Current per phase ?

[Microheli]

Hi all:

I´m new to this , and while waiting to receive the wires, stators and controller have some questions:
1.- Which is the expected ratio for each phase current to the battery Amps?

2.- Which is the maximum current density ( Amp/ mm2 ) normally used in this motors???

Thanks
JAlex.

[Q]

WOW, pretty difficult questions here. :? :wink:

Considering the fact the current is always devided over 2 phases, I think you know the answer to your first question.

I'll leave question 2 to somebody else as I don't have a clue what to look for.

[LBMiller5]

I started this post about 1 hour ago, and just as I was about 2 minutes from finishing it, there was a 5 second power outage that re-booted my computer. I have a UPS at home, but not at work! Go Figure!

Anyhow, I figured I would type this up in Notepad off line, and save it every couple minutes to avoid the same thing from happening again. When I am done I can cut and paste the text into the reply window.

In regards to how much current a wire can hold. First I believe that it would be benificial to explain a little bit about wire, and the standards used to determine wire size.

The amount of current a wire can carry is directly proportional to the cross-sectional area of the wire. There are 2 standard units of measurement for Cross-Sectional area in wire. In the metric system it is mm^2, in English units it is Circular Mils. Since all my Electrical Engineering experience is in Circular Mils, I will use those units, and give a brief explanation as to their origin.

Almost all of the wire used today is less than 1" in diameter, and more specifically, the wire we use to build our motors is around .010" to .025" in diameter. For all of you who remember your 6th grade math, to find the area of a circle you take the radius squared and multiply that by Pi (3.14159). Well, when you take a wire that is only 0.010" in diameter, the radius, which is half of the diameter, is 0.005". When you square this number you get .005 x .005 = .000025. Multiplying this number by Pi you get .000025 x 3.14159 = .00007854 square inches. These are very nasty numbers to have to work with, and the math gets real messy real fast! (Even for us Engineering Geek types!)

To make the numbers and the math real simple, wire engineers developed a system that is much easier to use. This measurement system is called Circular Mils. Instead of using inches to measure wire diameter, they use mils. 1 mil equals 1/1000 of an inch, so a wire that measures .010" is also 10 mils. Then to avoid having to take the radius of the wire, square that number, and then multiply by Pi, they came up with a new method. They simply take the diameter of the wire, square that, and give the result in a unit called Circular Mils that takes into account all the decimals, and Pi and all the nasty math.

Looking back at our earlier example of a wire with a diameter of .010", we came up with a cross-sectional area of .00007854 square inches. (Yuck!) Using the Circular Mil units, the same wire is 10 mils x 10 mils = 100 Circular Mils. Now isn't that a much nicer number!

Besides, when you compare wire sizes, and divide one number by the other, the Pi's and all those decimal places cancel out anyway and you are left with a simple number.

Now you all have a little more understanding about Wire sizes and Circular Mils, so now lets get to the matter at hand and see how much current we are drawing through our motors.

The sizes of the wire used in building our motors are based on the American Wire Gauge standard. Each gauge size has a specified diameter and is published in charts. Here is a link to a website I found that has a nice Wire Gauge Chart.

http://www.calfinewire.com/services/wirech.htm

Most of the wire used to build our motors is 24ga, 26ga, or 28ga. Using data from the chart here are some numbers.

24ga wire measures .02010 inches in diameter or approximately 20 mils.

26ga wire measures .01594 inches in diameter or approximately 16 mils.

28ga wire measures .01264 inches in diameter or approximately 13 mils.

For a point of reference, 12ga household wire measures .0808 inches in diameter or approximately 81 mils.


Using these numbers, the cross-sectional areas of the wires are as follows:

24ga wire equals 20x20 or 400 Circular Mils,

26ga wire equals 16x16 or 256 Circular Mils,

28ga wire equals 13x13 or 169 Circular Mils,

and our 12ga wire would equal 81x81 or 6561 Circular Mils.


It is standard engineering practice to require a minimum of 300 Circular Mils of wire per Amp of current. This allows for many wires to be bundled together and not have any of them overheat or have too high a voltage drop across the wire. For any of you out there that have ever done any house wiring, you know that a 12ga wire is rated for 20 Amps of current.

If we take the above numbers for the 12ga wire and divide 6561 Circular Mils by 20 Amps, we get 328 Circular Mils per Amp, which as you can see, meets the minimum of 300 Circular Mils per amp engineering guideline.

In the case of our CD-ROM motors, we push a LOT more current through the wire. We can do this for several reasons:

First, the wires we use are very short. We normally have 5-6 inches of wire between our speed controller and our motor, and the actual windings in the motor are pretty short as well, somewhere around 2-3 feet per phase is common in the CD-ROM motors.

Second, the surface area is pretty small, and we do not need to worry so much about heat build up.

And third, there is a great big cooling fan (AKA Propeller) blowing a lot of air across the motor to keep it cool.

Let's look at a sample motor to get some real numbers.If you look in the GB Performance Database, motor #42 is a good candidate to look at. This is a Delta Wound motor with 21 turns of 26ga wire per pole. Looking at the numbers for the 10x6 prop, this motor pulls 12.6 amps of current at 11 volts.

