[Santa Fe Al]

Al: You must have missed a few pages.

CR.

Hi C.R.,

My wife says almost the same thing.

Lee,

Thanks for the info from you and C.R. Believe it or not, you guys have made me realize that I need to do a little more research before committing to buying stepper motors.

I still intend to use the new cheap Geckos, so will plan accordingly.

Thanks guys,

Al

[[e-]OneEye]

JustFx, I thought about that a few moments after I posted, then I got distracted. I haven't written HobbyCNC yet, but I can check in and see if he has more complete specifications for the motors.

So we're still looking for the perfect, highest-torque, motor for use with the Geckos. What's the highest torque drive out there running at 3-3.5A and with a 2.5mH winding inductance?

Keep in mind, the specs listed are probably for running it as a unipolar. All of HobbyCNC's drivers are unipolar. While the inductance might be the same running it as a parallel bipolar, the amps will definitely be greater - based on other motor specs I would guess around 4.2A.

From www.hobbycnc.com (no affiliation):

(NEMA 23)#23-305-DS8A 425oz-in bipolar rating, 305oz-unipolar rating,
4.2v, 3A, 200 S/R, 3.2mH, Size #23, Dual Shaft,
8 wire MOST power consumed, best with idle current reduction,
Faster speed over 2A version below. $49 each

At 3.2mH, SQRT(3.2)*32 gives 57.24V. Running them at 50V isn't a huge amount less than their max and might give close to max torque at higher speeds.

[wolfdagon]

I hope to have my router finished by the time the G540 is released and like everyone else I am wondering what the best motor to use with it will be.

In the mean time, I have three HP Laserjet motors that I plan on trying on my machine before actually purchasing any more.

This is all of the info that I can find on these motors, ,but it does not include inductance. Is there any way to calculate inductance using the known values? Or maybe someone already knows the inductance of this motor.

[Mariss Freimanis]

If you are asking about the drives, don't worry. They will have the highest possible performance for their ratings. I cannot design a drive where something is held back, where something is not as good as it can possibly be. They use the most careful design I have ever done and they will flat-out outperform any other drive in their performance spec range. Rated at 50VDC, designed for 60VDC and having an ultimate "come apart" voltage of 68VDC means there is plenty reserve. Rated at 3.5A and using 15A rated MOSFETs means more reserve. It is a very sturdy design, the core circuit is borrowed from the G203V after all. Add a 4-layer printed circuit board and 4 premium International Rectifier half-bridge MOSFET drivers and you have a drive that is very serious about performance.

It will carry our name and my reputation. Our name is known for good drives and the G250/G251 will live up to that.

Mariss

[wolfdagon]

If you are asking about the drives, don't worry. They will have the highest possible performance for their ratings.

Mariss

I don't doubt that a bit, Mariss. And I wasn't worried about the drives. Just trying to figure out the inductance of the motors I have, wondering what the maximum power supply voltage for them should be.

By the way, I just did a quick Google search and found these motors. They are 4 phase, 3.6V, 3.0A, 2.4mH. I don't know the quality of the motors, but that 2.4mH inductance gives a max power supply of 49.57V, which sounds perfect for these drives.

Kenith

[dansutula]

I understand that G540 will come populated with 4-G250 drives. I do agree that the 4-drive capacity of the G540 is probably the ideal or best number, to satisfy most applications.

But for folks who might need say 5 or 6 motors, will the G540 "frame" be available with less than a full compliment of drives. I realize that it may not be a huge cost savings, and it may be nice to have some built in spare drives, but it seems like it would be a bit wasteful to buy (2)G540 and have 3-drives sit un-used, if you need only 5 motors. Then again to some it might seem like a waste to have the second G540 "frame" populated with just one drive – but it is such a slick and convenient package.

Which raises the question or idea – could or has anyone come up with a modular stackable motherboard- (for lack of a better name) for the G250/251(s)?

