After countless reading through this and other sites...
 

I still don't understand how to get these numbers correct. If someone could help me with a personalized response how to get the correct numbers from my selection.

3 x KL23H256-21-8B
2.4 Inductance ( Bipolar Parallel)
3 Amps Current Per Phase ( Bipolar Parallel)

After getting those numbers, how do I decide on drivers? Could I use 3 x kl-4030?

On the KL-4030 data sheet is says: "Suitable for a wide range of stepping motors of Nema 17 and 23". The motor that you've specified is a large Nema 34 motor.

The calculations that we normally use to figure maximum voltage was published by Geckodrive for Gecko Stepper Drivers. Those figures may or may not work with stepper drivers from other manufacturers. The Geckodrive G201x or the G203v would work very well with the motors that you selected. You could use a power supply up to 50VDC, with 35VDC being my suggestion. You might be able to use the G540. It falls within the specifications, marginally. You would probably be better off wiring the motors half-coil if you decided to us the G540.

Someone else will have to advise you about using Keling Stepper Drivers. I've had no experience with them.

They would be a tight fit, as you would put it to its maximum. you need around 45V for 2.4mH and 3Amp.
Better option if you are on a budget would be a KL-6050 or KL- 5042, and you might consider Gecko G540 which is good for 3.5Amp and 50Volts and has a BOB integrated.

What I failed to see on first looking at the datasheet is that is an small Nema23 motor with only 1.3Nm , that would not be sufficient to drive a machine such is the Mechmate

Thanks Danilo for catching my oversight. I looked at the KL23 Nema 23 motor and my mind played tricks on me (again). I thought that I was looking at a Nema 34 motor.

A size 23 motor can be used, if you use the right gearbox or belt-drive. It would be fairly easy to match pulleys to get a 6:1 or a 7.2:1 ratio. Doing that would easily multiply the motor's torque so that the motor could handle the job.

Notice that I used the words, "can" and "could" carefully. Using a long enough belt so that enough teeth were always engaged means that the size 23 motor would be more bulky than a size 34 motor with belt drive. A long belt also means that there would more elasticity in the belt, which just another way of saying "backlash".

My choice would still be a size 34 motor connected through a 1:3 to 1:4 belt-drive. That design has been tried and proven. It works. The costs are reasonable.

I just want to check out my calculs :
I've selected these motors : 5.5A 4mH (SY85STH80-5504B)
- 32*√4 = 64V
- 5.5A*4 = 22 ; 22*67% = 14.74A ; 14.74x64 = 950VA (943.36)

So I should get a 64V 1000VA power supply. Am I right ?
It sounds so huge to me (hard to find and pretty expensive) through, steppers does not look extremely powerfull.
I should have done a mistake somewhere...

http://www.antekinc.com/pdf/PS-10N63R.pdf


Here it's about 180.00 1000 watts @ 63 volts with 12volt and 5 volt taps too.

Axel, I ended up with your predicament when I 1st tried to work out the transformer size. That is what we were exposed to. BUT there is another way to determine the transformer size.

Logically, similar machine with similar mechanism, weight & duty should require similar power (watt) to operate. since VA is a very close approximation of watt. AFAIK, 300VA tranny is the accepted " standard "transformer size which most MM operated on without excessive hot transformer.
Disregard of the voltages of the transformer, 300VA transformer should do the job for you. If you like you can go for 400 or 500VA just to be on the safe side. but using a 1000VA seems like overly paranoid...

1000VA transformers need a big inrush current when they switch on. This gives problems to quite a few people.

Be carefull with this calc, you need 64 Volt AFTER the capacitors!
the transformer voltage should be √3 times smaller than 64V

I use a 17amp 34V transformer for two years with no problems, standard EI core not toroid

Thanks you all for your wise advices. And for the link Sean ;)
I was listing all the power supply and motors specs in the 'MM already cuting section' :o
That saves me a lot of time.

I'll go from a 300VA to 500VA depending on oportunities.

Just a fiew last questions on these subjects sirs,
I see some kit configurations with several "in box" power supplies.
I mean : 2x 200W 48VDC 4.2A (for 4x1200oz.in steppers)
or 4x 350w 60vdc 5.8A (for 4x1600oz.in steppers)

How to calculate a multiple power supply configuration ? A simple division ?
For example 300VA needed = (4x75VA) power suplies ?