Here is a link to this motor's data:

http://www.gobrushless.com/testing/motor_test_details.php?Motor_ID=42&email=&password=&sort_dir=ASC&old_order_by=&ORDER_BY=2

Our speed controllers are 3 phase units that alternate through the phases with 2 phases on at any given time. With a Delta wind motor, all 3 phases are always receiving current, but one phase has twice as much current as the other two at any point in time. In the case of this motor, with 12.6 amps of current flowing into the motor, one phase will receive 8.4 amps, and the other 2 phases will receive 4.2 amps. Since the current is constantly changing through the phases as the motor spins, we can calculate the average current by adding the current in each of the 3 phases and divide the result by 3. In this case 8.4 + 4.2 + 4.2 = 16.8. Dividing 16.8 by 3 equals 5.6 Amps of average current per phase.

Going back to the recommended maximum current for a given wire size, the 26ga wire this motor is wound with has an area of 256 Circular mils. Going with the minimum recommended 300 Circular Mils per amp, this wire is safely rated for 256/300 or .85 amps. We are pushing 5.6 amps through it in the above motor, which is 6.56 times the rated maximum! But, the motor runs fine because of the reasons listed above.

This motor is running with only 45 Circular Mils of wire per amp, and it will get warm when running. Looking at the data for this motor, the power in is 138.6 watts, and the calculated power out is 92.2 watts. The difference is 46.4 watts. This power has to go somewhere, so let's see where.

Our Li-PO batteries have an internal resistance of about .030 ohms per cell. For a 3 cell battery pack, this is a total of .09 ohms, and if we round this up to .1 ohm, it will take the battery wire into account as well. The loss in the battery equals the current squared times the resistance. In this case 12.6 x 12.6 x .1 =15.87 watts. The resistance of our typical brushless speed controllers is less than 1/100 of an ohm. In the case of a Castle Creations Phoenix 25, it is .0065 ohms. The loss through the controller would be 12.6 x 12.6 x .0065 or 1.03 watts. Adding this to the battery loss gives 16.90 watts leaving the remaining 29.5 watts to be dissipated in the motor. Did you ever grab the end of a 30 watt soldering iron? Now our motors won't get that hot because there is plenty of air blowing across them. But if you were to replace the prop with a flywheel pulling the same load, the motor would burn up in a few minutes or less.

This leads us full circle back to MicroHeli's original questions:

1. What is the expected ratio for each phase current to the battery amps?

The answer is for phase current = input current divided by 3, times 1/sin(120)^2 which in plain numbers equal the input current times .4444.

2. "What is the maximum current density normally used in these motors?"

It varies from motor to motor ranging from about 30 Circular Mils per amp for really hot racing motors to over 100 Circular Mils per amp for more sedate winds, with most of our CD-ROM motors in the 45-70 range.

I have tried to keep this fairly simple, And I am sure I lost a few people with the math, but that is the story based on my understanding.

Any Comments?

Lucien

[spoke2570]

That was beautiful.
Thanks!

pete

[Microheli]

Thanks for the answers. I understand now why so thin wire is running at 6,8, 12 Amps without burning! :D
Also circular Mils terminology is new me. Although i can handle inches and mm. we use Amps/mm^2 on this side of the world! I´ll have to rethink it.

Next question now: understand it is not possible to meassure those values on each phase using a regular DC meter. What instrument type are you using, if any of you do.???

Thanks
JAlex

[Pipistrelli]

Lucien,

Our speed controllers are 3 phase units that alternate through the phases with 2 phases on at any given time. With a Delta wind motor, all 3 phases are always receiving current, but one phase has twice as much current as the other two at any point in time. In the case of this motor, with 12.6 amps of current flowing into the motor, one phase will receive 8.4 amps, and the other 2 phases will receive 4.2 amps. Since the current is constantly changing through the phases as the motor spins, we can calculate the average current by adding the current in each of the 3 phases and divide the result by 3. In this case 8.4 + 4.2 + 4.2 = 16.8. Dividing 16.8 by 3 equals 5.6 Amps of average current per phase.

Can you clarify the point you made about the motor input is 12.6 amps, but when you do your 'average current' calculation you add up the current of all the phases and end up with 16.8amps? Is this correct, or should it be 12.6/4 = 3.15amps per phase, but with 6.3amps on one phase?

[juki48]

this reminds me too much of college, which reminds me I need to go back and finish :roll: I do hate college. to me, working is much better.

[LBMiller5]

The reason they don't add mathematically up is beacuse they are out of phase with each other by 120 degrees. If you look at 1 phase by itself, with no reference to the other phases, it carries more current that you would think, namely Current in divided by 3. The values given were DC current with no regard to phase differences.

Our brushless DC motors are a bit of an enigma. We think of them as DC motors because we power them with a DC Battery. In truth, after the conversion of the DC current through a brushless speed controller, our motors are technically 3-phase AC motors. The only real difference between a CD-ROM motor and a 3-phase AC motor is that our little motors never experience voltage swings from positive to negative. They do receive "alternating current" however, which technically makes them AC motors.

