Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: jimpinder on April 03, 2008, 02:53:16 AM
-
I'm trying to speed up my axis a bit - well a lot actually, since I cannot move them above 4in per minute at the moment.
Up til now, I thought it might be a mechanical problem with the lathe, but I took the belts off the motors and tried them on their own. I can't get them to spin any faster on their own.
I am running 1/8 step, 1.8 degree, 3 to 1 reduction, 1/10" leadscrew which, I think gives me 48,000 pulses per inch and therefore 192,000 pulses per minute for travel.
The motors are Arc Euro Trade Hybrid Steppers - 220Ncm, 2.5amp 7.5volt - and all in all I am satisfied with them - they are certainly powerful enough to do the job and at £22.95 each are relatively cheap.
They are 8 wire motors - and I have wired them in series.
The $64,000 question is - will they spin any faster if I wire them in parrallel - or is my computer - a Toshiba laptop - just up to its limit (says he looking longingly at the USB driver board just announced).
Anybody any ideas???
-
Jim putting them parallel will give you more torque but I would think lower speed.
Just out of interest are you using the Chinese drives that Arc Euro and others sell?
I have just built a small coil winder for a customer and I am having some issues, not sure exactly where the problem is lying yet but I am starting to think harware side of things.
Hood
-
The driver cards are marketed by Routout CNC shop in Wales - don't know where they are made - see Ebay.
Are the Welsh related to the Chinese???
Yes - that is what I though about the steppers - in parallel they will draw twice the current, so should give more torque - but I couldn't see an increase in speed.
I would try it, but the drives are rated at 3 amps max, and the motors are 2.5 amps per segment (7.5v). I am running on 24 volts and I am wondering if they will draw too much current and blow the drive
-
Jim,
Speed of a drive depends on how much torque the stepper can develop and the frictional drag of slideways and the leadscrew, hence the use of ballscrews and linear bearings. The drive system needs to be able to cope with what the stepper requires in terms of amps and volts. As Hood said parallel connection will get you more grunt at low speed at the expense of top speed.
At 4 ipm, 0.1 ins leadscrew and 3:1 reduction, your motor is only managing 120 rpm. Your top speed for the stepper might be about 800 rpm with a standard driver. You should be able to drive your leadscrew with at 1:1 with this size of stepper motor, so, with luck, you should be able to get up to about 80 ipm! The computer is probably not the limiting factor. Can you run Mach at a higher kernel speed, say 45kHz? Just keep pushing for higher top speed until the motor starts missing steps, then back off a bit.
Have a look at my post :-
http://www.machsupport.com/forum/index.php/topic,4764.0.html
I was trying to improve the speed of the G00 move. The Gecko's start out at 1/10 step at zero revs and "morph" to a simulated full step at a motor speed of 200-300 rpm. Micro stepping is a big help at getting the steppers through their resonant speed range, but not good for ultimate top speed.
Hope this is of some help,
Ian
-
Hi Jim
I am running 1/8 step, 1.8 degree, 3 to 1 reduction, 1/10" leadscrew which, I think gives me 48,000 pulses per inch and therefore 192,000 pulses per minute for travel.
Absolutely right - or - to put it another way a theoretical resolution per pulse of 0.00002 of an inch. Or more correctly taking out the microstep factor of 8, a theoretical resolution of 0.00016 of an inch. (microsteps should not be included in resolution calculations because they can't be relied upon to be equal in size)
So the first question is - do you require this theoretical resolution? The route from here on in is dependant on that answer. But for the time being let's just look at the status quo.
As you say, at 4 in/min you have a pulse rate of 192,000 ppm or 3200Hz. Clearly Mach/parallel port/USB etc. is not of issue here so let's look at your drivers and motors. Motors first.
The motors are Arc Euro Trade Hybrid Steppers - 220Ncm, 2.5amp 7.5volt - and all in all I am satisfied with them - they are certainly powerful enough to do the job and at £22.95 each are relatively cheap.
They are 8 wire motors - and I have wired them in series.
We need to be clear here exactly how you have them wired and exactly what each coils resistance is.
If indeed you have them wired in series and the 2.5Amps you quote is per coil then you may well be halfing your potentially useful current. But we need to be sure we're both on the same page here if we don't want to damage your drivers.
You need to get your multi-meter out and measure the resistance of 1 of the 4 coils. If you do that and get back we can continue.
Cheers
Ian
-
Thanks to the two Ians and Hood -
Yes - Fired up from the replies I went in the workshop and had a fiddle. I tried altering the incremental steps. I seemed to get more speed from 1/2 step. Full step was a bit too rough. I upped the speed to 10 ins per min. I got the crossslide travelling at that speed but the leadscrew baulked at it.
I get the idea that there is a bit too much resistance for it to accelerate up to top speed. I've tried to adjust the acceleration with some success, but not reliable. I think as Hood said, I could do with a bit more torque to get over the sticky patch.
My motors are wired in BiPolar series, i.e. coil A to coil C (A - A' - C - C') and B to D. The coil resistance is (across two coils in series) 5.8 ohms, across one coil is 3 ohms. This relates to the information about the motors - 2.5amps at 7.5 volts - which is on spec exactly.
My problem is, of course - if I now wire them in parrallel the net resistance will drop to 1.5ohms. I don't know enough about the makeup of stepper motors to know - if I run them at 24 volts how much current will they draw - my drives are RoutOut CNC and in the data sheet are rated at 2.5amps.
-
Having a devil may care attitude to money - I am going to wire up the leadscrew motor in parrallel and see if the driver will run it without having a terminal thrombosis.