And folks - please don't strain your brains on this question on my behalf - as I don't currently have a need for the 5-axis setup. In fact this question is merely a curiosity.

[Crevice Reamer]

By the way, I just did a quick Google search and found these motors. They are 4 phase, 3.6V, 3.0A, 2.4mH. I don't know the quality of the motors, but that 2.4mH inductance gives a max power supply of 49.57V, which sounds perfect for these drives.
Kenith

Those motors at first APPEAR to be perfect. But they are six wire motors. These motors can ONLY be wired in Unipolar or BIPOLAR Series--NOT in the more desirable Bipolar Parallel. The most you can get out of them is half the power at speed of a BP motor.

CR.

[ger21]

I understand that G540 will come populated with 4-G250 drives. I do agree that the 4-drive capacity of the G540 is probably the ideal or best number, to satisfy most applications.

But for folks who might need say 5 or 6 motors, will the G540 "frame" be available with less than a full compliment of drives.

Gecko has said that they will only be available in the 4 drive configuration. It would not be cost effective for them to offer multiple configurations.

[LeeWay]

I'd sell you two 540 frames with only 5 drives for the cost of two 540's all day long.

[Mariss Freimanis]

CR,

A 6-wire motor run half-winding with the drive set to the motor's unipolar rating gives nearly the same performance as an 8-wire motor wired in parallel. The difference is less than 2%. The inductance in half-winding is identical to an 8-wire parallel connection. Just be sure to insulate the 2 unused motor wires. They must not touch anything.

Mariss

[Mariss Freimanis]

dansutula,

The mechanical integrity of the G540 depends on all 4 drives being present. The drives screw mount inside to the bottom of the G540 case. The motherboard with the DB-9 and DB-25 connectors plugs onto and on top of the mounted drives. The cover of the G540 snaps over the DB connector flanges and to the case. This rigidly traps the motherboard in the aluminum package. If a drive is missing, the motherboard will flex and flop around in that area when a DB connector cable is mated to the G540.

All the sheetmetal has been delivered by the fabricators, the covers are out for silkscreening and the motherboards arrive from the pcb fabricators late next week. We have all the motherboard components and plenty of G250 drives to populate the G540s. It's coming together quickly.

Mariss

[automationtechinc]

Hi Marris,
I like to offer our 270 oz-in, 382 oz-in NEMA 23 motors to you to test with your new G540. I think these motors will work great with your new Driver. Check here for motors specification: http://www.kelinginc.net/NEMA23Motor.html. John

[Crevice Reamer]

CR,

A 6-wire motor run half-winding with the drive set to the motor's unipolar rating gives nearly the same performance as an 8-wire motor wired in parallel. The difference is less than 2%. The inductance in half-winding is identical to an 8-wire parallel connection. Just be sure to insulate the 2 unused motor wires. They must not touch anything.

Mariss

I respectfully call BS!

A Unipolar wired motor runs with only one coil at a time. Bipolar parallel powers TWO coils at once with the same inductance. BP should be twice as powerful.

CR.

[Mariss Freimanis]

CR,

Good objection! Good but wrong. Here's why:

Motor coil windings are bifilar. That means they are wound with 2 strands of magnet wire simultaneously. Ask yourself is there a difference in inductance between a coil wound with 100 turns of solid wire or a coil wound with 100 turns of 2-strand wire? The answer is "No difference". Or ask yourself does the inductance of a coil change with wire gage? The answer is "No".

Inductance only depends on the number of turns of wire, not the gage of the wire. Inductance goes up as a square of the turns of wire; if 100 turns gives you 1mH, 200 turns will give you 4mH.

Motor torque depends on ampere-turns only and is proportional to ampere-turns. If 1A through 100 turns of wire gets you 100 in-oz, it takes only 0.5A through 200 turns to get the same torque. Wire gage doesn't matter.

So let's get back to comparing a half-winding operated 6-wire motor to a parallel connected 8-wire motor. Let's say both motors are 1A rated, are 100 in-oz and have 5 Ohms resistance.