Getting back to your original question, to do a full evaluation of the current in the motor I would have to introduce the mathematical term "i" (Lower case letter I), which is the square root of -1. I know you can't take the square root of a negative number, but engineers use this notation to express what they call the "Imaginary" portion of current that comes into play when you discuss multi-phase motors. Because the voltage and current are not in phase with each other in multi phase motors, a portion of the current into the motor does no work. In a 3 phase motor this portion of the current is equal to the cosine of 120 degrees squared which is cos(120)^2. This equals -.5 x -.5 which is 0.25.

Now the value of what you would think the current would be, is 12.6 divided by 3, or 4.2 amps. To correct this for the fact that the voltage and current are 120 degrees out of phase, you take 4.2 and divide it by (1-0.25) or 4.2/0.75, and you get the 5.6 amps that I showed in the earlier math.

Now mind you this assumes no other losses in the motor, which of course there are, but theoretically this is what you get.

I hope this helped to clarify things a little bit, and if others are out there that really want to know the nitty gritty details, I could possibly put together an in-depth article on 3-phase motor theory, but frankly it would be a very dry read!

If anyone else out there would like to chime in their opinions, I would be happy to discuss them with you.

Keep those props spinning!

Lucien

[swatson144]

The reason they don't add mathematically up is beacuse they are out of phase with each other by 120 degrees. If you look at 1 phase by itself, with no reference to the other phases, it carries more current that you would think, namely Current in divided by 3. The values given were DC current with no regard to phase differences.

Chop

Now the value of what you would think the current would be, is 12.6 divided by 3, or 4.2 amps. To correct this for the fact that the voltage and current are 120 degrees out of phase, you take 4.2 and divide it by (1-0.25) or 4.2/0.75, and you get the 5.6 amps that I showed in the earlier math.

Chop

If anyone else out there would like to chime in their opinions, I would be happy to discuss them with you.

Keep those props spinning!

Lucien

Chime, could you break this down in black and white? Not all of us can take a snapshot of the current of all 3 phases at a single moment in time. So assuming a fellow had a snap on ammeter (ac) of good accuracy and took a reading on 1 phase, on a proper motor. How would this individual determine the overall current passing through the motor? I see that this could be used to determine if indeed the motor was wound correctly vice 18t 19t 17t motor. I assume that one could also figure the current from the battery etc.

how would this be affected by a "true RMS" meter

I really enjoy your posts as I have lots of experience on the deckplates, though we had no concern of why the services were installed as they were nor what the overall current of the motor was, but more of is the motor running right.

Thanks.

[LBMiller5]

Almost all of the clamp-on style AC current meters are designed to read pure sine wave alternating current at exactly 60 Hz. (Or for you chaps on the other side of the pond, 50hz). The readings that you get from them are not accurate for any other condition.

The power going to out motors, after it leaves the brushless speed controller is, for the most part, a square wave with a frequency of somewhere between 7,000 and 50,000 hertz. The frequency varies with controller brands. For example, in the Castle Creation line, the Phoenix 10 switches at 12,500 Hz, the Phoenix 25 switches at 11,000 Hz, and the Phoenix 35-80 controllers are user selectable to 11,000, 22,000 or 41,000 Hz. Because of this, common Clamp-on AC current meters are pretty much useless for measuring the current in our motors.

There is a very easy way to measure the current in our motors which I will explain here. Most of the speed controllers that we use on our CD-ROM motors are in the 10-25 amps range. The on resistance of these controllers is extremely low, in the range of 1/100th of an ohm. Because of this fact, very little power is lost in the controller. For most applications, this amounts to 1-2% of the total power used. Since so little power is lost in the controller, if we measure the voltage and current going into the controller, we have a very close measurement of what is going into the motor.

Most of us have a basic digital DC Multimeter that measures Volts, Ohms and Amps. You can buy them at Harbor Freight or similar supplier for under $10.00. You can get a very good one these days for under $50.00. Here is a picture of one I picked up at Harbor Freight on a sale special for $4.99:

http://innov8tivedesigns.com/rcgroups/meter.jpg

This one reads DC Volts, AC Volts, Ohms, DC Current, tests diodes, and can read the current gain of both PNP and NPN transistors. The standard current scales range from 200 micro-Amps up to 200 milli-Amps, and it has a seperate current scale that goes up to 10 Amps. I have checked this unit side by side with a $500.00 lab grade voltmeter with different voltage inputs, and resistance measurements, and found the meter to be accurate to less than 1% on all scales. Most of the time it was accurate to less than 1/10 of 1%. Not bad for 5 bucks!

A word of warning! To use the 10A scale on most of these digital meters, you must unplug the red lead and move it to another input hole. If you look in the photo above, the 10 Amp hole is directly above the current position of the red lead. Failure to move the lead before applying more than 200 mAmps of current will blow the internal fuse and possibly damage the meter.

Some of these meters do not have a 10 Amp scale, and sometimes we have a motor that pulls 12 or 15 amps. So how do we use our meter to measure these motors? Simple, we use a current shunt.