On a serious note - I put my meter in line with the motor card supply and then jogged up and down. I appreciate it might draw a bit more when working, but jogging (wired in series) the motors only draw 0.5 amps. Even if that went up to 1.5 amps under load, I will still be under my driver capability, even wired in parrallel.
I'm going to try it and see what happens.
The little Chinese man in Wales might get some more trade.
-
Jim - don't do this - you'll likely blow your driver. I'm currently composing a solution for you.
Your meter is not telling you the truth.
-
Sorry Jim - had to be quick off the mark there - don't want to see you trashing things unnecessarilly ;D
I think as Hood said, I could do with a bit more torque to get over the sticky patch.
My motors are wired in BiPolar series, i.e. coil A to coil C (A - A' - C - C') and B to D. The coil resistance is (across two coils in series) 5.8 ohms, across one coil is 3 ohms. This relates to the information about the motors - 2.5amps at 7.5 volts - which is on spec exactly.
My problem is, of course - if I now wire them in parrallel the net resistance will drop to 1.5ohms.
This is good news. Now to get that torque. Wire your motors in "bipolar single coil" configuration.
Basically this means you only wire 1 coil of each phase and forget the other two coils. Doing this you'll have a phase resistance of your 3 ohms (actually it should theoretically be 2.8 but it's near enough)
You'll effectively double your phase current and yet still be (just) inside or at your limit (of the driver) of 2.5Amps at around 2.3 to 2.5 Amps. (actually you should be fine because the manufacturers allways give a little leeway).
You should now see quite an improvement but next we'll look at how to use that extra torque to get even more speed.
BTW the reason your meter is not giving you the true story is that if you're measuring the current "drawn" by the drivers from the PS then remember what I said in the other thread about chopper drivers drawing less current than they output. Well this is part of it. But also, the current is being drawn in high frequency pulses and no "regular" meter can read this oscilating current fast enough so it gives you a much lower reading. BUT the real killer is that it's the power transistors that are delivering the FULL motor current that will be maxed out.
Cheers
Ian
-
Jim,
The other Ian is correct - DON'T TRY IT.
Resistance is not a good measure in this case, inductance is the important parameter. Measuring current with a meter, only gives you the average current draw, the waveform is a pulsed, square wave and as far as your drives are concerned, it is the peak current that will blow them. I have Routout drives on my Novamill and will post later giving Mach max velocity and acceleration.
Ian
-
OK - coffee made - where were we?
Hopefully when you've wired them single coil you should see quite an improvement in torque (should be around double). Now ideally (given the spec of your drivers) we'd like to up your voltage from 24V to 30V which I think is the max your driver allows so that we can maintain as much of the the torque as we can while we up the speed. However there's no quick fix to your PS to do that. You could get/build a new PS but frankly you're not going to get much extra for your money. You'd need 48V to double your speed at whatever the current stall speed of your motors is with your new torque (~8in/min?).
So first off I'd get rid of your 3:1 reduction. You'll increase your speed by 3 obviously (~24in/min). But I think I remember you saying you were about to get some new ballscrews? well if you do - consider a larger lead than your current 0.1 inch. In my first post I asked if you really needed the theoretical resolution of your current system. Decide what resolution you can work with and spec your screws accordingly.
So for example - if you go to 1:1 reduction your res will be 0.00016 * 3 or 0.00048. Double your lead and you'll be around 0.00096 - still within a thou. (@ ~48in/min) Anyway you get the idea.
EDIT: I'm just going to throw this in to possibly spoil a nice party - sorry ;D - You're motors are currently doing up to ~3 revs/sec. With these changes we'll hopefully see that increase. I'm now thinking we'll see how good these Chinese drivers are at mid band resonance compensation. :-\
Cheers
Ian
-
Ian
WWWWHHHHJEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE. Not quite the sound of my motor. I'm sorry, I was quick off the mark, having read your post just now - particularly the bit about the meter reading.
No harm done, it would seem. I wired in parrallel and got the motor up to 30 ins per min. I ran through 1/8 to 1/2 steps, but I have ended of 1/8 phase again. I settled on a modest 15 " per min for reliability. This is still nearly four times faster. Even at 20" per min it was good, with only the occiasional stoppage.
I will take heed of your warnings however, and disconnect two of the coils, C and D, and see if it makes any difference. It certainly did so far.
Meter reading for parrallel - 0.7 amps. I must admit I thought this was far too low.
-
Jim - That's a shame - I bet you could have fried an egg on your drive output transistors. Talk about destructive testing :(. You've been lucky - but if you do this again you're running the risk of considerably shortening the life of your drivers.
-
Right -
I rewired all three steppers - to two single coils, refitted them all.
Great - I can now get 16 in per minute on all axis. This is fabulous - 4 times what I was getting on bipolar series wiring.
This for me is fine. My mill table is only 12" by an effective 10" so I can move any where in 45 secs. In lathe mode the length of the leadscrew is 18" so I can move from end to end in about a minute.
Thanks for the other suggestions, but altering the gearing, which are timing belts is a bit difficult, because I have glued the gear onto the motor shaft, although this would increase the speed. The Z axis I would leave because I want the torque when drilling and it can be up and down in seconds.
The voltage increase is also a bit difficult, because I use a pair of traction batteries to provide a very steady 24 volts.
That leaves the leadscrew, which I will replace eventually, to cut backlash, and I will certainly consider increasing the pitch of that.
If there are any more suggestions thanks - and thanks for the help etc already - a fourfold increase for no expenditure is excellent.
