For 208Y/120 you could have wire colors: black, red, blue, white, and bare/green

For 240/120 delta you could have wire colors: black red, orange, white, and bare/green

For 480Y/277 (or 416Y/240 or 600Y/347) you could have: three of: brown, orange, yellow, purple plus white/grey and bare/green

If there are 4 wires, you may have delta, with ground. If there are 3 wires, you may have delta, without ground.

Wire colors will be the same as above without the white/grey.

If the wires are colored, record the results of testing the voltage between each different color pairing. If they are not colored, find some way to label them for consistency, using letters A, B, and C.

For 240/120 delta, you should see 208 volts nominal (203 is within most standard ranges) between the "stinger" phase (orange wire) and neutral (white). You should see 120 volts nominal between either of the other phase wires (black and red) and neutral. You should see

240 volts between any pairing of (black, red, and orange). You should see 0 volts between neutral and ground, and the same voltage between any phase and ground as between that same phase and neutral.

Post your complete results, including wire colors. If you have any major machinery, try to do separate tests with that machinery running, and not running if you can get a chance to do that.

Will post the results in the morning after I go in there.

Thanks for the response

Regards

Daveb

Went in this morning and all phases checked normal as one would expect

240vac (a,b,c) and 120 to the neutral.

All plant equipment operated correctly,so have to assume it was something related to the power company (Edison).

Thanks for the information

Regards Daveb

From what you described, there was nothing wrong with what you initially measured and what the power company measured. You have a 4-wire delta service made up of 3-240 volt windings. One of the 240 volt windings is mid-tapped and tied to ground.

So if you label the mid-tapped winding a-b, you should measure 240 volts. a-n and b-n is then 120 volts. The other 2 windings b-c and c-a will measure 240 volts. However, if you measure c-n, you'll get something close to 208 volts.

Contrary to what was said earlier, you should see 1 to 3% unbalance between the phase-to-phase readings.

Since you're having problems with equipment dropping off, I'd be more concerned with how the voltages vary when the equipment starts. I'd hook up a PQ monitor and see how the readings change as the key equipment is started. A likely culprit to your problems is that you get excessive voltage sag (dip) when particular equipment is started. The monitor will tell you that.

If that's the case, you might have to get with the power company and increase the size of the supply transformers and/or service conductors. But the key is to hook up a monitor and take some readings.

Dan

| From what you described, there was nothing wrong with what you initially | measured and what the power company measured. You have a 4-wire delta | service made up of 3-240 volt windings. One of the 240 volt windings is | mid-tapped and tied to ground.

Or maybe a (Scott) Tee arrangement with a normal 120/240 volt transformer and a single 208 volt "stinger" transformer at 90 degrees (one of the transformers would be wired to the primary as P-N and the other as P-P to the other two phases). The advantages of this include no circulating currents in the secondary windings under imbalanced conditions, such as the 120/240 volt loads exceeding 5% of the load, and being able to size the stinger to the expected 3 phase loads. And it's cheaper, which utilities tend to like, since only 2 transformers are needed, instead of 3, and the total number of HV bushings would be 3 instead of 6 (unless they wire up wye-delta which adds the troubles of backfeeding when a phase is lost).

| So if you label the mid-tapped winding a-b, you should measure 240 volts. | a-n and b-n is then 120 volts. The other 2 windings b-c and c-a will | measure 240 volts. However, if you measure c-n, you'll get something close | to 208 volts.

That's where he may have gotten mixed up. 203 volts there would be within normal.

| Since you're having problems with equipment dropping off, I'd be more | concerned with how the voltages vary when the equipment starts. I'd hook up | a PQ monitor and see how the readings change as the key equipment is | started. A likely culprit to your problems is that you get excessive | voltage sag (dip) when particular equipment is started. The monitor will | tell you that. | | If that's the case, you might have to get with the power company and | increase the size of the supply transformers and/or service conductors. But | the key is to hook up a monitor and take some readings.