So what exactly is a current shunt? A current shunt is a device of a known low resistance, typically 1/100th of an ohm or 1/1000th of an ohm. It will drop a known, very small, voltage across itself that is directly proportional to the amount of current flowing through the shunt. They are rated in millivolts per amp, and the output is measured with a voltmeter.

All of our digital meters have a 200mV scale, and this is the one you use with a current shunt. If you have a current shunt rated at 1mV per Amp hooked up between your battery and your speed controller, and you measure the voltage across the shunt, the reading on the display will be the amps of current you are drawing. 1 amp of current will be read as 1 millivolt. 6.5 amps of current will be read as 6.5 millivolts and so on. The beauty of the current shunt is that you get an accurate measurement of the current without adversely affecting the circuit.

If you want more accuracy, you can use a current shunt that gives 10mV per amp. On these you have to divide the meter reading by 10 to get the proper current. For example, if you used a 10mV per Amp shunt on a motor that was drawing 7.85 amps, the meter would read 78.5. You simply move the decimal point over 1 space to the left and convert your answer from 78.5 millivolts to 7.85 amps.

It is very easy to build your own current shunt. One of the properties of copper wire is that it has a known resistance per foot, and this value can be looked up in a standard table. Solid copper 10ga wire, the type used in house wiring, has a resistance of exactly .001 ohms per foot. Because of this, a 1 foot piece of 10ga wire makes a perfect 1mV per Amp current shunt!

To make the shunt you need a set of connectors just like the ones you use to connect your Li-Po battery to the speed controller, a 12" piece of 10Ga wire, and a 2" piece of 10ga wire. If you use a 2 pin connector, like a Deans, then you need both pieces of wire. If you use 2 seperate mini-banana plug type connectors, you only need the 12" piece of wire. You can go to Home Depot or Lowes and buy 10ga wire by the foot. It only costs about 14 or 15 cents per foot, and it comes in Red, White, Black and Green. Make sure you get the solid wire, since the stranded wire has a different resistance value. Once you have a couple connectors and some wire, you can begin construction.

Step 1. Take 1 piece of 10ga wire that is exactly 12" long and wrap it around a broom handle or similar sized tube to form a coil. Leave about 1 inch of wire at the beginning and end of the coil so it is shaped like this: _O_

Step 2. Strip about 1/4" of insulation from the wire at each end of the coil and solder a male connector at one end, and a female connector at the other end of the coil. (Make sure you keep the Positive and Negative pins straight and don't mix them up.) Do not put any heat shrink over these pins, since you need to be able to hook your meter to the ends of the wire.

Step 3. If you use the single pin connectors, you are done. If you use a 2-pin connector, solder the 2" piece of wire between the other pins of the 2 connectors. You may have to flex the coil to get the proper spacing for the short wire. Put heat shrink sleeve on these connections so you do not accidently short your battery when you hook the meter up.

Step 4. Install the shunt by plugging it in between your battery and speed controller.

Step 5. Connect your volt meter positive lead to the end of the coil closest to the battery, and hook the negative lead to the end of the coil closest to the speed controller. This works best if you have test leads with little alligator clips at the end. This way the voltmeter stays put, and you still have both hands free.

Step 6. Now you are ready to measure current! Set your meter to the 200mV scale and turn it on. Now as you apply throttle, the current draw will show up in the meter display. If your meter shows 3.4, then you are drawing 3.4 amps. If it reads 10.3, then you are drawing 10.3 amps. If it reads 34.5, then you better shut it down and blow on the motor because it is getting really hot! :P (Ha-Ha)

So that is how to build a current shunt. If you only build small motors that use less than 10 amps of current, and you want a more accurate shunt, you can build a 10mV per amp shunt in the same way, only for this one you need a different wire coil. 18ga solid wire has a resistance of .0066 ohms per foot, so a piece 18" long measures exactly .01 Ohms and makes a perfect 10mV per Amp current shunt. To build one of these, follow the instructions listed above, except make your coil from an 18" piece of solid 18ga wire. This size is commonly used for underground sprinkler system wiring or doorbell transformer wiring, so you can get 18ga solid wire at your favorite home improvement store as well.

Now you all know how to measure the current going into your motors. If you have 2 meters, you can connect the second meter across the input to the speed controller and measure the voltage in at the same time. To calculate the power into the motor, just multiply the reading on the voltmeter by the reading on the current shunt meter and you have Watts of power in. For example if the current meter reads 7.9 and the volt meter reads 10.8, then the power in would be 10.8 x 7.9 = 85.32 watts.

Hopefully that will explain an easy and accurate way to measure our motor current.

Any more questions?

Lucien

[paulvi]

You are my God.... Ok well at least a very wise man...

I have been looking for plain english and real world application theory..


Thank you and now I do not need to go buy that $50 watt meter.

using the 10mV scaled wire what would be the recomended max current through this wire.


I plan to build a dual range black box with bannana plugs on both ends and clips vor VDC and AMPS i can put a switch on the input to select 1 or 10mv outputs, put this all in a Chicken Shack Box and TaDa...