Jim.
-
Jim,
Fourfold increase, excellent and without frying your drives!
Details of my Novamill/Routout drive setup as promised.
Set to 1/8 microsteps
Sanyo Denki 2A size 23 motors 0.83Nm bipolar (four wire) 0.8 ohms/phase, 3.8 mH/phase
Ballscrew 5mm pitch
2.5:1 reduction
CNC4PC C11 BOB
Max speed 2.0 m/min motor speed 1000rpm.
Max Acceleration 383 mm/sec/sec
Mach set at 45kHz an 800 steps/rev
To give you some idea of the friction reduction using ballscrews, with the stepper disconnected, I can push my table by hand and rotate the ballscrew. That would be impossible with a 10 tpi leadscrew and nut. So your idea of fitting a new leadscrew would certainly make a big difference. A Gecko 203V would give you the amps to drive the
stepper windings in parallel, no need to alter your novel power supply. You might be able to reduce the diameter of the leadscrew pulley and use a shorter belt.
Ian
-
I may get shot down in flames but this is my experience with the need for speed.
First off i bought an all in one system that sells on ebay it goes on about not needing a dodgy printer driver. But although it has some advantages it just wasn't up to the task. i did a lot of research about cnc "which i should have done first" i found Mach.
I disconnected the drives from the system and tried it on a BOB from cnc4pc every thing about it was so much better in terms of smoothness but i still couldn't get the top speed. More research and i came to an article on the Geko site it mentioned the ratio of voltage supply in relation to motor voltage.
I measured the power supply on my ebay cnc box the voltage was 25 volts similar to you Jim. So Mach was much smoother but i still couldn't get the top speed as the motors stalled.
I sold the lot and bought drivers from Motion Controlled Products "Keling" with 80 volt power supplies and MDS980 drivers.
This setup is fantastic. So my overall statement is more voltage = more speed, more amps = more power you need both.
As this is on a mill I CAN run the motors at 3500mm/min with no stalling BUT i only run it at 2000mm/min max feed rate this gives me a big comfort zone in not missing steps.
I have had to tweak the acceleration as with a larger mass it takes a while to get up to and slow down from speed.
So IMHO Jim you will not get the desired effect from your current setup but i have no idea if 25volts would be OK for a router as the tables have less mass so maybe that voltage is OK then.
HTH
Phil_H
EDIT i have just remembered i did the original test 25V ps and Mach CNC4PC bob, the motor was unconnected on the shaft IE i was only spinning the motor, i couldn't go above 1500mm/min without stalling "Sorry Bad Memory"
So my statement is that For high top speed you need maximum voltage.
-
That's weird when i first posted i didn't see all the earlier replies so that's why i am somewhat late.
Phil_H
-
Thanks for the replies - I'll bear in mind the Gecko 203V card if any of my present ones fry. I can put the motor(s) back to parrallel then FLYYYYYYYY. :D :D :D
-
Hi Jim - I've been doing a bit more thinking on this.
It all comes down to the fact that steppers are generally over driven in terms of Voltage in order to extend their torque range over the increased speed. However they should not be overdriven in terms of current.
All drives must therefore have some mechanism for limiting the current through the motor regardless of the Voltage. I know how this is achieved in chopper drives but not so clear on how in non-chopper drives.
Anyway - The bottom line is that IF your drive is a chopper, then you should be able to set the current it at its output to the current it "delivers" to the motor. If it isn't a chopper, then at the moment I don't know. However IF you can limit your drive to its max current of 2.5Amps or 3Amps (you've stated both so I'm not sure which it is), then you WILL be able to wire your motors in parallel with no ill effects on your drive.
The advantage of this is that if your drive is 2.5A max then the motors will still only get 2.5A but the current will be less than the rated parallel current of your motor - therefore you'll get less motor heating than with single coil wiring - not that I suspect you get much heating anyway.
IF indeed your drives are max 3A then you win even more - that is still less than your motors rated parallel current but obviously more than the max 2.5A of your motors single coil current.
Obviously the ideal would be to run your motors at their rated parallel current but given your driver isn't up to this then this is probably the best you can do with what you have.
Ian
-
Yes Ian - I,ve thought of that too.
I can set the output current on the drives. If this is current limiting to the drives, then I would think that I can put them in parrallel, and get the benefit from using both coils.
On the quick experiment I did on parrallel, it would seem that there is more power than in single coil config, and I get a little more speed, which gives me a bit more choice.
I will get on to the supplier on Monday and find out. If I can use them in parrallel without problems I think I will
-
Hi
I have the bigger drives (10 amp) from Routout CNC and I also looked at the smaller ones they supply, both have current limiting, the larger you set by using a volt meter across some special contact to set the current, a table is supplied. Can't remember how the smaller ones are set.
Just looked and here's the pdf setup file for the smaller 2.5 amp drivers http://www.routoutcnc.com/RoutOutCNCStepper.pdf
Hope this helps.
Steve
-
Yes- thanks -
I have the instructions, the set up is much the same as for the 10 amp, you turn a pot til the voltage on two pins in 1.4v multiplied by the amperage you need. It is working well - as must be by definition, current limiting. I feel I could go back to the motors in parrallel and run them - as long as I set the outputs right.
Someone is lending me a Gecko drive - rated I think, at 7 amps and voltage way up.
I am going to try this on my leadscrew so I can put the full current through the motor. If I add another battery to my system, I can put the voltage up to 36 volts on the Gecko - that sould make the sparks fly.