And there is the remote possibility of an intermittent loose neutral.

That's interesting, because I've never seen Scott-T configurations done with pole pigs. If they skimp and use only two transformers, it's open delta. Of course I don't know the internal wirings of "three phases in a can" pole pigs or pad mounted jobs.

If it's Scott-T, it can't possibly be wye-delta, can it?

Also, if you wire a Scott-T primary so one is wired A-C, and the other B-N, this produces a substantial neutral current even with a perfectly balanced 3 phase load. Normally it's done to a center tap on the A-C primary (of course this means oddball transformers which may be an answer to why pole pigs in Scott-T are unusual.

For what this is worth, I noticed in this small industrial park some of the poles ,I counted like 3 servicing about 5 units.

Some of the poles had 2 transformers and some had 3 and they all have three phase service .

The unit I was in was just a delta connection that had some problem that day and since has went away.

The pole people went today (again) and said things were A-OK (scary) but all phases read 240 as they should between A-B-C and they ran all equipment without problems as they have done for the last 10 years.

Great responses, learned something from the thread

Regards

Daveb

On Fri, 7 Jan 2005 22:33:56 +0000 (UTC) Michael Moroney wrote: | snipped-for-privacy@ipal.net writes: | |>Or maybe a (Scott) Tee arrangement with a normal 120/240 volt transformer and |>a single 208 volt "stinger" transformer at 90 degrees (one of the transformers | ... |>expected 3 phase loads. And it's cheaper, which utilities tend to like, | | That's interesting, because I've never seen Scott-T configurations done | with pole pigs. If they skimp and use only two transformers, it's open | delta. Of course I don't know the internal wirings of "three phases in | a can" pole pigs or pad mounted jobs.

I've seen setups which I guessed are Scott-T. One big fat transformer connected P-P with 3-wire secondary and one baby one connected P-N on the remaining phase with 2-wire secondary (had 3, but only 2 connected).

I've looked at the instruction manual or specs for several pad mount transformers, and they are mostly delta-wye with a few delta-delta.

|>since only 2 transformers are needed, instead of 3, and the total number of |>HV bushings would be 3 instead of 6 (unless they wire up wye-delta which |>adds the troubles of backfeeding when a phase is lost). | | If it's Scott-T, it can't possibly be wye-delta, can it?

Scott-T is a tee-tee. You can't characterize primary as either delta or wye since there are only 2 of them. You can have one of the T's turned

90 degrees, too. You can get 2 90 degree phases at different voltages either way. But one of them has to be P-P and the other P-N to get that.

| Also, if you wire a Scott-T primary so one is wired A-C, and the other | B-N, this produces a substantial neutral current even with a perfectly | balanced 3 phase load. Normally it's done to a center tap on the A-C | primary (of course this means oddball transformers which may be an answer | to why pole pigs in Scott-T are unusual.

That depends on your loads. A Scott-T won't serve a wye load properly. But it can serve a delta load (e.g. no neutral, all P-P windings or elements or whatever) without any neutral current. There will be current on the interconnection point at the transformers which would normally also be grounded. But you get that with a wye secondary, too.

Which is why I was wondering why you are talking about wye-delta configurations and backfeeding.

It should be able to if you tap the secondary of the transformer that is the stem of the T, and ground it (not the junction of the two transformers). The tap is at 0.577.

Of course this oddball tap is going to mean off the shelf transformers aren't going to be usable.

On Sun, 9 Jan 2005 03:22:51 +0000 (UTC) Michael Moroney wrote: | snipped-for-privacy@ipal.net writes: | |>|>since only 2 transformers are needed, instead of 3, and the total number of |>|>HV bushings would be 3 instead of 6 (unless they wire up wye-delta which |>|>adds the troubles of backfeeding when a phase is lost). |>| |>| If it's Scott-T, it can't possibly be wye-delta, can it? | |>Scott-T is a tee-tee. You can't characterize primary as either delta or |>wye since there are only 2 of them. | | Which is why I was wondering why you are talking about wye-delta | configurations and backfeeding.