As far the "in box" power supply's output voltage is alowed by the stepper driver,
(for example from 18 to 50vdc) is there a preferency to choose 48v instead of 24v or 36v ?
I apreciate the Intensity might be superior with 24v and inferior with 48v,
but is it right to think it will be pretty the same ?

1200 oz.in or 1600 oz.in steppers seem awfully big for this application. That could be why your power supply calculation ended up so high. Check the motor selection threads :)

You get me wrong Brad, that was just examples picked up into existing retail kits. ;)
I've read a lot the Motors selection tread.
It's 4.6 Nm motors, with a 3:1 belt reduction, they should be fine for the work, not overpowered.

SY85STH80-5504B motors are just right at 650 oz.in

Sorry, I missed that Axel. :) As long as you aren't going into unreasonable levels, E.G. >32x inductance, then higher voltages may give you better acceleration than lower, so 48v would be preferred to 24v.

I get it, thank you Brad ;)

VA = AMPS * (VDC * (1 / (SQRT(2)))

So, 15 A * 65VDC * ( 1 / (SQRT(2)) = 15 * (65 * (0.7071)) = 750 VA

The AC volts rating is used, not the DC volts rating. The large filtering capacitors remove most (95%) of the AC ripple on a DC power supply. If you use 15,000 uF to 20,000 uf capacitors (several capacitors in parallel, if needed), a 750 VA power supply and a 25A 400V bridge rectifier, you'll have just what you need for heavy duty work.

The bridge rectifier specified costs about the same as smaller units, which is why I don't use smaller units.

75 volt to 100 volt "snap caps" in the 6,000 uf to 10,000 uf range are plentiful and fairly inexpensive compared to the large "can" types that most people use. Look at some of the photos to see how builders have connected "snap caps" together in parallel with nylon wire-ties and heavy gauge connecting wire.

You need caps whose voltage is rated at 150% of the AC value of the transformer because AC is measured using RMS and DC is measured at peak values. DC is SQRT(2) X AC, or 1.414 times higher than the the RMS AC value. That's why you need caps whose voltage is rated at 150% of the AC value of the transformer.

Many thanks Mike, it's all I needed. ;)

Am I correct in my assumption that Geralds recommendation in the openingspost is still valid (with my specs-see below), and that the reasoning behind this added 20% is that the calculation before the added 20 % will give you the unregulated DC power output, and this voltage will drop when under load ( which can be as much as 20 % ) hence the recommendation ?

My calc then gives me: V4.1x32= 64.79V + 20 % = 77.75 V = 77 volt DC output ( just within driver max 80 V )

Bipolar parallel current = Unipolar current X 1.414
5x5.65=28.28x0.67=18.94x64,79=1227.61 VA/Watt
which could be halved ( read this on the forum ) because all motors will not be asking full power simultansouly ?
1227.61 VA/Watt - 50 % = 613.805 = 600 VA ?

(5xMotor DEITECH 86HS9801 - driver LEADSHINE AM882 bipolar parrallel)

Sounds safe/no risk of magic smoke apparent ? Thnx in advance, bit scared to make stupid mistakes on this, this is my last check before ordering.

why 5x motor?

Hey Ken,

I get all exited that someone answers me and confirms/denies my electrical assumptions (that's really what they are).... so I can order the last stuff.... turns out you posted only a question :(....so do have any answers to my question before I answer yours ? :D

Just kidding; My plan for world domination requires 5 motors instead of the usual 4 because of a 4th axis or double Z

Your quest for world domination still would not require all 5 motors running at the same tim... so you can take it easier with your PSU VA rating. IMHO, 4x5.65A would be rational.

**PS,
Btw, I use 9801 too, it works wonderfully with 3A direct drive 24 teeth 1/8step, I bet you will use the belt reduction, hence I assume you won't be needing anything more than 3A per motor to dominate the world :)
with that, you can have even smaller PSU....
Its not about money or being cheapskate on parts, its about appropriate...

