PDA

View Full Version : How do you hook-up panel meters to a rotary phase converter?



Fear
03-26-2019, 07:22 PM
I snagged some cheap Chinese volt/amp digital panel meters off eBay to put on the front of my RPC. They are intended for single phase with a ring amp probe to go around the load wire and two more leads to neutral and line. To check each leg of the converter for voltage do I hook to one leg and neutral? I didn't know how well neutral applies once it's created the false leg. One way or another I need to know how to read the power (voltage and amps) of each leg.

chipmaker4130
03-26-2019, 07:58 PM
. . .do I hook to one leg and neutral? . . .

You could, but a more relevant reading would be leg-to-leg. Are you putting in separate displays for each leg, or are you going to switch the connection manually and read one leg at a time?
It's really mostly gee-whiz info anyway once you get it set up. The 'third' leg is going to vary quite a bit with varying loads, and that is expected, normal and there's nothing practical you can do about it anyway!

Fear
03-26-2019, 08:13 PM
Are you putting in separate displays for each leg, or are you going to switch the connection manually and read one leg at a time?


Yeah, I bought three up to 300v and 200a meters. I was looking at page after page of pictures of home made RPCs and several had the meters mounted in their panel to check stats at a glance and I thought that seemed like a great idea.

BadDog
03-26-2019, 11:54 PM
Several years back when I built my 3 stage 5/10/15 HP RPC I included 3 inexpensive cabinet mount analog volt meters between each of the 3 phases so that for whatever idler level and load I was running I could see if the run-caps were keeping things reasonably balanced. Theoretically also see when/if things start go go wonky. Once I got everything all leveled out, it never showed me a problem, but added some piece of mind, and combined with the internally lit AB switches also added a bit of Frankenstein like cool factor (at least in my head).

J Tiers
03-27-2019, 12:27 AM
because these are "cheap chinese volt/amp meters", the first thing you need to know about your meters is:

Do they have an isolated power supply?

If so, what is the isolation voltage?

Panel meters may have an SMPS or transformer that isolates the measuring input from the power supply, OR, cheaper ones may have a DC or low voltage AC supply input that has the reference / ground side of the input common to the power supply input. If isolated, it needs to be good for the voltage you are measuring, plus a good bit of margin.

Obviously, if there is any chance of the power supplies being non-isolated, then you cannot power all of them from the same source, or there may be a "smokin surprise".

johansen
03-27-2019, 02:02 AM
if there are only two terminals, the meter runs off the voltage that is measured.

you may find two adjustable resistors on the circuit board, they are for calibrating volts and amps. the first one i bought was out by 5% on the voltage so i fixed it.

keep in mind that the current from the capacitors circulating through the motor does not exit the rotary phase converter. so you may want a 4th amp meter to display the amps flowing through the third winding of the idler motor, if you intend to push your converter to the limit.

J Tiers
03-27-2019, 10:04 AM
if there are only two terminals, the meter runs off the voltage that is measured.

......

Quite so.... I was not considering them as possibly being BOTH the measuring AND power connections. But for AC power measurements there would be no significant issue with doing that.

BadDog
03-28-2019, 05:43 PM
Mine were also as described, and readings were very comparable to what I got on my Fluke MM.

Fear
03-28-2019, 09:18 PM
They get power from the source they read. I say cheap, I just mean when I was pricing panel meters, many cost more than any of the Flukes I've bought. These are $15+/- each.
If you read between legs, then what are you reading? T1 and T3 are the original legs, T2 is the manufactured leg. If I'm reading between T2 & T3 then how am I reading the manufactured leg?

J Tiers
03-28-2019, 09:39 PM
They get power from the source they read. I say cheap, I just mean when I was pricing panel meters, many cost more than any of the Flukes I've bought. These are $15+/- each.
If you read between legs, then what are you reading? T1 and T3 are the original legs, T2 is the manufactured leg. If I'm reading between T2 & T3 then how am I reading the manufactured leg?

If T2 is the manufacture leg, and T3 one of the pass-thru legs, then you are reading one of the three "phase voltages" which are "delta" voltages. There is T1 to T2, T2 to T3, and T3 to T1.

The individual legs do not have meaningful voltages unless there is a three-phase neutral, in which case they have a voltage to neutral (wye voltage). Example would be 480V T1 to T2, and 277V from T1 or T2 to neutral. With 240V, the T2 to neutral voltage is irrelevant, because the wye voltage is non-standard and never used with 240V 3 phase. (with 208V 3 phase, the wye voltage is 120V)

plus, the RPC does not provide a 3 phase neutral.

Paul Alciatore
03-29-2019, 06:11 PM
True, the RPC does not normally provide a three phase neutral. But you could create one with three appropriately selected resistors. You would need to know the input impedance of your meters so you could calculate how much current they would draw while measuring the Voltage. This current would include that needed to power the meters as well as what was used for the measurement but the meter specs should include both in one figure.

Once you know the amount of current that the meters draw, you would want resistors that would draw 10 or 20 or even 100 times more current when they are between the three phases and a three phase neutral point (which they create). These line to neutral Voltages would be about 115/cos(30) or 133 Volts. I attempted to look up the specs on some similar meters, but they don't seem to provide the current draw on Amazon. So I have to assume a reasonable(?) figure and I choose 15mA. Yours could be more or less, I don't know. You could find out by hooking one up across an AC line and using a more sensitive current meter to see just what the current draw is.

Anyway, assuming 15mA for the meters, you would want a multiple of that for the resistor network: the reciporical of that multiplier would be the APPROXIMATE, expected error. So, 10X would give a 0.1X error or 10%. And 25X would give a 0.04X error or about 4%. I will calculate with that 10X or 10% error factor. So, 10 X 0.015A = 0.15A and that will be the desired current in the resistors.