Thanks again you are proving to be a killer asset to GB.. and this forum

[LBMiller5]

Paulvi,

Thanks for the compliment! I am glad the information was useful to you.

To answer your question about the max current through 18ga wire, here is the math.

As I stated earlier in the thread, the max current a wire can carry is proportional to the cross sectional area of the wire. From the wire gauge charts, 18ga wire has a diameter of .0403" or 40.3 mils. The Circular Mil area for this wire is 40.3 x 40.3 = 1624. Using the accepted 300 Circular Mil per Amp standard, the maximum current would be 1624/300=5.41 Amps.

Now remember, this is a very conservative rating used for house wiring and Mil Spec cable harnesses. You can actually put more current through the wire safely, about double in fact. I would never put more than 10 amps through the 18ga solid wire, since that is about the upper limit for that size.

A quick bit of math to show heat loss: Assuming you are running 10 Amps of current throught your 18" piece of 18ga wire in the 10mV per Amp current shunt described earlier, the known resistance of the piece of wire is .01 ohms. The voltage drop across the shunt equals the current times the resistance. In this case it would be 10 Amps x .01 Ohms = 0.1 Volts.

The heat loss in the wire is equal to the voltage times the current. In this case it would be 0.1 Volts x 10 Amps = 1 Watt. 1 Watt of power is not very much heat. To get a perspective, the average night light bulb is 4 watts. If you grab one of these bulbs that has been on for a while, it is noticeably warm, but not so hot as to burn your skin, probably in the 50-60 degree celsius range. The current shunt in the above example would be producing 1/4 as much heat as the common night light bulb, so heat build up would not be a problem. The wire might get a little warm, but it will not melt the insulation, or burst into flame, or anything else nasty like that.

If you plan on building a switchable dual shunt, use the 10mV per amp shunt on motors up to 10 amps, and use the larger 1mv per Amp shunt on motors over 10 amps.

Now before you get a chance to ask me, "What is the maximum rated current of the 10ga wire shunt", here is the answer.

10ga wire has a standard houshold rating of 30 Amps AC. The wire measures .1019" in diameter or 101.9 mils. This gives a Circular Mil area of 101.9 x 101.9 = 10,383.6. To get the max safe current divide 10,383.6 by 300 and you get 34.61 amps.

You can see that the 30 Amp rating for houshold use is a little on the conservative side, and in real life, with a 1 foot piece of wire, you could safely run 50-60 amps of current for short periods of time with no damage to the wire or insulation. The truth of the matter is that you could not put big enough wire in a CD-ROM motor to get any where close to a 50 amp current draw, so you don't have to worry about the 10ga shunt. You would burn up your motor and speed controller before you hurt the shunt!

So there you go, hope that answers your questions.

Till next time!

Lucien

[paulvi]

Ok so i feel silly..

That is what happens when you have to many people at your door (work) and you get distracted from your hobby reading.

You actually stated every thing I needed to figure this out. But I failed to reread before asking..


Thank you very much for your time on this..

I just about have every thing built out and am looking at about $15.00 in parts and that is to use terminal lug bannana connections and 20 amp switches.. But that is a far cry from the $50 for a LHS device. Of cource I could buy a mV digitall display and get it up a bit closer HAHAHA..

On a diffrent note. (Or the same Topic) how does voltage play in all of this I know that switches are rated at 120 or 220 vac and I seem to remember that when dealing with lower voltages it up rates the device and higher voltages derate it. I assume it has to do with contact resistance and wattage that the switch can handle

[LBMiller5]

The switch rating you need to need to be concerned with is the amperage rating. The amperage rating is determined by the internal contact size of the switch, and has very little to do with voltage. The maximum voltage rating of a switch is determined by how far apart the contacts are when the switch is in the off position. This rating is to make sure that the the voltage does not arc across the contacts.

If you take a standard room light switch that is in the on position, and slowly turn it off, you will hit a point where the light starts to dim and you hear a crackling sound like bacon frying come from the switch. At this point the electricity is arcing across the contacts and will quickly heat up the switch and burn up the contacts. At this point the contacts are not far enough apart to break the circuit. When you flip the switch all the way to the off position, the contacts are far enough apart to prevent the arcing.

Since a motor shunt will see somewhere between 8-12 volts from the battery, the voltage rating of the switch is of little concern. Do make sure that the switch is capable of handling the maximum current you expect to use and you will be fine.

Lucien

[paulvi]

This is the build from the info in this thread. I may be a little off on my reading I will litnus test it agains a buddies watt meter..

But it works one set of terminals is for curent the other voltage.

I droped the switch for now. I am using 10amp speed controllers so thsi will work for just all I own

[paulvi]

did some testing and we found that the 18g wire if mesured at the 18 and 1/8 inch mark gave best results compared to a medusa watt meter.

the 10g wire needs to be just a tad longer as well.

But all things considered this workes great. now all i need is a home made tac..

[LBMiller5]

The extra wire needed is due to the fact that you probably soldered your test lead socket wires about 1/16" from the actual "end" of the wire, making the shunt 11-7/8" long in effect. This would explain why you would need a little bit more wire to be accurate.

