Are the Zapp 3NM steppers good with their PM-752 drivers?

[Jonathan]

Hi,
A bit of background first...
The motors I currently have on my milling machine are only 1Nm, with a Unipolar driver which I've just sold to raise money for the new stuff. These motors are mounted to the side of each screw and connected via timing belts etc to give a ratio of about 1:2 (I tried various ratios). With this setup the maximum feedrate I could get was 1.4mm/s, and that was pushing it. This is with the standard 2.5mm pitch ACME screws that came with the machine.

Anyway, I've finally decided to buy 3 (or 4) new stepper motors and drivers from Zap so that I can get a higher feedrate. I was about to buy them yesterday but thought I'd just check first:

PM752

SY60STH88-3008BF

I'll be running them at 72V. I'll probably experiment a bit with the pulleys to see what works best.
My question is what feedrate can I realistically expect to get with this setup (without ballscrews), and are those motors/drivers a good combination? I'd like maybe 25mm/s.

Might these 3.5Nm motors be better? I'm guessing not as Nema 34 are generally slower, being bigger?

http://www.arceurotrade.co.uk/Catalogue/Stepper-Motors/Stepper-Motors
(I live just a couple of miles away from Arc-Euro so those would be pretty convenient.)

Here's a link to me using the machine to turn something, just to show which milling machine I have (skip to the end of the video):

http://www.youtube.com/user/Jonblissett#p/u/1/4wGDMSiZpyg

Thanks in advance...

[irving2008]

for 25mm/s (1500mm/min) you need to spin those motors at 300rpm. At 72v those motors have a corner speed of 1023rpm so you will be near optimal as regard their torque curve, indeed you might be better off with 1:1 gearing (motors running at 600rpm). Assuming a 20mm screw then you should get pretty close to your goal.

The 3.5Nm motors have a higher inductance and are more marginal than the 3Nm ones at the speed you require with 1:1 gearing... stick with the 3Nm

[Gary]

We sell the PM752 with the SY60 and they work well together.
Also 72V is a bit too high for the PM752 driver, it will work, but is a bit close to the max voltage.
At a high voltage the Arc motor should perform well, but will not spin as fast as the Sy60, but will have a lot more torque at the slower speeds.
I would also point out that the datasheet is wrong, and if connected in parallel the current would be 6A not 4.2.
It is odd that they say the motor is rated as 4.2A in parallel because parallel current would be double the series current.
All i can think of is that the motor has a problem running at 6A, so if you decide to use this motor, i would ask the question to them.

[irving2008]

Those 3Nm motors you sell Gary will do fine for this application at 60 - 65v with no meaningful loss of torque as even at 60v the corner speed is 850rpm so he'll still be in the flat region... in fact it would still just be ok at 48v but I wouldnt go any lower than that..

[Jonathan]

for 25mm/s (1500mm/min) you need to spin those motors at 300rpm. At 72v those motors have a corner speed of 1023rpm so you will be near optimal as regard their torque curve, indeed you might be better off with 1:1 gearing (motors running at 600rpm). Assuming a 20mm screw then you should get pretty close to your goal.

The 3.5Nm motors have a higher inductance and are more marginal than the 3Nm ones at the speed you require with 1:1 gearing... stick with the 3Nm

Thanks for the fast replies - great to get a decisive response!
How did you work out the corner speed of the motors?
When I said 1:2 earlier I meant smaller pulley on the motor to double the torque...the 1Nm wouldn't move without.

The screws are about 25mm I recall, though that's a guess...
Based on a corner speed of 1023rpm, and 1:1, would if be reasonable to expect 40mm/s for traversing? I guess this is difficult to say since there are so may factors.

If 72v is a bit high what would you advise? 60v?
I'm still not sure what to use for the power supply, rewinding a microwave transformer I've got seems an attractive (i.e. cheap) option, so does putting 5 or 6 computer power supplies in series, isolated, for the same reason.

If I buy the motors and drivers tomorrow how soon will you be able to post them. I ask because your website says the PM-752 isn't in stock?

Thanks again for the help...

[Jonathan]

Those 3Nm motors you sell Gary will do fine for this application at 60 - 65v with no meaningful loss of torque as even at 60v the corner speed is 850rpm so he'll still be in the flat region... in fact it would still just be ok at 48v but I wouldnt go any lower than that..

Didn't see this before I posted. I'll probably use 65V then. If I use the microwave transformer it shouldn't be hard too wind it accordingly...

[Gary]

The PM752 is running low, but have about 15 in stock and another 100 cominig in in about 2 weeks.

[Jonathan]

I've been waiting for money to transfer to my bank account from paypal. I didn't pay with paypal directly as I don't think it's fair to make a website pay paypal fees when I can pay using a debit card. Now I've just found that the stepper drivers have gone up from £44 to £50 today :( Now I've not got enough pocket money...

How much difference will getting the PM542 and running the lower voltage make?

[Gary]

The price has gone up because of the exchange rate, out of my control im affraid.
have a look at the torque curve in the datasheet, and this will give you an indication of the difference in power.

[Jonathan]

My mum lent me £20 so I was able to buy them on Friday :).
Order was dispatched on Monday and arrived yesterday - excellent service, thanks!

Been doing some testing...
I've not got round to making pulleys for a 1:1 ratio, so for the moment the closest I can get is 22T on motor, and 42T on screw.
The motor stalls above 1250rpm (no load), yet it still has enough torque to drive the X-axis at 25mm/s (1150rpm). :cool: Trying higher feed-rates than this gets a bit dicey.
I think this means that if I put a 1:1 ratio, then above 25mm/s is almost certainly attainable since the motor will only have to run at 600rpm to get the 25mm/s. Having said this the current speed is pretty scary! I had to put a lot of tension in on the belt to stop it slipping.