The parallel connected winding has two 5 Ohm coils in parallel. It measures out at 2.5 Ohms. The half-winding coil has the same number of turns but its resistance is 5 Ohms. Ampere-turns and inductance is identical. The only difference is a 2.5V drop at 1A for the 8-wire motor while the 6-wire motor has a 5V drop at 1A. At 50VDC, the 8-wire motor coil "sees" 47.5V while the 6-wire motor coil "sees" 45V. This makes no difference at low speeds.

At high speeds (past the motor's corner RPM), motor current drops rapidly. When it is at 0.5A, the 8-wire motor "sees" 48.75V while the 6-wire motor "sees" 47.5V. Bumping up your 6-wire motor supply voltage 1.25V (2%) cancels the performance difference.

Your own words: "A Unipolar wired motor runs with only one coil at a time. Bipolar parallel powers TWO coils at once with the same inductance. BP should be twice as powerful." Both have the same inductance as you say and only inductance and supply voltage determines power output. Your mistake is there are not 2 coils, just one coil with fatter wire (two strands instead of one).

Finally, you can run an experiment. Disconnect the paralleled windings on your 8-wire motor. Run it with only 4 of the 8-wires connected. See if you can find any difference in performance (greater than 2% of course).

Mariss

[Crevice Reamer]

I really find this hard to believe. Why then are the majority of step motors capable of Bipolar wiring? It would seem that Unipolar is the weak sister of motors.

I can't get AT my shrink tubed motor wires to try that test. How about YOU fire up the (Obviously better test equipment) dynomometer and post the results?

CR.

[Mariss Freimanis]

I did. Years ago. The results are in our "Step motor basics.pdf" on our commercial website. I'm not trying to be difficult but there is no point in me revisiting an investigation whose results were unambiguous. Particularly so when the physics behind the experimental results are in complete agreement. Nothing will have changed over the years.

I suggested you try it only because I'm not the one in need of convincing.:-) I thought the joy of you discovering something new on your own would be pleasure enough. Nature showing you what you thought was wrong and her showing you how it is. I hope your curiosity gets the better of you and you can get past the 4 shrink insulated wire obstacle. Maybe it will help if you say "Mariss is full of it. I'll prove him wrong."

Either way, knowledge is what you test for yourself and find it wrong or right. It gives enormous power against experts like me or others. If it's true, you can test it and say "I agree". If it's wrong, you can say "Bogus. It didn't pass."

Mariss

[High Seas]

This is a repeat from Post 1100 ( a few days back) Thought I'd repeat as long as were talking about 540s

With all the effort into the G240/250, I hope I don't ask these out of turn:
SIZE:
Haven't found a reference to this in the over 1000 posts (maybe missed it) but it looks like it is about 7 x 2.5 x 1.25 inches. [From the posted photos- mockups] Close?
AVAILABILITY:
Certainly after the G240s -- but in the next 30-60 days?
COST:
Will there be a Limited Time Offer like the G240/250s? Ballpark costs ?

Maybe too soon to tell - but the sheet metal has been cut so, maybe a size answer might be forthcoming?

I am interested in the G540 - planning a case and need some availability/cost/ and size details.

[Mariss Freimanis]

High Seas,

You are very close. See the attached picture. The panel dimensions of the G540 are 6.375" wide and 2.375" high. The panel cutout is 5.625" wide and 2.375" high. The depth from the mounting surface to back is 1".

The attached picture is the artwork sent in to the screeners yesterday. Any comments or questions are appreciated.

The G540 will be available in 2 weeks barring a disaster. The price will be $299 and no quantity discounts until we have a handle on what support volume will be. Should it be ordinary, discounts will be applied.

There will be no time limited special sale on the G540. Drives we know and knowing them we can run an introductory sales. A complete system we don't know because we never designed one before. We don't know what we are getting into.

