PDA

View Full Version : Rotary phase converter with high voltage on made up leg



jeremy13
12-05-2012, 07:54 PM
My home built rotary has 193V on the made up leg. First some info rotary is a 25Hp I have start and run caps. The rotary is powering a 15Hp hydraulic pump on my plastic injection machine. Iím getting L1-124V L2-123V L3-193V with all the run caps removed I still get L3-174V. With the hydraulic pump running and all the heaters on I can get it down to 165V on L3. So whatís next, will moving the legs on the rotary motor help balance the voltage? I know this will work for the amps.

lakeside53
12-05-2012, 08:40 PM
Are you measuring between phase and neutral/ground? Neutral or ground are not valid with the generated leg of an rpc.

L1-L2 should be 240 (or you local mains which will be close to this) period; T1- T2 will be (and are) the same. Then measure L1/T1-T3, and L2/T2-T3 and report back.

The Artful Bodger
12-05-2012, 09:15 PM
Can you please post a circuit diagram of your RPC and show where you are taking the measurements?

Thanks

jeremy13
12-05-2012, 09:39 PM
I will get the line to line voltages tomorrow; I don’t recall them good enough. I was checking the load on the motor leads one the injection machine and got L1- 28A L2-27A L3-9A witch made me start checking voltages.

Diagrams of RPC Ha-ha Diagrams? We don’t need no stinking diagrams!

I copied a bought one from a gas compressor. I took the voltage readings at the injection machine motor mag switch.

The Artful Bodger
12-05-2012, 09:59 PM
Diagrams of RPC Ha-ha Diagrams? We don’t need no stinking diagrams!


Well I do if I am to make any constructive suggestions. However I am sure you will get it sorted without me.

Fasttrack
12-05-2012, 10:00 PM
I copied a bought one from a gas compressor.

What size was the one you copied and, if it is not the same size as the one you built, did you adjust the run capacitors accordingly?

flylo
12-05-2012, 10:07 PM
Diagrams of RPC Ha-ha Diagrams? We don’t need no stinking diagrams!

:mad:Gotta have Stinking Pics & Diagrams Or you may not get any Stinking help!:p LOL! Just Kidding:rolleyes:

Rich Carlstedt
12-05-2012, 11:19 PM
Forget the voltage...Thats why its called " potential"
Pay attention to amps, thats where your work is done

Rich

J Tiers
12-05-2012, 11:31 PM
Assuming it is indeed measured from ground, or at least neutral, as the voltages seem to show, that is maybe a tad low, but not too bad. Try measuring line to line, as has been mentioned, and see what you get. From your measurements I would think it will be much better balanced as to voltage when measured that way.

Yes, the amps are the real issue, but if the thing is truly balanced, the amps and volts will go together. Balancing really means getting the power factor to a decent point, and that brings both the amps and volts closer to balance between the legs.

In a true 3 phase system, the three legs should be balanced as to both current and voltage. The RPC cannot quite be balanced, unless you add some inductance in the "pass-through" lines, because the manufactured leg has the impedance of the idler motor in series with it, while the "pass-through" legs do not.

Adding the balance caps tends to bring up the voltage and correct the phase.of the manufactured leg. The correction is somewhat load-dependent, which is a trade-off that is normally made optimum at full load, the point at which balance is most important.

jeremy13
12-06-2012, 08:43 AM
The Artful Bodger-I meant no disrespect that was an old off color movie comedy quote (Blazen Saddles) adapted to fit.

The gas compressor was a 15Hp. And I upped the capacitors accordingly.

I’ll get a few pics when I check voltage later probably around lunch.

rdhem2
12-06-2012, 10:48 AM
What is the real problem? I have not seen one posted yet. You stated most of the real problem yourself, third, or made up leg. IMHO rotary phase converters are mostly all a bunch of crap anyway. Use one if absolutely NOother alternative is available. Most are all homemade because no manufacturer wants the responsibility and problems. Electricly there are tons of things that will work, but that does not mean they are worth much. These units work on the open delta theory with a center neutral tap. Hense when you read voltage A to N = 120, B to N = 120, and C to N = 197. Why? Because the electrcal path is through one complete winding for phase A and B, but through 1 1/2 windings when reading phase C. Hard to explain without a drawing. Like trig, just think of the triangle and you can find the answer to the universe!
If it runs and does not smell funny do not worry about it. Don't go looking for trouble because it will find YOU! And cost both time and money! :rolleyes:

lakeside53
12-06-2012, 12:15 PM
rdhem2.... he already stated in a subsequent post (see #4) that his current was way low on one leg, so he was looking for the problem.