On the topic of accuracy, I know that some of you guys (and possibly gals) out there have access to a clamp-on style DC amp meter. My local hobby shop carries one that is a AC & DC voltmeter, Ohm meter, and AC and DC clamp-on Amp Meter. The only problem with some of these is that they have very high Amp scales. For example, the one my LHS carries only has 100 Amp and 200 Amp DC current scales. It is tough to be accurate when you are trying to read 4.5 Amps on a meter with a 100 Amp scale. As engineers we always try to shoot for something in the mid-range of the instrument to get accurate measurements, so how do we get a 100Amp meter to accurately measure 4 or 5 amps?

You you a current amplifying loop. Here is how it works. The amp meter reads the current going through the wire that is placed within the inside of the the meter's measuring coil. If you have 1 wire carrying 5 amps of current, the meter will read 5 amps. However, if you take 2 wires, with each wire carring 5 amps, with the current flowing in the same direction, and put both of them inside the meter coil, The meter will read 10 amps! The meter adds up the total of the current flowing through it.

To help visualize this, imagine that you have a piece of stranded 10ga wire that has 7 strands inside the insulation. This is actually 7 seperate wires running in parallel with each other. If you put 7 amps of current through this wire, each strand will carry 1 amp of current, the the total is 7 amps. If you place your current meter loop around this wire it will read 7 amps, but what it is actually reading is 7 wires with each wire carrying 1 amp, but the meter does not know any better and it thinks that there is just 1 wire carrying 7 amps.

We can use this feature to our advantage when measuring small amounts of current. Back to our original question: How do you read a 5 amp current accurately on a 100 amp meter? Answer: If you take a piece of wire large enough to handle the motor current, and wrap it around the coil of the amp meter 10 times, it will read 10 times the current in the wire. In the case described above, the meter would read 50 amps.

Here is a photo to illustrate this. I don't have a clamp-on current meter, so I used a roll of electrical tape to simulate the amp meter coil, and wrapped wire around it. You will notice that there are 10 wires on the inside of the loop, but only 9 wires on the outside. The meter only "sees" the wires that are inside the coil, so remember that when you wind the wire.

http://dmiller1036.homestead.com/files/L/10turn.jpg

Now you can simply take the meter reading, move the decimal point over 1 space to the left, and you have a reading that is 10 time more accurate than before. This reading is also in the middle of the meter scale which provides additional accruacy. This is especially useful if you have an older meter that has an analog pointer scale instead of a digital readout. On a meter like that, the width of the needle on the 100Amp scale is equal to about 1/2 of an amp. It is much easier to tell the difference between 50 and 51 amps than it is to see the difference between 5.0 and 5.1 amps!

So if any of you motor builders out there have access to a clamp-on current meter, now you have another great way to use it!

Later,

Lucien

[swatson144]

I have a hard time believing that. I don't have a clamp on dc meter to test with.The main reason I don't is the accuracy is generaly like 3% of full range which would leave it near useless. If it is really this simple I'll have one in a couple of days. .3% I could live with. I really am from MO. and someone'll have to show me...Well just saying you tested and it really works would be good enough. Simple test like put a battery on charge and loop a wire around a couple times and compare it to the single wire current reading. Seams more like the flux the thing is reading would be canceled by the loops outside. Very possible I am wrong and actually hope I am.

[paulvi]

The extra wire needed is due to the fact that you probably soldered your test lead socket wires about 1/16" from the actual "end" of the wire, making the shunt 11-7/8" long in effect. This would explain why you would need a little bit more wire to be accurate.


on the above pic i would say you are right

On the 18 g wire we took a lengt of it 20 inches long striped off 2.5 inches from each end of insulation cliped the meter lead to one side and made a slid of the other meter lead set up a constant lamp load and a motor cycle battery for a stable power source (ok slow to discharge power source) and and slide the wire back and forth wile we compaired both meters I assume there may have been some resistance in the clip lead connections but once soldered we found that with 18 and 1/8th inches were were closer to the comertial power meter..

Now it is possible that meter is off, That would be scary,..

I still thing small variations in resitance can play havoc. If i had nothing to compare my meter with I would have only been off by .2 to .3 of an amp

close enough

[paulvi]

I have a hard time believing that. I don't have a clamp on dc meter to test with.The main reason I don't is the accuracy is generaly like 3% of full range which would leave it near useless. If it is really this simple I'll have one in a couple of days. .3% I could live with. I really am from MO. and someone'll have to show me...Well just saying you tested and it really works would be good enough. Simple test like put a battery on charge and loop a wire around a couple times and compare it to the single wire current reading. Seams more like the flux the thing is reading would be canceled by the loops outside. Very possible I am wrong and actually hope I am.

He is right I have used this method to amplify curent readings on low current systems that we were using, 100 turns this was on a system that used computeres to monitor current being applied to the soil. and we were using adam modules to convert the amplified signal to RS485 to be xmited back to the control PC


but it does work.

