Is this a good motor to use?

I found what looks like a very good motor at an unbeatable price. Only four wires with 1.8 degree step; but the motor is 2-phase bi-polar. Comparing the model "SHINANO KENSHI #STP-58D4002-02" to what is the closest model number available on the Shinano website the motor _MAY_ have additional specifications of:
7.2 volt
1 amp per phase
7.2 ohms resistance per phase
9.2 kg/cm holding torque

I've asked for clarification of the specs from Shinano Kenshi on Monday with no reply. Previous experience tells me very few of these companies will help out a scrapper, so this might be the best I can do without actually having the motor in hand. At $8 a piece I've ordered a few...

Thought I would see what everyone here thinks.

[www.allelectronics.com]

The page you linked says 0.42 Ohm coil so I would guess it is lower voltage than 7.2V otherwise it would take 17A.

---

[www.hydraraptor.blogspot.com]

net3d Wrote:
-------------------------------------------------------
> I found what looks like a very good motor at an
> unbeatable price. Only four wires with 1.8 degree
> step; but the motor is 2-phase bi-polar. Comparing
> the model "SHINANO KENSHI #STP-58D4002-02" to what
> is the closest model number available on the
> Shinano website the motor _MAY_ have additional
> specifications of:
> 7.2 volt
> 1 amp per phase
> 7.2 ohms resistance per phase
> 9.2 kg/cm holding torque
>
> I've asked for clarification of the specs from
> Shinano Kenshi on Monday with no reply. Previous
> experience tells me very few of these companies
> will help out a scrapper, so this might be the
> best I can do without actually having the motor in
> hand. At $8 a piece I've ordered a few...
>
> Thought I would see what everyone here thinks.
>
> [www.allelectronics.com]
> 400600/STEPPER_MOTOR._1.8_DEG__STEP_.html

I guess the key question here is what are the implications of using a 1.8 degree step motor rather than the baselined 0.9 degree step motor? I'm assuming there would be resolution losses and software tweaks required.

Nathan

There appears to be firmware support for half step control, currently unused for the default motors. So it looks like someone already planned for controlling 200 step/rev steppers. Simon will no doubt know a lot more detail about that, incluidng how well tested that support is or is not (!) -- but, at least in theory, a quick edit of pic14.h to #define HALFSTEP appropriately and a firmware rebuild might be most of the work needed to support such a stepper? The host software should never need to know.

Jonathan

I've found a 6 lead unipolar stepper for $3.00!
UNIPOLAR |Lin Engineer 5618S54 with 21.5 ohms across phase to center tap, 43 across each phase. These appear brand new surplus and are in a 23 NEMA housing
Are you saying a C variable can be re-defined to handle the 1.8 degree, (200 step)
for the firmware?
Motor data can be found at [www.linengineering.com]








1.62"
41.1 mm

2.01"
51.1 mm
Lin

5618S-54
Amp per phase 0.5
Torque in NM 0.42
resistance ohm per phase 21.5
inductance mh per phase 28.4
Weight lbs.1.12

Ack, no good. You'll have to make an entirely new driver board. Unipolars have a constant common high voltage line and switch each individual coil to ground to actuate. Bipolars change the current flow through each coil set to move. They take entirely different driver configurations to accomplish this. While one board can be made to do both drive schemes, I don't think the RepRap driver boards are made that way from what I've seen.

They also tend to have poorer performance than a similarly sized bipolar.

I reviewed the data sheet for the ST5709S1208B stepper. When you place the two coils in series to make a bipolar motor, you have 10 Ohms per phase. Now 12VDC (and 10 Ohms
make a steady state current of 1.2Amps. The current builds up by I(t)=V/R(1-exp**-(R/L)*t) and the motor in this configuration is rated for 0.85Amp.
now (1.2/0.85)**2=1.99 or 99% overload at steady state, I bet it gets hot!
The 5418S-54 has 43 Ohms if we use the coil ends and cap off the center tap. The L is 56.8 MH for this motor verses 46.8 for the ST5709S1208B, so current buildup isn't as quick, but it won't overheat on hold torque. Physically the motors are indent cal, except this motor has a sheave installed with a retainer lip to keep the belt from slipping axially off! What is the fastest you run this motor anyhow? If the fan was to cool the steppers, then if you don't overload the steady state, you don't need them. Steppers run hot normally, just so they don't breakdown the insulation.
Look at time constants, 0.0568 henries/43Ohms=0.0013293 sec**-1 and 0.0468H/10ohms=0.00468 sec**-1. The 5418S-54 has a smaller time constant, so it will build up to steadt state current faster, due to the greater resistance! So resistance is futile! The recommended motor is better in resolution only! I think it is a hot potato for a 12 VDC system!