I was comparing it to wye-delta in terms of total number of transformer bushings (one of many cost factors). I understand many utilities prefer P-N primary connections, and one of the reasons I've heard is that single bushing transformers are cheaper.

|>That depends on your loads. A Scott-T won't serve a wye load properly. | | It should be able to if you tap the secondary of the transformer that is | the stem of the T, and ground it (not the junction of the two | transformers). The tap is at 0.577.

Then you won't have 120/240 volt single phase power. All the P-N connects would be 138.5 volts.

| Of course this oddball tap is going to mean off the shelf transformers | aren't going to be usable.

Might as well just use a plain wye setup. But you have to decide what voltage you want.

There are a couple issues with mixing lots of single phase loads with some use of three phase power. A wye doesn't give you both 120 and

240 volts (you either do 208Y/120 or 240Y/138.5 or something in between like 220Y/127). Delta with a center tap limits the single phase loads to minimize circulating currents in the secondary windings. The Scott-T works out for big single phase loads and small three phase loads.

Getting both 240 volts and 120 volts from three phase, which are often needed due to the proliferation of single phase power and products that are designed for it, is a pain. A 6-pole 7-wire circuit would work, and could keep things in balance, but there are no distribution products for anything like that.

If the single/three load is particularly big, you need to balance between phases, then even Scott-T is not good. At that point you have to take power in a plain three phase and have three separate reasonably balanced single phase systems derived from it (which is essentially the 6-pole thing but each phase angle is kept separate).

Umm, yes, if one were to do this, it would mean simulating a 208Y/120 setup. No 240V. I was just addressing whether it could serve a wye load.

You'd need two transformers, one 208VCT, the other 180V, tapped at 2/3s (120V), with this tap grounded. Which means I got the tap wrong in my previous post.

Again, with the odd voltages/taps, no advantage to this over 3 120V in a wye configuration, except reduced pig count, except perhaps custom configurations with the pair permanently married, such as pad mounts.

On Sun, 9 Jan 2005 20:14:20 +0000 (UTC) Michael Moroney wrote: | snipped-for-privacy@ipal.net writes: | |>|>That depends on your loads. A Scott-T won't serve a wye load properly. |>| |>| It should be able to if you tap the secondary of the transformer that is |>| the stem of the T, and ground it (not the junction of the two |>| transformers). The tap is at 0.577. | |>Then you won't have 120/240 volt single phase power. All the P-N connects |>would be 138.5 volts. | | Umm, yes, if one were to do this, it would mean simulating a 208Y/120 | setup. No 240V. I was just addressing whether it could serve a wye | load.

A wye load would involve a neutral for some reason, such as the need to get a P-N voltage, such as 120 volts. The Scott-T, like the center tapped delta, gives you the expected P-N voltage on 2 of the phases, but not all three. If you can narrow down which phases need P-N and not more than 2 then you could hook it up carefully and succeed.

But what if you have a staged start motor that connects the windings to P-N first, then switched to P-P when it reaches a certain speed. You need a genuine wye supply for that.

| You'd need two transformers, one 208VCT, the other 180V, tapped at 2/3s | (120V), with this tap grounded. Which means I got the tap wrong in | my previous post. | | Again, with the odd voltages/taps, no advantage to this over 3 120V in a | wye configuration, except reduced pig count, except perhaps custom | configurations with the pair permanently married, such as pad mounts.

Scott-T cannot replace 208/120 wye. It probably can replace 120/240 center tapped delta in most cases (the move motor would not work here, either). If you need genuine 240 volts, you don't get that directly with 208/120 wye. If you have lots of single phase load with 240 volt needs, and a minimal about of 3 phase loads that can be wired as delta, then Scott-T should work. Otherwise get more transformers.