I am thinking of adding a forth axis to the system - so that will be a chance to get a more powerful driver card.
-
As your driver can only source 2.5A it can not supply the rated parallel current of your motor which is 3.5 Amps. The data sheet describes it as a PWM driver which I think is synonymous with it being a chopper driver. If so, then I don't think it will be damaged by your proposed parallel wiring scheme - but don't take my word for it. That said, at 24V in both cases your motor will have the same available power of 60W. Had you been able to source 3.5A then clearly the power would have been 84Watts in parallel as opposed to the 60W in single coil.
Despite the fact that in the first case you're "overdriving" voltage wise by a ratio of 3.2:1 and in the second case by 4.52:1, I can't see it actually makes any difference - power is power is power after all. But from your reaction when you briefly tried it I get the feeling it did make a difference. So if you feel it's better in parallel go for it. The only thing I can think of is that if it is better in parallel then I'm guessing it's because reluctance is coming into the equation - back to the books!
If you do switch to geckos (other great chopper drives are available) and up your V - make sure your rig's well bolted down - that thing's going to go so fast it'll make your head spin. ;D
-
Stirling,
Just some thoughts on current into a stepper. When you apply a voltage, because of the inductance of the coil, it will generate a current in the coil. This will start at zero amps and rise rapidly to whatever maximum you have set your driver to, then decay to zero with a time constant depending on the inductance. So the driver is not generating a square wave pulse of current and you cannot just multiply amps and volts. In a reply on the Yahoo Gecko group, Mariss said that a 23 frame 2 stack(2.5A) motor would generate about 35 watts with 30 volts applied and doubling the voltage would double the watts(power) available to overcome inertia, friction, detent torque and drive the cutting tool.
So how does this help Jim? It looks like you were right to point out that he could use parallel connection of his 8 wire drive if he sets the current limit on the drive and gain some useable torque perhaps at the expense of top speed and that if he can try his borrowed Gecko and increase the voltage even more power will be available.
Keep th information coming Jim, we are all learning!
Ian
-
Hi Ian
I think we're on the same page more or less with this. I understand that the driver does not induce a square wave of current. As I understand it, that's exactly why we use over voltage in order to try to get the leading edge of the current waveform as steep as possible.
I wasn't my intention to suggest that the output mechanical power of the motor is VA, that's why I said "your motor will have the same available power of 60W", meaning from the PS - which is VA. Clearly some of this is not converted into mechanical power otherwise we'd have a 100% efficient motor and we'd all be rich ;D
I agree that the output power of the motor effectively doubles as the PS voltage doubles - Mariss told me that too when he said this (I'm sure he won't mind me including it here.)
"Look at it this way: Doubling the power supply voltage doubles the mechanical power output of a step motor. The efficiency of the motor doesn't change, so twice as much electrical power has to be supplied to the drive. Electrical power is Volts times Amps. Because you doubled the voltage, you doubled the electrical power at the same amperage.
Conclusion: Power supply current is independent of supply voltage for a given motor. It is the same value regardless of power supply voltage.
Mariss"
What I still don't really understand though is if Jim does indeed get more oomph by wiring in parallel as opposed to single coil, where is that extra oomph coming from? Like I said, there is 60W of electrical power available to drive the motor in both cases. As Mariss says - the efficiency of the motor doesn't change at the same amperage, and as you've said - not all of that will be converted into mechanical power, - but whatever proportion it is, is the same.
The only thing I can think of at the moment is prbably absolute rubbish but I'll throw it in anyway. We know that the current is induced in pulses. We know that the quicker we can get the current to rise to its 2.5A the better. Inductance of parallel is less than single coil. This presumably allows the current's leading edge to be steeper. Therefore the 2.5A is "present" in the motor for longer. Therefore the power is available for longer on each pulse than it is for single coil. - just a thought.
Ian
EDIT: As you correctly say - we're all learning. A thought just crossed my mind. A chopper will apparantly only draw about 2/3 of the rated current of the motor. Is this (I wonder) because as we know it's applying that current to the motor in pulses, it's only actually applying it for 2/3 of the time?
-
I cant nsupply the answer yet, becasue I've been away and only got back tonight, Went straight down to the post office and picked up my Gecko drive - looks good. Will try and fit it tonight.
As far as power in a motor is concerned Watts is the measurement. Watts = Amps x Volts. The only trouble is you need to increase voltage to increase the amperage becasue Voltage = Resistance x Amps - and for the sake of the argument resistance of the winding is constant ( in this case at 3 ohms - given in the rating of 2.5amps at 7.5volts).
There are, we have found, several different ways to connect steppers ( the four coil variety).
1.) In my first effort - with two windings in series - I was effectively putting 6 ohms of resistance across my voltage. Using the 7.5 volt rating as a rough guide, then this would mean that the two coils would draw 1.25 amps. Again using 7.5 volts this would give 9.375 watts x 2 = 18.75watts absolute maximum.
I am not aware what effect the chopping of the applied voltage (24v) has on the final voltage at the motor. It must be similar to the PWM output and vary with the number of pulses per sec opposed to the carrier frequency. Since the carrier frequency is usuall well above the maximum chopping frequency, then there must be quite a cut in apparent voltage at the motor - hence I will use 7.5 volts in all the examples. (It does - however - increase with speed)
2.) In the ONLY WIRING TWO COILS ON THE MOTOR scenario - each coil is 3 ohms resistance - and draws 2.5 amps (the rating for the drive) but bear in mind there is only one coil operating - this give 18.75 watts - BUT THE EFFECTIVE VOLTAGE INCREASES AS THE SPEED INCREASES - THEREFORE THE APPARENT POWER OUT INCREASES. To me it seems that twice the speed (of pulses) would give twice the apparent voltage right up to the point where current limting came into play. It seems to me also, that the current in the coil would rise much more quickly than in No.1 and therefore be higher longer and give more power earlier - perhaps helping it get over the limits of the series wound option.