SOI,
You can always drive a unipolar motor from a bipolar drive, but not vice versa. You have the choice of using the coils in series by ignoring the centre tap, or just use half the coil. When in series you have to reduce the current rating by root two to give the same heating effect (I^2 * R). This can give root two times more torque provided the magnetics don't saturate (twice as many turns). But the inductance is times four in this mode so top speed will be less. Alternatively you can just use half the coil. In this case the current rating is the same to get the same heating so no torque increase and the inductance is the same so no speed reduction.

Dean,
The driver chip loses a few volts so the RepRap motors are not overdriven much if at all.

Chris

Edited 3 time(s). Last edit at 06/03/2007 06:22PM by nophead.

---

[www.hydraraptor.blogspot.com]

Dean Schrickel Wrote:
-------------------------------------------------------
> Are you saying a C variable can be re-defined to
> handle the 1.8 degree, (200 step) for the firmware?

Sort of, but not quite. A #define'd constant, not a variable.

I've skim-read some of the firmware source code, trying to get a feel for it. I didn't write it, and have never run it. This is an important distinction to make.

I'm saying that at some point in the development of that C code, someone (most likely Simon) made provision for "half-step" control of motors, which (*if* that support is complete and working!) I think could be well worth a look for anyone trying out a 200step/rev stepper). A quick find and grep shows me:

## > find reprap/firmware/ -type f -name "*.[ch]" | xargs grep -l HALFSTEP
reprap/firmware/devices/stepmotor/stepmotor2.c
reprap/firmware/devices/stepmotor/stepmotor.h
## >

So those are the files to start look at, not pic14.h, my mistake. Do go ahead and read them, then maybe try to build firmware with #HALFSTEP support enabled, test it on your stepper controller board and a stepper (carefully), and let the forums know what you discover :-) If as a result your hardware smokes or otherwise misbehaves... blame Simon, not me :-) :-)

Even if you can't use that halfstep support at all, all you really need is accurate linear motion on each axis... how many rotations of the stepper motor it takes to move the extruder head some specific distance is (I think -- bear in mind I am a software person and do not yet own Reprap electronics, never mind stepper motors!) "just" a per-axis calibration/preferences issue which you can (in theory) take care of in the RepRap host software. There are preferences labelled XAxisScale, YAXisScale and ZAxisScale in there for you to try out. They default to 400, and are in steps/mm, so apparently we're expecting 1mm of linear travel per full stepper motor rotation in a default RepRap Darwin setup.

Again, this advice/info is coming from someone who has read the code, at least somewhat, but not built any hardware... beware, I perhaps know just enough to be dangerous here :-)

Really, I was thinking more that if RepRap mechanics need the precision of a 400step/rev stepper, then using a 200step/rev motor and half-stepping it could get you very close to the same amount of precision. Simon is the firmware guru, I'm just a wannabe at this point :-)

Jonathan

I meant constant, I'll have you know I read "C for dummies" cover to cover!
Using only half of the total coil keeps the same L/R and shorter time to max current. Are the H bridges in the L298 made of bipolar transistors or MoSFETs? If bipolar, the loss would be 0.2 voltsX2=0.4 or 11.6 VDC. Is anybody working on Standard size parts and software, for us poor Americans who deal with feet, inches, pounds and Gc! I can find standard size hardware, just not metric!


O

See [forums.reprap.org] this thread for some at least partly related stepper motor discussion.

Jonathan

Dean,
The driver chip is a ST L298 bipolar device. It has a saturation voltage much higher than 0.2V. At 1A the total voltage drop is typically 2.55V. The outputs will be Darlington pairs.