I = E/R or R = E/I

E = 133 V
I = 0.15 A
And
R = 887 Ohms

The closest standard value would be 860 Ohms.

Now the power dissipated in the resistors is given by

P = E I
P = 133A x 0.15V
P = 19.95 Watts

I would suggest a 25 or 30 Watt resistor with 860 Ohms resistance. You need three of them. They are connected in a Y arrangement with the free end of each resistor connected to one of the three phases. The common point of the three resistors is your created, three phase neutral. And your three meters can use that as a common point from which they can measure the Voltages of the three phases. WITH a predicted maximum error of 10%. Since the circuit will be balanced, the actual error would be less than that, perhaps only a few percent, but there are no guarantees on any particular level of accuracy.

For greater precision in the measurements you can either use resistors with smaller resistance values (and higher power ratings) which will draw more current or get meters that draw less current.

PS: I do not think you will have any power/smoke problems with these meters since they use only two electrical connections and a magnetic (loop/transformer) probe for current.




If T2 is the manufacture leg, and T3 one of the pass-thru legs, then you are reading one of the three "phase voltages" which are "delta" voltages. There is T1 to T2, T2 to T3, and T3 to T1.

The individual legs do not have meaningful voltages unless there is a three-phase neutral, in which case they have a voltage to neutral (wye voltage). Example would be 480V T1 to T2, and 277V from T1 or T2 to neutral. With 240V, the T2 to neutral voltage is irrelevant, because the wye voltage is non-standard and never used with 240V 3 phase. (with 208V 3 phase, the wye voltage is 120V)

plus, the RPC does not provide a 3 phase neutral.

lakeside53
03-29-2019, 09:15 PM
But what is the point? Why are you concerned with voltages that reference neutral when neutral is not part of the final delivery (delta)?


And... great theory but the "neutral point" in practice isn't fixed - it moves depending on RPC balance and load.

Paul Alciatore
03-29-2019, 10:27 PM
The point? Well the OP wanted to be able to judge how well the system is working. Is this the way to do it? I don't know. I would think the current readings would tell you the most. The Voltages? I don't know. But I do think that reading the Voltages from a proper neutral point would tell you more than reading them to the center of the 230 Volt line used to power it. Those Voltages will, in theory, read 115, 115, and 190. Now suppose you get 120, 120, and 185. What does that tell you? Can you mentally calculate the actual vector Voltages? I sure can't.

As for the neutral point moving with circuit conditions, yes it will. But so will the point that a three resistor network will create. I can't think of a better reference point for a three phase circuit. Can you?

I am not arguing for creating a three phase neutral as opposed to just reading the line-to-line Voltages. That also would be easier to understand than using the power line neutral (ground). I don't care which he uses. I was just saying that there is an alternate way to do it. It can be done, if you want to.




But what is the point? Why are you concerned with voltages that reference neutral when neutral is not part of the final delivery (delta)?


And... great theory but the "neutral point" in practice isn't fixed - it moves depending on RPC balance and load.

J Tiers
03-29-2019, 11:47 PM
The argument for using the line-to-like voltages is that the motor uses them.

It may or may not have an internal neutral (most US motors do), but only has external connections for the three line-to-line voltages. So you may as well read those directly.

PStechPaul
03-29-2019, 11:50 PM
If they have LCD displays, the current draw should be fairly low, and will not change depending on what is displayed. But if they are LED, the current will be higher, and will vary depending on how many segments are lit. 111.1 would be lowest, and 88.8 would be highest. 60 watts seems a bit much to dissipate. Maybe you could use three lamps tied to a common point, but I don't know how well matched they might be.

You also might try adding three 1:1 transformers from each phase to GND. The outputs will be in phase with the L-N voltages even if there is no actual neutral, and current will be in phase with the line.

http://enginuitysystems.com/pix/electronics/3_Phase_2.png

Fear
03-31-2019, 12:52 AM
Reading up it makes sense just to get readings for delta voltages.

That's what AR lists specs and diagnostics for. I was just confused. I spent most of my life chasing problems in DC automotive circuits and am still trying to grasp household single phase and now conversion to three-phase. I'm used to voltage, current and resistance. To see voltage and current on the same gauge it seems funny that the current is accurate to the leg but the voltage is a difference between two legs. I'm used to the voltage being the easy part that just gets probed and amps being the hard part that (before we all had access to amp clamps) had to be taken in-line.

I am reading up on basic single and three phase, I swear.

J Tiers
03-31-2019, 01:09 AM
Well, it is a bit confusing until you wrap your head around it.

And, actually, you are correct in one way..... The actual "phase voltage" for delta is the voltage between wires. AND the "actual phase current" is the current between those same two wires.

The current in the wire (such as the generated leg) is actually the (vector) sum of the currents in two delta phases.... the two that connect to it.

But, in most systems, the actual delta current is not accessible, and in a way it does not matter, because the currents are given for the supply wires, and the wire sizes are selected based on current in them, even the power can be calculated from them (with power factor).

The only issue is that the specific wire current is not traceable to a particular load phase without more information.

It does not help that 3 wire delta is commonly referred to speaking about "phase currents" and meaning the wire currents (which are more properly "wye" currents). That is somewhat careless since current in a specific delta phase (circuit between supply wires) is usually not directly related to that wire current.

One can convert wye to delta or the reverse when calculating things, because one may make things simpler for a particular calculation. But you have to keep which you use straight and consistent, or all the answers will be wrong.

Point of that techie stuff is that while your meters will give numbers, they are actually reading a DELTA VOLTAGE and a WYE CURRENT. The numbers are useful enough taken by themselves. No problem, you can check balance etc. But those two numbers do not directly combine into a useful number for apparent power.