3.) When I connected the motors with two coils in parrallel, the effective resistance Rt = R1 x R2 / R1 + R2 = 9/6 = 1.5ohms. This has the potential to draw 5 amps - and the current limiting would be working over time to keep it down to 2.5 amps - BUT there are TWO windings working on each pulse therefore twice the power - so although there was only 2.5 amps to share between the coils, I think it was sucking a bit more out and then with twice the number of coils, giving a little over that at No.2. Again because resistance was less, current would rise more quickly giving more power per pulse.
It certainly had a bit more Omph.
I will wire the Gecko up and try it wired to give the full five amps to a motor connected in parrallel and let you know the result.
-
Sorry Jim but there is waaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaay more to it than you've suggested. You're talking pure Ohms Law which is wholly inapropriate other than to work out the static ratings for a motor. Once that motor moves you're into a whole different ball game.
Can I suggest you re-read Ian and my posts because we've moved a lot further on since you were last here.
Thanks
Ian
-
Jim,
Like Ian Stirling said, you are not quite there with your understanding of what's happening when you apply a voltage to a stepper coil. It is as if the coil had the mechanical equivalent of inertia, you apply big volts to get it to move, but you need the amps quicker to get power. Yes, two coils in parallel have less resistance, but they also have less inductance and therefore the current rise time to maximum is shorter, so within a given pulse, maximum current is on for longer.
Ian (Who lives in hope that one day he might reach a reasonable plateau of understanding!)
-
I will wire the Gecko up and try it wired to give the full five amps to a motor connected in parrallel and let you know the result.
Jim - if you do this you will be overloading your motor current wise which is a really bad idea. You run the risk of significantly shortening its life. The rated current of your motor in parallel is 3.5Amps - not 5Amps. Putting figures to it, the coils will heat up to just over twice their rated value. (This is not to be confused with acceptable motor heating which is the result of overdriving VOLTAGE wise. I know it's confusing but the two are entirely different.)
the rated parallel current of your motor which is 3.5 Amps.
-
Ian and Ian - Yes I know I was missing out a great deal about inductance, and current lag and all the other things I've read about. I would understand it if I spent enough time reading and studying it - but I haven't the time.
We are, though, near the end. The Gecko works a treat and I have it up to 30ins per minutes, which from a start of 4" in quite remarkable.
Ian or Ian - sorry are you saying from coils in parrallel I should settle at 3.5amp - not 5amp. I will keep checking the motor - it was not getting warm (although it was only on test - and not working). I can drop the rating on the Gecko.
Certainly where I went wrong was - when the motors said they were rated at 2.5amp at 7.5 volts - per coil - I then opted for 2.5amp drives. There are four coils of course - two pairs - and it follows that you need more current capability.
I have done all this by merely altering the motor wiring and upping the current capability of the system - I have the voltage side to go at yet - although that would mean Gecko's right through, rather than just the one I have at the moment.
I better stay as I am for the moment (otherwise the financial director will complain) - I have the two motors using only 2 coils on the y and z - using their original drives upped to the Max at 2.5amps. The X (or Z on the lathe) with the four motor coils in two pairs, powered by the Gecko.
Either way we have a big improvement in performance !!! Y and Z at 16 ins per min and X at 30. I am settling for 14,12 and 28 for accuracy and see how it will go.
-
Ian or Ian - sorry are you saying from coils in parrallel I should settle at 3.5amp - not 5amp.
If you don't want to exceed the manufacturer's rated values, yes.
Certainly where I went wrong was - when the motors said they were rated at 2.5amp at 7.5 volts - per coil - I then opted for 2.5amp drives. There are four coils of course - two pairs - and it follows that you need more current capability.
It depends on the application, you might decide that series winding is advantageous, in which case a 1.77A driver would do it. I'm currently working through this very process for a system I'm designing and at the moment it's looking like serial may be the way to go - but I need to learn more first. Like Ian R say's - it's a long and steep old learning curve.
I better stay as I am for the moment (otherwise the financial director will complain) - I have the two motors using only 2 coils on the y and z - using their original drives upped to the Max at 2.5amps.
post #28?
-
Jim,
Congatulations on the increased speed! You might have got similar results with the Leadshine driver from Arceuro or Motion Control but I have not tried them.
You asked "Ian or Ian - sorry are you saying from coils in parrallel I should settle at 3.5amp - not 5amp."
Ah - Here I think Ian S and I have a different view. For the Routout drives set it to 2.5 Amps (single coil). For the Gecko - set the output control resistor for 5 Amps (2.5 + 2.5) ( Edit by Ian R. This wrong see later) and note that the Gecko needs to be mounted on a heatsink above 3 Amps. It is recommended to match the driver output to the coil and not overdrive it as Ian S said.
Ian
-
For the Gecko - set the output control resistor for 5 Amps (2.5 + 2.5)
LOL - I'll come clean - up until a short while ago I'd have agreed with you Ian.
As I think we agree, the rated values are specified such that we don't overheat a coil.