Besides, at $75 / axis it's probably less than your motors cost. For that, you hook-up your motors, connect a few limit switches, a power supply and plug in your cables. No need to puzzle-out a $100 breakout board, no need to cut and strip a hundred wires to hook everything up, no need to stress over having screwed up miss-connecting one of those 100 wires. No need to come up with heatsinks and messy thermal grease and fans that have to be oriented just so. You don't need to worry about grounding issues and circulating ground currents.

In effect, the G540 turns many hours of work into minutes and guarantees it will work right from the get-go. Your mission in life is not to learn how to be a competent electronic technician. Your mission is to just get a machine running and do things with it.

Here is what you get:

1) A 4-axis 3.5A, 50VDC system ported to Mach3.
2) Optoisolation on every parallel port I/O line.
3) Short-circuit protection.
4) 4 general purpose outputs rated at 1A and 50VDC max. Pullup loads.
5) 4 general purpose inputs requiring a SPST switch to GND to activate. Filtered.
6) Mach3 charge-pump watchdog timer.
7) 4 G250 drives. Autostandby, silent operation, 10 microstep, midband compensated, morphing, etc. etc.
8) 4 female DB-9 motor connectors. Motor side cable connector carries current set resistor.
9) Female DB-25 parallel port connector.
10) FAULT and POWER LED indicators. You miswire something, you light the FAULT LED.
11) Supply Voltage connection. Apply your 15VDC to 50VDC power supply here.
12) DISABLE input. E-stop input. Kills everything while activated. Clears FAULT when released.

What you have to provide:

1) Male DB-9 motor cable connectors. Wire your motors to them.
2) Current set resistors. Solder these to your motor cable connector. Pin 5 to Pin 9.
3) DB-25 parallel port cable.
4) +5VDC from your PC.
5) Motor power supply. +15VDC to 50VDC. Your motor choice determines the Amps.
6) That's it. It's that easy.

Mariss

[yngndrw]

Sorry to beat a dead horse, but getting back to the unipolar Vs bipolar:

If the bipolar motor has double the guage, can it not be run at a higher current for the same amount of heating ? (Assuming that you don't saturate the core ?)

[Mariss Freimanis]

It's not a dead horse, it's a very interesting and alive horse.:-)

Your question answers itself. Yes. Look at any 8-wire motor datasheet. You will find 3 current ratings for it; series, unipolar and parallel. Assuming a 1A series rating, the unipolar one will be 1.41 Amps and the parallel rating will be 2A. The relationship will always be Unipolar = SQRT 2 the series rating, parallel = 2 times the series rating. Check it out. All motor manufactures use the same formula.

Motor heating has two principal components; copper losses and iron losses. Copper losses are the easiest to understand. They are I^2 *R losses anytime current passes through a resistance. The series -> unipolar -> parallel current ratings. Run the numbers and you will get the same Watts for 2R -> R -> R/2 respectively.

Motor manufactures are not drive manufacturers. The ratings naively assume their motors will be run with crap-hole L/R drives. The kind you see on the internet advertising "10A, 100VDC drive for $0.99!" What they leave out is the need for two 1,000 Watt current limit resistors.

Most modern drives are switching type drives. Obnoxious, noisy choppers for cheap ones, silent PWM for ours. Either kind incur iron losses. Iron losses are primarily eddy current and to a lesser extent hysteresis losses. Eddy currents are caused by a changing magnetic field in iron laminations. Switching type (high performance) drives regulate current in a winding by oscillating between an upper and a lower threshold. Usually at about 20kHz where only dogs can hear them. This oscillating current causes the iron magnetic dipoles to swing in one direction, then another one 20,000 times a second. This causes a "friction" in the iron manifested as heat. In fact it induces an electric current passing through the resistance of the iron but the analogy works better for me.

Motor manufactures make no accounting for these losses. These losses far outweigh I^2 R at high supply voltages and they are what prompted us to come up wit the 32 times the SQRT mH rule. Iron losses cause the majority of motor heating at high supply voltages. The limit is the motor reaching 85C.

Mariss