I'd be more inclined to get 480/277 wye and divide the single phase loads up 3 equal ways and put 3 480->120/240 single phase transformers in. But that's only if there's extra money available. But converting 208/120 wye to the 3 single phase 120/240 systems is even more expensive due to the costlier transformers for the odd ratio.

Do the math for the oddball configuration I mention. For phase B, it's

2/3s of 180V, or 120V. For phase A it's the vector sum of A-T (104V) and T-N (60V), or sqrt((208/2)**2+60**2) which is 120V. For phase C same math as phase A. Remember this is grounded at point N below, not at T.

B | | N | A-----T-----C A-C = 208V CT, B-T = 180V (B-N = 120V, N-T = 60V)

Voltage A-B = sqrt((208/2)**2+180**2) = 208V. B-C same. A-C = 208V by definition.

Again, with the screwball voltages/taps, no real logical reason to do this, unless you like spending more money than you have to, but you wrote "cannot" which is wrong.

If you ground at point T rather than at N, and increase the voltages somewhat (A-C = 240V, B-T = 208V) you get a 240V delta replacement with

120/240V single phase available.

Actually, the teaser transformer can be a standard transformer, as long as the pri/sec ratio is correct. For example, take 2400 delta primary and convert to 240 delta secondary using a straight T connection. Use standard

2400/240 transformers for both. The main transformer is obviously happy; the teaser runs at a primary of 2080 and of course a secondary of 208. Compared to a "correct" transformer, the copper loss is higher, due to "too many turns," but the core loss is lower, due to lower flux density and the attendant savings in hysteresis losses.

T-connected transformers may indeed be provided with neutrals and supply line-to-neutral loads; as a matter of fact, dry-type distribution transformers below 15kVA are generally T connected. Larger Ts are rare; the two halves of the main transformer operate at opposite power factors; one leading, one lagging, with the result that the mid-point voltage may drift. This may cause problems with phase unbalance when 3 phase motor loads are served, but for apartment buildings and other uses where single-phase loads predominate there is absolutely no reason why a T may not be used. When all is said and done, however, it still takes more transformer capacity with a T than with a proper set of 3 wye-connected transformers.

On Mon, 10 Jan 2005 22:47:21 +0000 (UTC) Michael Moroney wrote: | snipped-for-privacy@ipal.net writes: | |>| You'd need two transformers, one 208VCT, the other 180V, tapped at 2/3s |>| (120V), with this tap grounded. Which means I got the tap wrong in |>| my previous post. |>| |>| Again, with the odd voltages/taps, no advantage to this over 3 120V in a |>| wye configuration, except reduced pig count, except perhaps custom |>| configurations with the pair permanently married, such as pad mounts. | |>Scott-T cannot replace 208/120 wye. | | Do the math for the oddball configuration I mention. For phase B, it's | 2/3s of 180V, or 120V. For phase A it's the vector sum of A-T (104V) and | T-N (60V), or sqrt((208/2)**2+60**2) which is 120V. For phase C same math | as phase A. Remember this is grounded at point N below, not at T. | | B | | | | | N | | | A-----T-----C A-C = 208V CT, B-T = 180V (B-N = 120V, N-T = 60V) | | Voltage A-B = sqrt((208/2)**2+180**2) = 208V. B-C same. A-C = 208V by | definition. | | | Again, with the screwball voltages/taps, no real logical reason to do this, | unless you like spending more money than you have to, but you wrote | "cannot" which is wrong. | | If you ground at point T rather than at N, and increase the voltages | somewhat (A-C = 240V, B-T = 208V) you get a 240V delta replacement with | 120/240V single phase available.

I'll stand by my "cannot" because the above is not exactly Scott-T. But it is close. I know what you mean, and that _can_ substitute for

208Y/120.

You can get 3-phase out of network power like this, too:

A * \ / \ N B / C

There are all kinds of crazy schemes that could be cooked up to get the voltage and phase you want:

B / * \ N---* / \ A---* C

If you can get 160 volts on 3 pairs of secondary windings, you can have

480Y/277 from the above.