The problem comes about because we all get too hung up on the purely electrical side of things and we forget that motors are physical beasts - like when you drop one on your foot it hurts. ;D
The two coils that make up a single phase are not two isolated coils. They are inter-wound together forming one physical mass of copper. Each winding is not just heated by the current induced in it, it's heated by its partner as well. Because heating is a function of power (and hence why we buy 250W kettles and not 10 Amp 2.5 Ohm kettles or whatever), the total power and thus heating of any wiring combination is arrived at by summing the power of the individual windings that make up that phase and not by summing the currents.
Ian
-
Jim,
Excuse us for a moment we have come to a fork in the road :). It is relevant to how you set up the drivers, but as long as neither the driver nor the motor are overheating, you should be OK.
Ian,
I was under the impression that the steppers we use were all four pole devices and that in an eight lead motor they were just brought out separately. Are you saying that they are wound as a pair of wires serving two poles? Can you point me to a diagram, as I have yet to dismantle a motor. What are the advantages?
Ian
-
You two chat amongst yourselves - I had the back off one of the motors - it fouled the lathe bed and cut the wires (lost a drive) - it needed re-wiring so -
It looked to me as though there were eight coils in my motor - wired as four pairs (one pairing the opposite).
On diagrams I have seen of stepper motors, there have always been four magnets - which mean that they must be seperately wound. If they make eight wire motors - they must be four seperate coils, but having said that - how you wire them gives several choice
-
Ian,
I dropped the map! Jim's motors are 8 wire 2.5A per phase, so he should set his Gecko to 2.5A max for parallel connection.
I was also confusing the number of poles with the number of phases and being missled by the simple diagrams that are used to describe stepper motors. A bipolar stepper has two phases 4 wires or 8 wires. This to enable the motor to be stepped by alternately feeding each phase with a pulse of current. It may have many poles depending on it's design.
I think I am begining to understand why the torque drops off with speed. More later.
Ian
-
Hi Ian EDIT: posted this before I realized you'd posted again. :)
just to be clear we're on the same page, we're talking 2 phase motors - yes?
So, yes - 4 poles, 2 phases (each and every magnet, electro or otherwise has two poles - yes?). A four wire motor has 2 coils - 1 per phase, an 8 wire motor has 4 coils - 2 per phase. You don't need to take a motor apart to consider what would happen if the coils of a phase were in different physical locations. Apart from anything else - how would you pack them in to the available space optimally?. Will try to find a diagram.
Meanwhile, here's some datasheets for motors where the manufacturer has been good enough to give the ratings for each wiring option so we don't have to work them out. If what I'm saying about heating is wrong - why else are the rated values as they are?
http://www.kelinginc.net/NEMA23Motor.html - some are 4 some are 8 - for our purposes check the 8s. Obviously 4 wire motors only have one wiring option.
PS The reason I suspect Jim's motors never get particularly warm is that with the voltage he's using I doubt if they're ever conducting the excess current for long enough. When Jim was running his motors in serial he was also over currenting them by around 100% and they didn't get warm then. What this shows (to me anyway) is that Jim has never been running his system optimally and until he ups the voltage he never will be. The original question concerned getting more speed - voltage not current, gives speed.
Ian
-
On diagrams I have seen of stepper motors, there have always been four magnets
Then you're not looking at 2 phase motors.
-
My apologies - four poles - two magnets.
I don't think we are going to resolve this unless the manufacturers give figures for their motors - all from the same Hymnbook.
If I can run a motor at different voltages - Routouts sheet with the driver say it is the normal to drive motors with 3 or 4 times the rated voltage, then how are we ever to know the current they will draw (at that voltage). I noted the figure given for the motors above and the current drawn by the various different wirings was in the ratio Unipolar 4, Bipolar series 3, bipolar parrallel 6 - but in quoting the current, nowhere was an appropriate voltage quoted - except one, where the voltages were different for each method of wiring.
I think it is going to be a matter of intuitive guesswork - and a continuous check on the warmth generated to see if your setting have any adverse effect. Yes - I must admit I try and stay conservative - I am only a hobby/part time pro - I build maybe two engines a year - so I do not need to try and get every last ounce from the machine.
It would still be nice to know though - and to be able to advise others - what is the best way to wire steppers for the quickest, reliable performance,
-
I don't think we are going to resolve this unless the manufacturers give figures for their motors - all from the same Hymnbook.
I don't know what to say Jim. I've tried to explain the best I can HOW and WHY you calculate the rated values. I'm not sure I can improve on what I've already said.
If I can run a motor at different voltages - Routouts sheet with the driver say it is the normal to drive motors with 3 or 4 times the rated voltage,
Maybe with their drives - I don't know - but as a general rule this is a pretty meaningless statement. The torque curves for the motors I use are generally calculated at around 8 times. Gecko drives recommend a minimum of 4 and a max of 25.
then how are we ever to know the current they will draw (at that voltage).
Again I'm not sure what to say. I've tried to explain that motors "draw" the current made available to them by the driver. Voltage is effectively independant of current in a dynamic system. That is the whole point of how motors and drives work together.
I noted the figure given for the motors above and the current drawn by the various different wirings was in the ratio Unipolar 4, Bipolar series 3, bipolar parrallel 6 - but in quoting the current, nowhere was an appropriate voltage quoted - except one, where the voltages were different for each method of wiring.
Again.... These are STATIC RATED values.
I think it is going to be a matter of intuitive guesswork - and a continuous check on the warmth generated to see if your setting have any adverse effect. Yes - I must admit I try and stay conservative - I am only a hobby/part time pro - I build maybe two engines a year - so I do not need to try and get every last ounce from the machine.