I would be happy to show you a test but the clamp on I have left is an AC unit only All the DC stuff was owned by the company I was working for at the time.

[LBMiller5]

I know for a fact that this method works, because in my years as an Electronics Technician I used to do it all the time to measure small currents. This was back in the 1980-1985 time frame when most of the test equipment we had in the lab was still analog with meter displays, and you got the most accurate readings when you were between 50% and 90% of full scale.

You can use any number of turns as long as you divide your displayed value by the number of turns. If you have a motor that draws about 4.2 amps, and you have a 100 amp meter, you can put 20 turns of wire around the meter coil. In this case you would read 84 amps of current, and when you divide 84 by 20, you would get your 4.2 amp current draw.

Hope that helps to "Show You". :wink:

Lucien

[paulvi]

LBMiller5

I am going to retest every thing I want to make sure that I have not over looked a small step. also I will be soldering my test leads to the shunt to see if the 18 inch rull plays out.. (eliminate stray resistance)

This method works fantasitic and I doubt I will ever buy a amp meter again for the hobbie stuff..

Thanks again

Any thoughts on how to count the spining blades of a prop 2, 3 or 4 blades kind of a poor mans tach

[LBMiller5]

The best way to read the prop speed is to use an optical tachometer. Globee makes a real nice one that will read 2, 3 or 4 blade props up to 32,000 RPM. It sells for $39.99 in the Tower Hobbies Catalog, and it runs on a normal 9 volt alkaline battery.

When you say "Poor Man's Tach" I assume you mean something real inexpensive. There are a lot of different way to measure the speed of the motor, Oscilloscopes, Frequency counters, Strobe Lights and the like. With any of these you can use a very inexpensive sensor of 1 type or another, but then you need a several hundred dollar piece of test equipment to read the sensor!

In the long run, the cheapest way to read a prop is to buy a Tach!

Sorry about that!

Lucien

As a side note, (And this is directed at all of the people who post photos on this site) Please, Please, Please re-size the photos so they are no more than 600 pixels wide before you add them to a post. The browser window in the system automatically adjusts the width of the screen to accomodate the wider pictures, even if they end up going off screen to do it. The problem with this is that the text window also widens up to the same size as the largest photo posted in the thread. When this happens, the text ends up being wider than the monitor screen and the readers have to keep scrolling from side to side to read the post. (Like this one has now done.)

My digital camera takes photos at 2048 x 1536 pixels. When displayed full size, these pictures take up 4 full monitor screens! (Assuming your monitor is set to 1024 x 768 pixel resolution like mine is.) I always run my photos through Adobe Photo Deluxe and re-size them down to 533 x 400 pixels before I upload them to the site. At this resolution they are still very clear and visible, but they stay within the standard width of the screen so the text box does not get messed up.

I am sure many people were unaware of how posting large pictures messes up the view of the thread, and that is why I am saying it here. I am not casting any blame to anyone in particular, just making a comment that will help to improve the readability of the postings.

Thanks for your help in keeping the posts readable!!

Lucien

[paulvi]

I supose i could use a photo detector and my freq counter But you right $39 is hard to beat.. for a nice portable unit..

Pictures fixed.. I am not to fond of the wide screen look myself HAHAHA I will They download faster too when you mak em smaller..

[ScubaSteve]

Hmm Lucien if you wanna do a write-up on 3-phase brushless motor theory, I'd gladly host the article for you. :wink:

Your bit on max current density... Jay, maybe your 10 amps/mm2 safety limit should be reconsidered. The source of that info probably didn't take into account our specific application and the high level of cooling our coils get. :?:

[LBMiller5]

Thanks Paul

Much better now!

[Jay C]

Not untill I see proof otherwise ... is there a quick way to conver circular mils to mm^2 ;) ?

Jay

[LBMiller5]

This thread has gotten into some pretty heavy motor theory, and there has been 2 different conventions brought up as to current carrying capacity for wires. A lot of people have used Amps per square millimeter as the measure of current, while I introduced the units of Circular Mils per Amp as another common convention.

There have been a lot of "Rules of Thumb" passed around as well as safe current guidelines. Typical household wiring in the US goes by the minimum of 300 Circular Mils of wire per Amp of current for safe use. There have been other people shouting "Don't exceed 10 Amps per square millimeter" as another safety guideline.

I have gone ahead and put together 3 spreadsheets that compare the 2 different conventions with wire sizes and currents typically used in our size motors. I am making them available on my personal web server in 2 different formats. For those of you that have Excel, and know how to use it, I have the .xls files available. You can download these and play with them, change numbers, and do as you like.

For those of you that do not want to mess with the numbers, I have saved the spreadsheets as HTML code that will show up like a regular Web page. Clicking on those links will pop the chart up on the screen, where you can view it, or print a copy if you would like.

I discovered some interesting things when I looked at the numbers. 300 Circular Mils per amp is equal to about 6.5 Amps per MM^2. Going the other way 10 Amps per MM^2 equals about 200 Circular Mills per Amp.