Hi ken, thanks for all the info. Good to have first hand experiences. Yes, I am using 4:1 reduction. Are you driving the motors at the higher (+20%) voltage !?

nope, used 32*sqrt(mH) =56.34V, custom wound a 40V tranny & get stable 54V dc (loaded). no rocket science involved.... :)

ok, so ignore the +20% recommendation of Gerald it is, to be on the safe side.
I just thought he said that because it's the difference between the regulated and unregulatedpowersupplies under load, and was curious about that assumption being correct.

But...... your calcution seems to be faulty, or am I making a mistake ? 86HS9801 motor Deitech ? V4.1*32=64.79 V ?

The plus 20% comes from my personal experience - We have increased the voltage of our 4 tables to about 20% over the conservative Mariss formula value, to give really good cutting performance. If there is a heat problem, it is really simple to control the heat with the current setting.

Sorry, yours is the correct numbers, I made a mistake in my coil inductance.
Tend to forget numbers...
I can still remember I didn't go there because my motor driver is 80V max, , a big fluctuation in my line voltage & I risk burning all the drivers. I wasn't brave enough...
Also, 80V rated capacitors are a lot cheaper than a 100V one. again, one won't use 80V DC on a 80V rated capacitor unless one crave for fireworks inside the control panel.

Hey Gerald,

Ok so there's goes my theory, but (your) experience counts... Since 77 V is very close to the max 80 of my drivers, I will follow your example but a bit conservative... and probably make it a nice round number in the lower 70-ties volts area.

@ Ken Thanks ! I am so out of my comfort zone with these electronics I already got afraid that I was the one with a stupid beginner mistake again :D

No wories, I came from the same dark corner too. :o
My example may not be the right one but its mine :D

End of the day, learnt from those before you & apply common sense.

PS, I don't use Gecko...

- me neither (LEADSHINE AM882).

- Ah the Darkside...yes...and Gerard is our Yoda:D,

I must say sometimes I really hate(d) it* that the info/theory is scattered over the forum/internet, or stuff 'seems' contradicting, but even so.... I feel Gerald was right with this approach. I've already learned sooooooo much, it was not easy, and very time consuming.....but when it is finished it will be my machine, and I will know every inch of it. If I want to repair/change/ improve on it in the future I can.

* Often I have only a 2-3 hours I can spend on it in a week, an hour or less a day. It becomes then very hard to make significant progress, which frustrates, because when you start to understand it, you have to put it away again, and the next time you look at it, you read something else and you are left second guessing yourself and losing a lot of time catching up again....

But I am (slowly) getting there and my questions directly related to this thread are answered...so I will continue in my build thread as to not pollute this thread.

There are several "unknowns" in the use of the 86HS9801 motors used with a Leadshine stepper driver. Geckodrive published the formula for use with Geckodrive stepper drivers. No one should assume that the same formula is safe to use with other stepper drivers. It would be prudent to contact Leadshine and get their recommendation.

If the 86HS9801 motor motor is used at full Amps with a Geckodrive stepper driver and if it is loaded to the point that it will miss steps if pushed any harder, it will be generating about 80-degress C (yes, I have tested that on my testbench when I used a 86HS9801 stepper with a G203v stepper driver). You will have to read the data sheet for that stepper motor to see what temperature the insulation can handle, but 80-degress C is hotter than I have chosen to run my motors. Reducing the load also reduces the heat generated - all other things being equal.

Reducing the current that is pulled by a stepper also reduces the torque of that motor. Reducing the torque means that the motor will not be able to cut as deep as one might hope.

In a general sense, voltage determines the speed of a motor under a light load. Current determines the load that the motor is able to handle. However, both voltage and current are interrelated. In a resistive load, voltage X current = watts. The higher the voltage or the greater the amount of current being drawn by the stepper motor, the higher the heat. Many modern stepper drivers have an "automatic" current reduction feature that reduces the current by 50% or more whenever the motor is idle for more than 1 or 2 seconds. That reduces the heat. In a 3-D cut, the motor may be required to make short moves at short intervals, never coming to a complete halt for the time required to enter "reduced current mode". In that situation, the heat will build up faster than expected.