Here is the datasheet [www.st.com]

Chris

---

[www.hydraraptor.blogspot.com]

Wow, what a voltage loss! I'll use a frying pan for a heat sink and cook eggs while it running. I'm sure their is a better driver IC, perhaps in surface mount, than this. Eventually, aren't the boards going to be made on the machine itself?

The key word being "eventually". If you have a Repstrap that can manufacture PCBs and do surface mount component soldering on them... *tell us* :-) Maybe Reprap v4.x will be able to do that??

Meanwhile, soldering surface mount components at home is not for those with even slightly shaky hands or poor eyesight... I don't think we should be expecting Reprappers to do that.

And oh yes, you *will* need a decent sized heatsink on those drivers :-)

Given the replicability goals of the project, spending a dollar (or five) more in cost for bigger heatsinks, and high-ish power consumption, in exchange for easier construction with only hobbyist tools, is a worthwhile tradeoff, at least for this first generation of Reprap, IMO. Further cost reduction and perhaps power consumption reduction are definitely worthy goals, but they can probably be left for future offspring of Repstrap Darwin.

Jonathan

"Further cost reduction and perhaps power consumption reduction are definitely worthy goals, but they can probably be left for future offspring of Repstrap Darwin. "
That's what Tommelise is all about.

[www.3DReplicators.com]

With a little luck by the time Darwin is up and running I'll have the technologies you're talking about running on Tommelise. At that point you can pull them into RepRap 2.0 or Mendel, as Adrian has named it.

Ack, I never remember what works where, and a lot of the motors I deal with have an additional case ground wire, so I thought all center taps were tied together from the "six wire" setup. I see I'm wrong after a little reading here and elsewhere. I tend to live and die (professionally) by PMAC three phase servo motors for motion control. Tends to warp my perceptions sometimes.

A discrete FET bridge based on TO-220's could be built, plus some driver chips, but you're replacing one part with half a dozen. Integrated FET stepper drives are usually surface mount.

Time to get back to lunch...

Ok, the motors arrived in the mail today. I had hoped that by some weird fluke they had measured the Ohms wrong, like the surplus store forgot to zero their meter (WRONG)... So, here I am with three perfectly good low wattage motors. They look great by the way. A++ condition!

I wanted to hook these up to provide some information about the torque, however I stopped short of holding one of the motors at 28 amps by connecting directly to the 12VDC power supply. I need to look at Ohm's law. I want to shoot for 3.5 amps and see if it overheats.

I = R / V

3.5 Amps = 12V / (0.42 + R) where R is the needed resistance.

Solving for R gives 3.

And I need to get the power correct.

P = E

Edited 2 time(s). Last edit at 06/06/2007 02:02AM by net3d.

Thought I would add, I found a neat little web script that does this all pretty quickly. Almost requires no thought. Just figure out what your resistance needs to be.

[www.sengpielaudio.com]

Why not use the 3.3V output on your PSU. That way you only need 0.5R 7W resistors.

---

[www.hydraraptor.blogspot.com]

That's a good idea! And really... I wonder how long they would hold up to 8 amps? I gave it a try this morning with direct 3.3V Excellent holding power!

I wonder if you could PWM it with 12V. I know the stepper controllers don't have current sense resistors or any of the other stuff you'd want to do it right, but maybe you can fake it. I'm just thinking out loud here, never tried this, but if it were my motor this is how I'd go about destroying it. Set up a 20kHz PWM at 0% duty cycle. Connect the motor through your ammeter (remember to switch to the unfused side :-). Gradually increase the duty cycle until you hit the motor's specced current.

Would that work? I'm probably going to go the cheap-o motor route myself, but not for a few more weeks.

Copper resistance.

As an experiment I held both coils on with full 3.3 voltage. As expected the motor quickly heated up to 100F or 40 C at which point I disconnected the power. The Ohms went up to 0.7 for this temp, and I assume normal operating temp could be hotter as most steppers and servos can operate at around 70+ C with continuous use. It may have been much hotter internally that what I perceived....

So, knowing that copper becomes nearly fully resistant when it melts (that would be bad!) and I want to keep as near to 3.5 amps as I can I should be able to apply a temperature coefficient. This is probably going a little over the top and my mathematical product is not going to be scientifically accurate. Thought I would run it through.