Earlier in the post, I ran some numbers and showed mathmatically how the DC current in each phase was equal to the total current in x .4444. If we return to Motor #42 in the performance database, it was drawing 12.6 amps of total current, and had a current of 5.6 amps per phase. This motor was wound with 26 ga wire, which has an area of 252 circular mils, or 0.128 MM^2. With these numbers, the motor in using 45 Circular Mils per amp, which is equal to 43.75 amps per MM^2!

Well that shoots the old 10 Amps per MM^2 rule right out the window. (Right along side the old 300 Circular Mils per Amp guideline!) Now I know that I have just opened up a whole big can of worms on this one, but I am sure it will keep this thread active! :wink:


Here are the links to the spreadsheets I spoke of earlier. The first batch of 3 are the actual Excel .xls files for those of you who want to play with the numbers. These will open up in a seperate window. You can do a "Save As" and put them where you want on your computer.

Chart #1 Wire Data
http://dmiller1036.homestead.com/files/L/Wire_Chart_1.xls

Chart #2 Wire Currents
http://dmiller1036.homestead.com/files/L/Wire_Chart_2.xls

Chart #3 Conversion Table Between Circular Mils per Amp and Amps per Square Millimeter
http://dmiller1036.homestead.com/files/L/Wire_Chart_3.xls



Here are the links to the Web Page versions of the charts. They will also open up in a seperate window and can be viewed or printed.

Chart #1 Wire Data
http://dmiller1036.homestead.com/files/L/Wire_Chart_1.htm

Chart #2 Wire Currents
http://dmiller1036.homestead.com/files/L/Wire_Chart_2.htm

Chart #3 Conversion Table Between Circular Mils per Amp and Amps per Square Millimeter
http://dmiller1036.homestead.com/files/L/Wire_Chart_3.htm


That should keep you all busy for a little while at least.! :D

Have fun crunching numbers!

Lucien

[paulvi]

Built the 2nd shunt much cleaner this time much smaller too also used the 18g soild core wire. 18 inches did it. being off just a hair and using the clip leads on the test set up crept some errors into the readings.. so with my meter having a 300mv scale this will safly read 29 amps I would assume that the 18g wire would start heating up if it was anything but burst readings. Which seems to be most of what we do anyway

Again thanks for the info and the idea's

Pictures of the 2nd one if anyone is intrested..

[LBMiller5]

Looks Good Paul,

You packaging looks very nice. I am going to build a similar one to that for myself to do my motor testing. I will hook up 2 meters at the same time so I can read both voltage and current through the motor. I am also going to build a little 3-phase unit that can go between the motor and the speed controller to measure the current in all 3 legs at the same time. I will probably put a 3-position switch on the outputs so I can read all 3 phases with just 1 meter.

I am assuming you have a 3 cell Li-Po pack there and an automotive 1156 Tail Light bulb as a load. With 17.8 showing in the meter display, you must be pulling 1.78 amps. That would be just about right for an 1156 bulb, which has a hot resistance of about 6 ohms, and is rated at 25 watts @ 12 volts in.

These things really do work!

Lucien

[paulvi]

It is a 2s2p pack but that is right on the bulb leftovers from the 1963 impala my son sold this year.

I have it wired so I can hook up a 2nd volt meter. I also put a dean on thet line so that the bannana plugs do not short out whan I am only using the amp meter.

here is a pic of the control panle I designed for a 440 volt 3 phase soil heater box on the left is the computer interface box on the right is the ssr and the current feed back system. The box was connected to the site office by fiber RS485 and monitored 255 temp probs and the 6 25 foot electrodes it took a month but we got the soil up to and maitained 250 deg's

We were steaming fuel for apache heli's out of the ground at the end of the runway on fort hood texas. I had a blast with that project the whole thing was connected by dial up to berkley so that we could plot the soil temp and monitor current draw.. we went through 3 xformers before we finally stoped smoking them... also lost one ssr to the blue smoke gods.. that was a frantic call from the site forman OMG you should see the smoke comming out of your boxes. HAHAHAHA the system had shut down and already dialed us up to alert us that it failed. HAHAHA best part of going and installing the system was getting to watch the trainees learn to hover heli's (6 hours a day of wop wop wop...)
Notice the size of the heat sink on the SSR

[LBMiller5]

Steve,

I could put together a "sort-of in depth" article or series of articles on Brushless motors. Sort of a "Brushless CD-ROM Motors for Dummies" type of booklet. We could go over Speed Controller theory and operation, 3-Phase motor design theory and practical applications, Current and wire size considerations, and a host of other parameters. I could generate some color CAD drawings illustrating some of the principles of the motors, and how, for example, an ABCABCABC wind interacts with 12 magnets versus how an AaABbBCcC wind interacts with 8 or 10 magnets.

Perhaps we should start a new thread to get input from the rest of the Go-Brushless Forum. We could ask the members to post the questions that they would like to have answered, and based on that, write an outline for the article.

Let me know what you think about this.

Lucien

[ScubaSteve]

Lucien, I think it's a magnificent idea. :P

[LBMiller5]

So How do you want to get this thing started?

Lucien

[Q]

So How do you want to get this thing started?

Lucien
By just another thread I'd say