I use an infared thermometer to measure the heat produced by the stepper, the stepper driver and the load. For my peace of mind, I keep the stepper's temperature at 60-degrees C or lower and the temperature of the G203v at 40-degrees C or lower. Because I use Geckodrive stepper drivers, I use Mariss's formula to determine the maximum voltage, then I choose a voltage at or less than that voltage.

The 86HS9801 motor has Rated Current (A) = 4.0 and Phase Resistance (Ohms)= 0.98. The voltage marked on the motor would thus be 4.0 x 0.98 = 3.92 Volt.

Most websites and producers (other than Geckodrive) would still tell you to aim for 20 to 25 times marked motor voltage. That is 78.4 to 98 volts for the 86HS9801 motor.

Geckodrive will tell you 32xSQRT(4.1) = 64.8 Volts

I suggest 1.2x Geckodrive = 38xSQRT(4.1) = 77 Volts

Mike Richards has mentioned before 0.75x Geckodrive = 24xSQRT(4.1) = 48.6 Volts

The choice is yours! :)

Fox, nobody said learning is gonna be easy, but its not impossible. Shriving through this treasure trove is the time best spent on the internet for me.

Don't forget voltage rating of other component parts in the system.

PS, careful with your current(A) selection, bigger A require fatter cable which = more money, more weight. mind you, cable cost is significant investment.

Ken, you do a lot of texture cutting, where your Z-motor is moving very slowly and almost continuously . . . . . . . does that motor get much hotter than the X or Y motors? Our experience is that the Z is cooler than the X or Y motor. The motor with the biggest cutter load gets the hottest.

Modern drives seem to have completely cured the "hottest condition at low speed/hold" which was apparently the plague for earlier (over 10 years ago) drive technology. Mike Richards, when saying "In a 3-D cut, the motor may be required to make short moves at short intervals, never coming to a complete halt for the time required to enter "reduced current mode". In that situation, the heat will build up faster than expected." appears to be referring to another era of drives that don't apply here anymore. Same applies to his logic on heating caused by selected motor voltages.

Our experience is that running a motor hard, fast, deep causes the most heating today. The earlier era experience was that the most heat came when the motors sat still, doing nothing.

I don't know what was the history of motor theory, my understanding of electric machine operation is based on electromagnetism which I nearly forgot after I left school.

All my motors are hottest when not in motion. Heat is produce from the coil resistance.
I configured full current when stop.
In static mode,
P static=IR & I=3A, R=coil resistance (Constant)

In dynamic mode, all electomachanical machines the current fluctuates according to load, the theory is way to tedious to get into, in layman terms, its like pushing a shopping trolley, you need more force to get it up to speed, & when in motion, you need less force to keep on the speed, if you run into a ramp, you need to apply more force to keep climb at the same speed & if the trolley is full its harder to push than an empty trolley...

In the stepper motor case, its identical, when it need to generate more force, it will ask for (draw) more current from the power source, its not greedy, it only ask for as much current it need to do the work..., . but wait, the driver has a max current setting, hence the motor can never draw full current it asked for all the time & it also won't be asking for the full current all the time, so on average, the current that passes through the motor coils will be lower than when it is stationary. I can lay out pages of equations & maths to proof this but I'll leave it for noobs who has the energy & interest to do the chore :)
Unfortunately, I can't get away without an equation,
P motion=IR, I= less than 3A, R=Coil resistance (constant)

hence P static > P motion

My Z-motor is the coolest of all motors. I reckon like you said because its lightly loaded. Especially, when it has the highest mechanical ratio of all 3 axis. (Ball Screw)

Actually, the temp difference between the X & Y motors are 2~3C, I consider they are at the same temperature, but Z-motor runs 5~10C lower, unless my IR temp gauge lied to me.

Conclusion, Cold motor is one that isn't working hard, hard working motors gets hot... Ain't that familiar :)

Our aim is to make sure motor don't get too lazy, yet not slave drive them to a burn out... :)

*More on operating V & I.~
We have a choice of using higher voltage or heigher current to achieve the same power.
P=VI
In transmission application, higher V is prefered, because P=IR, R=wire R, & wire R is proportional to wire length.
In short distance, higher I is prefered, because low V is safer to work with (still dangerous nevertheless.)
My personal preference is always go for lower I possible, as fire is almost always caused by high current. & if select your component's Voltage rating carefully, high V isn't that dangerous (still dangerous nevertheless).