Currently I'm using 1/4 microstepping. I've tried 1/8 but it doesn't seem to make much difference. What do you recommend I use? The computer I'm using is only 1.4ghz, this might be a limiting factor? I'm using Mach3 at the moment, I intend to change to EMC.
Also, I've set the port frequency setting in Mach3 to 25Khz. I tried 35Khz but it seemed worse, should one be better than the other? I've set the pulse width to 5us, is there any reason to change this? I guess not unless I start using 10000pulse/rev...

Thanks for all the advice so far.

[Jonathan]

Oh one thing I forgot to mention...I measured the current into one driver (using both an ammeter and a shunt resistor) and when idle it's about 0.3 amps. It goes up a bit when the motors running, but either way that seems a bit low to me? I've got the driver set to 4.09A peak, I take it this is correct for these motors in bipolar parallel? The motor is only a little warm, nowhere near 80°C.

I'm currently using a 225VA toroid for just the one motor. There's no perceivable temperature rise on it so maybe I can run more than one motor from this transformer?

[irving2008]

you need 66% of the total motor peak current... so 3 steppers at 4A/phase = 3 * 4 *.66 = 7.9A. @ 60v thats 420VA

remember that measuring phase current with an analogue device will show the RMS phase current. With a digital meter you are unlikely to get a useful reading because its highly spikey.

[Jonathan]

you need 66% of the total motor peak current... so 3 steppers at 4A/phase = 3 * 4 *.66 = 7.9A. @ 60v thats 420VA

remember that measuring phase current with an analogue device will show the RMS phase current. With a digital meter you are unlikely to get a useful reading because its highly spikey.

I thought it was 8 amps per motor, since there's 2 phases?
Anyway, I got lucky and won a 500VA toroidal transformer for £20 on eBay so I'll be using that.

I changed the pulleys round (had to bore them out to the right diameter and stuff) to get 1:1. Now I'm getting 42mm/s on the X axis :)
Bad news is the Y-axis for some reason refuses to even go at 25mm/s without stalling the motor. Not really sure why, maybe something to do with the belt tension and how tight the nut is...
Not tested the Z yet.

[irving2008]

I thought it was 8 amps per motor, since there's 2 phases?
Anyway, I got lucky and won a 500VA toroidal transformer for £20 on eBay so I'll be using that.

I changed the pulleys round (had to bore them out to the right diameter and stuff) to get 1:1. Now I'm getting 42mm/s on the X axis :)
Bad news is the Y-axis for some reason refuses to even go at 25mm/s without stalling the motor. Not really sure why, maybe something to do with the belt tension and how tight the nut is...
Not tested the Z yet.No, thats peak current. In general the 2 phases are never both powered fully all the time so a sort averaging occurs and anecdotally (although Mariss of Gecko fame did show the maths once) its 66%!

Its usual to quote feed speeds in mm/min for milling - so 2520mm/min is good, 1500mm/min isnt bad tho

Strange... if you haven't already, try swapping driver channels to validate its not a driver issue, then motors (assuming they are the same) to validate its not a motor issue

[Jonathan]

(although Mariss of Gecko fame did show the maths once) its 66%!

Thanks for the clarification. Do you by any chance have a link to the 'maths' behind this? I'd be interested to see..



Strange... if you haven't already, try swapping driver channels to validate its not a driver issue, then motors (assuming they are the same) to validate its not a motor issue

(I quoted the feeds in mm/s since personally I think it's better to work with smaller numbers and easier to visualize, but I'll endeavor to stick with the convention from now on)

I've only wired up one driver so far, so that can't be the issue. I'll try swapping the motors though. All the motors are the 3N-M ones.
The handle for the Y axis (before I removed it) did seem a little harder to turn than the X-axis so I may try loosening the nut. I tightened it to reduce the backlash. Annoyingly to get to the nut you have to take the bed off the machine and all sorts...takes ages :(

[irving2008]

Surely its the gibs that determine how tight the slide is? although I'll accept endfloat is another contender

[Jonathan]

Yes that's true, I'll check the gib strip. There's no end-float on the screw since there's thrust bearings, unless you're referring to something else?
The nut is slotted with a bolt to tighten the two parts enabling you to adjust the backlash. Quite recently I tightened it to reduce the backlash, but maybe I tightened it too much and increased the friction excessively.

[irving2008]

AH, I was confused by the statement


The handle for the Y axis (before I removed it) did seem a little harder to turn than the X-axis so I may try loosening the nut. I tightened it to reduce the backlash many people use the term backlash when they mean endfloat and i thought the nut you refered to was holding thr handle in place...

[Jonathan]

I've had some luck with the Y axis now. I loosened the gib strip which helped a bit however it strikes me as being too loose now. I adjusted the backlash adjustment and now I've got 2300mm/min on the Y axis, but about 0.2mm of backlash instead of 0.065mm.

Good news with the Z, 4600mm/min with 1:1 pulley!!!! It will do more if I change the motor to half stepping because the computer is limiting the pulse rate I can use at 1/4 microstep, however 4600 is plenty fast enough for an axis that's only got 92mm travel. I'll keep it at this ratio to preserve torque for drilling etc

So to summarise:
X, 2520
Y, 2300
Z, 4600

Overall I'm very happy with this - can't wait to start machining things again! Thanks for all the help that's enabled me to do this...

Also, could someone point me to a simple program to import a 2D .dxf and generate g code for '2.5D' milling from it?

[Vic66]

Try cambam, theres a free version and a pay version http://www.cambam.co.uk/