(R-R1)/R1 = A(T-T1) where R is the resistance and A is the coefficient and T is the temperature difference. I would have to take many (accurate) readings to find the curve for this motor. I've played with this formula a bit and can't get it to come out right....

(work)

Ok, with a little more research on the temperature coefficient. [www.allaboutcircuits.com]

R = (0.46 Ohms) [1 + 0.004041 ( 80 - 20 )]

R = 0.57 Ohms for the target operating temp.

So I think I shouldn't be playing with the 3.3V direct anymore having measured 0.7 Ohms. This puts the internal temperature of the motor at somewhere around 140 C if this calculation is right... :+)

Use the 3.3, just make sure you use it being switched and make sure you don't get near the full duty cycle. Since Adrian has been doing lots of work to make the console parameterizable, you could maybe set a max limit that you can't go past. Not sure if the software supports that or not yet.

The drive chip on the RepRap controller board drops up to 4.9V at 2A so you will need to start from more than 3.3V if you use that device. Also looking at the datasheet the supply voltage of the chip needs to be 2.5V greater than the logic one voltage so I think you are stuck with using the 12V rail.

You can use open loop PWM to limit the current provinding it is switching quickly compared to the rate at which the current builds up due to the winding inductance. That way, as long as you keep the duty cycle low it will never exceed the max peak current of the chip (2.5A). If it switches too slowly you could exceed the peak rating of the chip even though the average may be a lot less.

I have not looked at the RepRap PIC s/w so I don't know what speed it PWMs at. Unfortunately motors like these are likely to have a relatively small inductance. They are intended for high speed use driven from a constant current chopper driver which is the modern way of doing things.

---

[www.hydraraptor.blogspot.com]

Hi All:
I'm not a big electronics person so forgive my ignorance in this matter, please give me your feedback on the following stepper motor spec's.

Excellent for precision control
Can be operated in forward/reverse mode
Excellent torque/size ratio
Wide variety of supply voltages
Data sheet included
Dual shaft
Check for compatible power supplies
Step angle: 0.9 degrees
No. of phases: 2
Drive System: Bipolar
Voltage(VDC): 8.4
Phase resistance (

It looks compatible electrically but physically it is a lot smaller and probably doesn't have enough torque unless you want to build a scaled down version of the machine.

---

[www.hydraraptor.blogspot.com]

Nophead you must have read my mind, I was giving thought to scaling down the reprap design as I thought it would be easier to construct and a good entry point into the whole area. I could always scale up later once the mechanics are learned.

I believe it would be posable to use the existing software by using the lower left quadrant of the build platform only, and at just $6.00 a motor your not breaking the bank.

PWM is probably the way to go..

I'm experimenting right now with something called a "chopper". its an oscilator, a flyback connection and a current sensing resistor (1 ohm) in series with the "coil" which right now is replaced by a 4.7 ohm resistor - not ready to fry my $50 steppers just yet...

The pwm:ed signal is fed to a P-MOS transistor driving the coil.

It seems to be doing the right thing only the transistor gets to hot and starts to smoke after a while...

Not having a scope, I meassured the DC current at around 1.1A and the voltage drop over Drain-Source to about 2V.

Ok, I use an el-cheapo small cooling element for a TO220, but I still think 2.2W power loss should not be to much of a fuzz.

The only thing I can think of is if the AC component is whats bringing the CMOS to fry?

This evening I will try to meassure the frequency and duty cycle of the thing by using the fancy features of a multimeter bought at a gas station.

mimarob,
A few tips:

2W is about the most you can dissipate with a TO220 with no heatsink.

With a modern MOSFET you should be able to get a low enough RDSon to not need a heatsink at all, even at several amps. Looks like you are not giving enough gate drive to turn it on fully.

Sometimes a MOSFET will oscillate at very high frequencies causing it to get very hot. The easiest way to fix it is to include a resistor in series with the gate, about 1K, mounted close.

It is more usual to use N channel MOSFETS. They generally have a better performance / price ratio. For unipolar drive this is easy but for bipolar you either need to generate a rail above the supply rail for the gate drive or use P channel.

Edited 1 time(s). Last edit at 07/27/2007 09:37AM by nophead.

---

[www.hydraraptor.blogspot.com]

what is the voltage requirment for STP-58d2006-03 stepper motor??