Thanks Ken, you confirm our experience; in 3D cutting, the slowest moving, lightly loaded motor is the coolest.

It is surprising that you have chosen "full current when stop". Most of us opt for the current reduction feature.

In 1/2 current, I can move the gantry with 1 hand push. in full current, I need 2 hands to move the gantry... Hence, full current. Its not what others do, but I like it that way, give me peace in mind...

Ken,

What motor do you use on your Z-axis and what is the gear ratio on your Z-axis? You wrote that you're using a ball-screw. Two popular screws from Automation4Less have pitches of 5mm per revolution and 10mm per revolution.

A 7.2:1 geared motor with a 1.25" pinion has a "pitch" of 13.8mm per revolution. A motor with a 10:1 gearbox and a 1.25" pinion would move an axis the same distance per revolution as a ball-screw that moves 10mm per revolution. A motor with a 13:1 gearbox and a 1.25" pinion would move an axis the same distance as a ball-screw with a 5mm pitch.

In any case, when attached to a ball-screw, a motor would only have to produce a maximum of 70 oz*in of torque to match the 700 oz*in of torque produced by a PK296A2A-SG7.2 geared motor. Even a little PK264 motor could easily handle that much torque.

What I'm guessing is that your machine with a ball-screw on the Z-axis would only draw a fraction of the current that would be required if you used rack and pinion on the Z-axis. If it only pulls a fraction of the current, it will only produce a fraction of the heat.

Exactly, that what I call lightly loaded. or in another layman term very much over sized motor. I'm using 86HS9801. 640oz-in. I think the motor work harder overcoming the gas spring than the actual work :) I really can't remember what the pitch was 5 or 10... its chinese & I got it off ebay, stuff from Automation4Less & Oriental motor are too dear for me.

I didn't do much math as that was not important since I have too much power for Z-axis, I only do maths when this are marginal but its obvious that the overall mechanical advantage of ball screw is a few time over a direct drive. move 5~10mm per turn, vs 24mm (Pitch circumfrance of Mod 1 is the number of teeth in mm) one turn... I leave the details for you to ponder.

Mike, we have rack&pinion Z drives and always the Z is the coolest motor. Identical 7.2 geared motors on all axes.

Ken,

Thanks for the quick response. As you said, your Z-axis is lightly loaded so it won't draw much current.

If I've done the math correctly, a 24mm pitch diameter gear would move an axis about 75mm per revolution or about 7.5X futher than a 10mm pitch ball-screw, so if your motor were used with that 24mm pitch diameter spur gear, it would have to pull about 7.5X more current than it pulls with your ball-screw to move the same load. Of course there are other factors involved.

In 3-D work, the cutter is normally lightly loaded simply because much of the fine detail work requires a small cutter and a small cutter could not handle a heavy load without breaking. However, in 3-D work all three axes are constantly moving, usually short moves. Depending on the way the software is setup, short moves may never get "up to speed" which means that those moves are in the high-torque range of the motors. High-torque moves mean that the motor is pulling more current than it would pull in a low-torque situation. Depending on the weight, moving the X-axis and the Y-axis can take a lot of torque, especially when the moves are short. (One of the tests that I run to quickly heat up a motor is to move the motor constantly 1/4" CW and then 1/4" CCW with just enough pause between directions to allow the motor to come to a complete stop before changing direction. That "test" heats things up quickly.)

I think that the proper test in a 3-D cut would be to measure the heat of the motor that has to work the hardest. A lightly loaded motor would not develop much heat - just like you said, but the motor(s) moving the heaviest load would develop the most heat. Modern motors can handle about 80-degrees C rise with a maximum temperature of about 100-degrees C. That also assumes a duty cycle of 50% or less and a large heat-sink on the motor.

When all factors are considered, an efficient machine can be built that can handle the required job. Even if not all the factors are known, the price of the components is low enough that changes can be made as required. That's the "price" that's part of a do-it-yourself design. Commercial machines have been put through that cycle as many times as required until the machine does what it is supposed to do within the safety envelope required by whatever government regulations that may exist where the machine is sold and used.

Thanks for your input.