There are many manf of RPC.. you see home built are because they can, and try to save money. And please fill in your profile location...

The Artful Bodger
12-06-2012, 02:10 PM
The Artful Bodger-I meant no disrespect that was an old off color movie comedy quote (Blazen Saddles) adapted to fit.

The gas compressor was a 15Hp. And I upped the capacitors accordingly.

Iíll get a few pics when I check voltage later probably around lunch.

No problem, I am cool!:D

jeremy13
12-06-2012, 02:39 PM
Ok guys and girls if there are any looking. This is what I came up with.
Rotary running and injection machine not running. Taken from motor leads.
L1-L2 242.2V
L1-L3 226.2V
L2-L3 228.6V
L1 33.9A
L2 28.1A
L3 33.1A
Rotary running and injection machine running. Taken from motor leads on injection machine.
L1-L2 237.8V
L1-L3 216.8V
L2-L3 214.3V
L1 26.7A
L2 28.2A
L3 8.8A

http://i1224.photobucket.com/albums/ee374/Jeremy_Hanak/6A643A33-379D-4CAB-BA56-49790CF751F2-1955-00000177A3BC303F.jpg

lakeside53
12-06-2012, 03:31 PM
For an RPC it isn't all that bad on voltage even when loaded. You could balance it a bit better but I doubt that would explain the large difference in the machine T3 lead. It is sufficiently out of balance that you need to derate your IM motor. At 5% unbalance (nema recommended limits) you would be at 75-78%. At your 10-15% it falls rapidly to about 60% of 15hp... So take care. If you are running this hp intermittantly, it's less of an issue, but still important, and it's why static converters suck for some applications, and poorly unbalanced rpc can be almost as bad.

This is the highly technical explanation...http://users.encs.concordia.ca/~pillay/12.pdf
Easier reading :
http://what-when-how.com/electric-motors/voltage-unbalance-electric-motor/
http://www.elongo.com/pdfs/voltages.PDF



Your high rpc idle currents are because the capacitance from t1-t2 is insufficent (I can't tell if you even have any) to to adjust the power factor closer to unity. That may also improve your overall performance.

I suggest you add bleed resistors on your start caps (for safety, not performance).

I can see your design varies a bit from this but the general principles apply http://www.practicalmachinist.com/FitchWConverter.pdf

The Artful Bodger
12-06-2012, 03:39 PM
First observation, a very neat build! One phase sure is not pulling its weight but I will have to ponder those figures before I can make a useful comment.

The Artful Bodger
12-06-2012, 06:04 PM
My contribution:-

Assuming the RPC idler motor is internally in good order, with no load and no capacitors connected to the output the voltages between each pass through leg and the generated leg should be equal but if they are not the culprit will be poor power factor on the supply side. (Please comment if you dont agree...:)

Therefore voltmeter(s) might be used to indicate power factor while correcting the supply side for poor power factor.

PF is effected by the load therefore tuning the PF of the supply side requires a load on the RPC. Unfortunately the load must be a non-reactive (i.e. neither capacitive nor inductive) resistive load which might be difficult to arrange at the power levels we are talking about here. One possible test load involves conductive bars hanging in a salt water solution. It may be possible to avoid the use of an electrical load for testing by putting a brake load on the RPC idler and adjust the supply side PF that way, monitoring the voltage(s) of the generated leg while doing so.

Once the supply side is tuned for PF at typical load the intended load can be connected and capacitors added to tune the load side for equal current.

Thats my theory anyway....:rolleyes:

J Tiers
12-06-2012, 10:32 PM
Actually, it's better to "tune" with the intended load in place. You need the unit to operate that load, after all..... I am speaking of the power factor on the LOAD side.

The low amps are not necessarily due to the load etc, but due to PHASE...... The amps are out of phase with the voltage on the manufactured load. The purpose of the "balance", "tuning", or "PF correction" capacitors is to correct the phase difference.

it is spoken of as "tuning out the inductive reactance", or as "providing a source of leading current", or any number of other descriptions which each have some validity.....

What it does is to actually cancel some of the series inductive reactance on the manufactured leg, which raises the voltage as well. Usually, you want that done for the full load condition, because at lighter load the driven motor is already "derated" by less load.

The pitfall is to over-correct (actually, "fully" correct) the PF and phase, which will lead to very high voltages at no load, which you do not want. So usually, you stop at 80 to 90% of the full correction.