???
It would still be nice to know though - and to be able to advise others - what is the best way to wire steppers for the quickest, reliable performance
???
Jim - do you remember Chioticones animated avatar where the guy beats his head on the keyboard until he's a bloody mess - well that's me now...
-
LOL Ian
May I suggest if you have the time to write a summary of the conclusions you have come to so far. From my understanding of steppers it is the voltage that determines the speed and the current that determines the torque and I suppose most importantly the way the motors are driven has no relation to V=IR, is this correct?
I am dealing with AC Servos at the moment and from my understanding they have a constant for Torque/Amps and Voltage/RPM and the drive is able to supply the current and voltage independant of each other, this means that I can get max torque no matter what RPM the motor is turning. Obviously it is different from steppers but the principle of voltage and current being independat of each other seems to be similar, and the relation of Voltage to RPM and Current to Torque seems to be similar.
Maybe my understanding of both steppers and servos is totally wrong so I have already donned my hard hat and have taken cover :)
Hood
-
Ian and Hood,
just to be clear we're on the same page, we're talking 2 phase motors - yes?
Yes they all seem to be two phase but can be multipole ie more than 4 windings.
I found this website :-
http://www.allaboutcircuits.com/vol_2/chpt_13/5.html
Scroll down to see the full story.
Yes, more voltage gives more top speed but if the motor does not develop enough torque to overcome detent torque and friction, you will never get there. So more amps to get enough torque.
Yes, heating effects, from eddy current losses and copper losses, limit how many amps you can feed the motor or it will get too hot and burn out the insulation.
Something for you to think about. A stepper turning at 1000 rpm is being fed by an 8 microstep driver at the rate of 200 x 8 x 1000 = 1,600,000 square wave pulses per minute i.e. about 27kHz. From a time when I tried (and failed) to design a high amps, high frequency transformer, this is realy pushing it with single wire coil and steel laminations. Bifilar (two wire) windings helped but the ultimate was lintz wire, with many small wires and braided. This reduced the copper losses and the iron losses were reduced by using ferrite E cores. With 27kHz square waves feeding into metal cored coils, I am surprised that we can even get these steppers to turn! It perhaps explains why I can get an extra 20% speed out of mine using Gecko's, because they effectively impose a full step at speeds above 2-300 rpm. Perhaps we can look forward to ferrite cores as the next move in steppers. (I could not find anything on the net}
Ian
-
LOL Ian
May I suggest if you have the time to write a summary of the conclusions you have come to so far.
Ah time, there's the rub - however maybe it is time to conclude this thread and continue elsewhere, I think between us we've long ago answered the original question and I think we're moving into areas well beyond its intended scope.
From my understanding of steppers it is the voltage that determines the speed and the current that determines the torque and I suppose most importantly the way the motors are driven has no relation to V=IR, is this correct?
Kind of, but not quite so simple - as Ian says below.
What I was really trying to say is this: In a real world motor, torque stays roughly constant from 0 RPM up to several hundred RPM and from there it drops inversly with speed. A caveat to this is that below about 500 rpm, torque is affected by patches of resonance, but let's ignore that for the moment. Suppose you have a system that is running with the rated current of 5 Amps, at 12 Volts. Let's also suppose that your system has the torque for the job but has limited top speed in that it stalls at 1000 rpm. If you now double the voltage to 24 Volts - you'll find that you've upped your top speed to around 2000 rpm. This is of course in an ideal system and in reality you may not quite get that increase but you get the idea.
The biggy to notice however is that you have not INCREASED the TORQUE of your motor because you havn't increased the current - it will still not go above 5 Amps. You've just increased the speed at which the torque is available. I suppose in effect you've stretched the curve.
With regard to Ohms Law - we have to be carefull - it's not that it has no relation - we just have to be aware of when it's appropriate and when it isn't. One thing I think that leads to confusion is that I've seen stepper motors variously described as being DC motors and as AC motors. I think I'm right in saying - well it depends on how they're driven. One thing is for sure though - they are not constant DC motors and Ohms law (as far as I'm aware) only "works" for constant DC. (Constant DC is not to be confused with limited DC)
I am dealing with AC Servos at the moment and from my understanding they have a constant for Torque/Amps and Voltage/RPM and the drive is able to supply the current and voltage independant of each other, this means that I can get max torque no matter what RPM the motor is turning. Obviously it is different from steppers but the principle of voltage and current being independat of each other seems to be similar, and the relation of Voltage to RPM and Current to Torque seems to be similar.
Maybe my understanding of both steppers and servos is totally wrong so I have already donned my hard hat and have taken cover :)
Sorry Hood - I have no knowlege or experience with servos but it certainly sounds from what you've said that there are or may be similarities.
Ian and Hood,
just to be clear we're on the same page, we're talking 2 phase motors - yes?
Yes they all seem to be two phase but can be multipole ie more than 4 windings.
I found this website :-
http://www.allaboutcircuits.com/vol_2/chpt_13/5.html
Scroll down to see the full story.
Thanks Ian - interesting read - one of the best I've seen particularly from the construction point of view. But oh boy, where do I start? - Oh well, here goes nothing as they say...
I think it's important to start by saying I'm not disagreeing with anything it says. Rather, I'll say it's allways difficult to be absolutely in synch when we use words to describe things. We use a word and assume the other person thinks the same as we intend them to think - but this isn't allways the case sadly.
I'll put it like this: Is "poles" the best word to use or would "notches" be better or maybe worse? Is "coils" the best word or would "windings" be better or maybe worse?