I wouldn't worry too much about the supply side PF...... if you have single phase and need an RPC, you are almost certainly NOT being charged for your power factor, so it does not pay to spend money correcting it.

The Artful Bodger
12-06-2012, 11:29 PM
Actually, it's better to "tune" with the intended load in place. You need the unit to operate that load, after all..... I am speaking of the power factor on the LOAD side.

Ahem, I thinks that what I said in my last sentence..




The low amps are not necessarily due to the load etc, but due to PHASE...... The amps are out of phase with the voltage on the manufactured load. The purpose of the "balance", "tuning", or "PF correction" capacitors is to correct the phase difference.



it is spoken of as "tuning out the inductive reactance", or as "providing a source of leading current", or any number of other descriptions which each have some validity.....

That is true but there is a limit to what can be done with capacitors where the maximum phase change possible is 90 degrees which in practice will always be less. Thats why in another post I proposed a phase reversing transformer for 180 degrees phase change combined with capacitors to subtract from the 180 degrees to give something closer to the desired 120 degrees. (I have not heard of anyone doing that so maybe there is a hole in my theory.)




What it does is to actually cancel some of the series inductive reactance on the manufactured leg, which raises the voltage as well. Usually, you want that done for the full load condition, because at lighter load the driven motor is already "derated" by less load.

The pitfall is to over-correct (actually, "fully" correct) the PF and phase, which will lead to very high voltages at no load, which you do not want. So usually, you stop at 80 to 90% of the full correction.

High voltage at no load, the sort of thing to be expected in any circuit of high "Q".




I wouldn't worry too much about the supply side PF...... if you have single phase and need an RPC, you are almost certainly NOT being charged for your power factor, so it does not pay to spend money correcting it.

Does this not require a heavier idler motor to carry the higher input current? Also, I am thinking that home shoppers wanting several HP from a single phase circuit might not have much surplus current carrying capacity in their supply?

J Tiers
12-06-2012, 11:35 PM
That is true but there is a limit to what can be done with capacitors where the maximum phase change possible is 90 degrees which in practice will always be less. Thats why in another post I proposed a phase reversing transformer for 180 degrees phase change combined with capacitors to subtract from the 180 degrees to give something closer to the desired 120 degrees. (I have not heard of anyone doing that so maybe there is a hole in my theory.)


The 120 already comes from the motor characteristics, even with NO capacitors. It's only with LOAD that the PF goes awry and needs correcting. Then it is an "error" off of the 120 deg, , not something that needs to be "pulled" from 180 deg to 120deg.....




Does this not require a heavier idler motor to carry the higher input current? Also, I am thinking that home shoppers wanting several HP from a single phase circuit might not have much surplus current carrying capacity in their supply?

That's a perfectly valid point. If that is an issue, a PF correction can help. No argument.

killawarra
12-07-2012, 12:54 AM
Check that the barrel heater loads are evenly distributed over the three phases

The Artful Bodger
12-07-2012, 01:06 AM
The 120 already comes from the motor characteristics, even with NO capacitors. It's only with LOAD that the PF goes awry and needs correcting. Then it is an "error" off of the 120 deg, , not something that needs to be "pulled" from 180 deg to 120deg.....

Ah! Of course, that is true! Thanks.

jeremy13
12-07-2012, 02:56 PM
Ok I'm wading threw the wealth of information y'all have provided me. it's a bit over my head but I'm waiting on a friend of mine to help me. I'm more like the Chinese good at coping but don't understand what I'm doing. I'm starting to wonder if there's a bad winding in the injection machine motor. I can make the band heaters share L1-L3 and L2-L3 to bring up more load on the made up leg. and I don't have Inductance or by the diagram provider caps between the L1-L2 input legs.

J Tiers
12-07-2012, 08:04 PM
You probably do NOT WANT MORE load on the manufactured leg...... 25 vs 15 HP is not a lot of over-size on the RPC, especially for an injection screw motor, which is probably overloaded at some point in the cycle. Any extra loads just draw down the RPC output.

I'd stick them on the input directly, and let the ONLY load on the RPC be the motor. The heaters don't care about 3 phase, and it all is coming from the single phase input anyhow. Eliminate the middleman.

That lets the motor draw as much as it needs, which is likely to be at least full load, maybe more when the machine is packing the mold full and is creating max pressure.

Then balance teh manufactured leg at as near to full power as you can conveniently get.

Bad phases you can measure on the motor with an ohmmeter.... look for differences going from each wire to the next, and taking them in turn.