The "notches" as I'd call them here are what gives a motor its resolution. The "poles" (as in magnetic poles) as I'd call them are the bits (as I know you know) that "intensify" the magnetism when you energise the "coils" or "windings" and so ultimately give it it's number of phases - in a 200 step motor 100 of these would be energised at a time - or is it 50? I lose track. Is one piece of wire wound (or coiled) round a stick a coil or a winding? Is that same piece of wire coiled (or wound) round one stick and then off to another stick, one or two coils (or windings)? In short I would still describe this as a 4 pole motor and only introduce the word "notches" if we were discussing resolution. Maybe I'm wrong, maybe I'm using the wrong words, maybe I'm just really bad at explaining what I mean. ( I know one thing - I'm going to give it a rest after this because this is beginning to do my head in).
The reason we're discussing all this b*ll*cks is because of where we started - remember "rated values" and how I said the interwound coils (or inter-coiled windings) have a heating effect on each other? Well doesn't this article confirm that? In fact doesn't it confirm it with a vengence when it says (near the bottom)...
"The center tap is achieved by a bifilar winding, a pair of wires wound physically in parallel, but wired in series. The north-south poles of a phase swap polarity when the phase drive current is reversed."
Granted, this is for a six wire (unipolar driven) motor, but an eight wire motor is just where the centre tap is brought out to the outside in two wires instead of one.
Yes, more voltage gives more top speed but if the motor does not develop enough torque to overcome detent torque and friction, you will never get there. So more amps to get enough torque.
Yes, heating effects, from eddy current losses and copper losses, limit how many amps you can feed the motor or it will get too hot and burn out the insulation.
We both agree that heating is the issue, but we were talking about our different conclusions regards the rated values. If you could treat a parallel wired pair of coils as totally separate entities then you would be correct with your current calculations but the data sheets suggest otherwise and I'm sorry that seems to have been forgotten. Anyway, I've done my best to explain why I think this is - again maybe I'm wrong. At the very least the fact that the root2 rule and the I^2.R calculation works every time in complying with the datasheets suggests to me at least that it's no coincidence or momentary lapse of reason that the rated values are as they are.
Something for you to think about. A stepper turning at 1000 rpm is being fed by an 8 microstep driver at the rate of 200 x 8 x 1000 = 1,600,000 square wave pulses per minute i.e. about 27kHz.
LOL - like we havn't given ourselves enough to think about already. But for the fun of it I'll play.
You could improve your motor by having 2 rather than one phase - 13.5KHz.
You could half your frequency by swinging positive/negative (bi-polar driver) 7.75KHz.
You're said you're using microstepping so your signal is nearer a sine wave than a square wave, but more to the point, half of that sine wave spans a whole step and is what actually manifests the microsteps - 3.87Khz.
And now nurse is coming with the jacket with no sleeves, so I have to hide. But before I go - group hug anyone? ;D
EDIT: Yes I've just reread this and "notches" is probably no good either because then we'll become confused as to whether we're talking about the notches on the rotor or the stator - poles it is - I give up!
Ian
-
Jim,
Stick with it you have already made great progress with your lathe leadscrew. I have an 8 lead stepper rated at 2.5 Amps per phase and a Gecko 203V. This will eventually be driven with a 50 Volt power supply and I will post the values of torque and maximum speed obtained with the 3 different wire connections as I did in this thread.
http://www.machsupport.com/forum/index.php/topic,4764.0.html.
Ian,
It has been a very fruitfull discussion, we must do it again sometime. I have learned a lot and had some of my assumptions questioned and answered. By the number of views, we seemed to have provided a few people with some food for thought.
Hood,
The thought of taming a PID loop on a servo scares me! Please keep us informed of your progress and perhaps I might gain some understanding of the problems.
Ian (Heading for a wee dram of Ardbeg :))
-
HA ha for a guy that didnt have time to do a summing up I think you did a fairly good job of it :)
Servos seem to be easy enough, I have 10NM AC servos on my lathe and they were a dawdle to set up, might find I am needing a bit of fine tuning once I get the SmoothStepper hooked up to it as it will allow me to release the full potential of them. The motor I am working on at the moment is hopefully going to be for the spindle of the mill which is being retro'd at the moment. I have AC Servos on the axis and I got a 5.5KW (20NM continuous) servo from eBay, now have a drive for it and just got a small motor that I have robbed the encoder off. All I need to do now is mount the encoder and line up the Hall signals and hopefully it will all come together and I can rip off the induction motor thats presently on there. One of the advantages of the AC Servo is it has constant torque from 0 to 4000RPM (in this case) so should be ideal for the spindle, also it will allow me to think about making a tool changer and rigid tapping may well be a possibilty.
Well hope you enjoyed your wee dram and thanks to you and the other Ian for the info and also Jim for starting the thread and posting his findings.
Hood
-
I've got to thank you guys for all the input to this post - which I will admit, I never thought would get this far.
Yes - I am a bright chap and can understand all that has been said, and realise it is far more complicated that I have made out.
I think the post ought to close now. In the end my system will work and be reasonably efficient and fast enough for me.
BUT
The main thing I was worried about are the vast number of motors and drives that are kicking about our hobby - and the "starters" or "newbies", with little or no experience, have no idea what to use.
It would have been nice to get to a point where we could say - yes - for this motor, use this drive set at this current, at that voltage.
I suppose that is a little pie in the sky.
Again, thanks again - Ian - don't bash your head against the wall - I'm not that bad. :o
-
Well done guys - a most educational thread.