PDA

View Full Version : bumping holding current - how it affects torque?



dsc
02-02-2014, 04:06 PM
I'm using a Nema 34 8.0Nm motor to drive a spindle via pulleys, the ratio is 2.66:1 (18th on the motor and 48 on the shaft), speed is 240RPM. Torque curve for the motor is like this:

http://www.americanmotiontech.com/Upload/TorqueCurve/86HS80_EC_1000.jpg

I'm running it at 68VDC from an SPS705 (5A) leadshine unregulated power supply. Motor is the 86HS80EC, with a HBS86 driver, running in a closed loop control (don't really need the closed loop at all).

I'm struggling with stalling in certain situations and was wondering whether bumping the current from 40% holding torque to say 80% would have a big effect on torque? the above torque curve is for 40% holding current, I've tried getting torque curves for high current percentages, but leadshine didn't have the required data.

It's been suggested (thanks Jazz!) to use higher voltage to get more umph from the motor, but sadly there's no space in the design for an AC powered driver like the Kinco 2280, so the best I can use is the AM882 + a bigger Nema 34 12Nm motor, but I'd like to experiment with what I already have before coughing up more cash.

Regards,
T.

EddyCurrent
02-02-2014, 04:42 PM
AM882 manual says this, have you tried using ProTuner ?

Standstill current setting
SW4 is used for this purpose. OFF meaning that the standstill current is software configured, and ON
meaning that standstill current is set to be the same as the selected dynamic current.
By default, the current automatically reduced to 60% of the selected dynamic current two second
after the last pulse. Theoretically, this will reduce motor heating to 36% (due to P=I2*R) of the
original value. Reduction rate and idle time can be configured in the PC software ProTuner. See
more information in section 13

8. Select configuration for different motor
There is a rotation switch to select the built-in configuration parameters for
different motor. User can select the corresponding parameter according the motor
part number. When it is 0, user can tune the parameters in the configuration
software or just activate the auto-configuration by SW4.

dsc
03-02-2014, 12:00 PM
Cheers Eddy, reading through the manuals, I think bumping the holding current won't change the torque in normal operation. There's a separate running current set to 100% and holding current tab set to 40%. I'm currently looking at upgrading to a bigger motor, which is an 8 wire that will be wired in series to get more torque at low speeds.

Regards,
T.

Jonathan
03-02-2014, 01:06 PM
Cheers Eddy, reading through the manuals, I think bumping the holding current won't change the torque in normal operation.

That's correct.

Is 240rpm the required motor speed, or shaft speed? From other threads I think it's the latter, so with the ratio you've chosen that makes the motor go at about 2.66*240=638rpm. According to the torque speed curve you've posted, the motor will only output about 2.8Nm (187 Watts) at that speed.

What I would do it make a copy of that graph in a speadsheet, by just reading off a few values and plotting them and use this to plot power vs speed. Pick the point with the highest power (which incidentally is around the corner speed) and choose your drive ratio to match that speed. To convert the torque-speed graph to power you'll need the following formula:

Power=Torque*rpm*\frac{2\pi}{60}

If the power you get using that approach is still not sufficient, that's the time to consider raising the motor voltage. Personally, for a coffe gringer I'd go for a generic motor with a planetry gearbox as it should be much more compact, efficient and cheaper.

Just to illustrate why this approach is better, one way that might seem logical is to run the motor at the speed with the highest torque output - i.e about 160rpm. So for instace 240/160=1.5 .. say 30T pulley on motor and 20T on output. That should get 7Nm at 160rpm, so 117W, which is clearly less power than the current setup so not a good way to approach this.

dsc
03-02-2014, 02:31 PM
Thanks for the reply Jonathan. I'm actually quite happy with less speed on the output, which is why I'm currently running the motor at 240RPM and the shaft (output) at 90RPM. At 240RPM there's plenty of torque, around 5.5Nm at 68VDC, this over 2.66:1 ration gives around 15Nm on the output (at 90RPM). This is on the verge of stalling, sometimes it runs fine, sometimes stalls. Max voltage both the HBS86 or the AM882 can handle is 80VDC, would that give enough bump to the torque? Here's the power / torque vs. speed data from Excel:

11500

With a Nema 34 12Nm, which is only 25mm longer than the HBS 8Nm I'm using now (it's got an encoder fitted which I don't need at all), I'm planning to run that at the same speed of 240RPM. This gives around 7.5Nm, which over 2.66:1 ration gives 20Nm on the output (at 90RPM). Hopefully that's enough to get past stalling. Similar data to the 8Nm motor:


11499

I've thought about geared motors, fitted with a planetaty gearbox, but those seem to be rather expensive and overall the cost was similar. I know size is different, but length is comparable and that's the only thing that limits me at this point anyway. With a planetary gearbox I could run with the same pulley size as the gearbox handles the ration, but even for a 10:1 planetary, I'd need a motor which can deliver 2Nm at 900RPM (to stick with the 90RPM on the output and 20Nm torque). This means typical smaller sized motors are out of the question, steppers probably as well, and it leaves me in the AC motor territory, which tends to go huge, bulky and loud. Unless of course I've missed something which is quite possible.

Regards,
T.

Jonathan
03-02-2014, 03:51 PM
Based on your new graphs, the peak power is at about 600rpm, yet you're currently running at 240rpm. So for the 8Nm motor you're currently getting 130W (@240rpm), but if you changed the ratio to 600/90=6.66, then you would have about 175W (@600rpm) available - so 35% more. That corresponds to 18.6Nm on the output shaft, which is a much bigger difference than you're going to get from changing to 80V from 68V. Changing the motor to the 12Nm version and leaving the ratio the same gains about the same percentage, but presumably at much greater expense...


I've thought about geared motors, fitted with a planetaty gearbox, but those seem to be rather expensive
[...]
Unless of course I've missed something which is quite possible.


2nd hand cordless drill motor and gearbox.

Edit: When selecting the motor and gearbox it's sometimes easier to think of things in terms of the required power output. If you need 20Nm at 90rpm, then look for a motor that can output a little over 20*90/60*2pi=188W, then see if a gearbox is available to match the speed requirement. In general higher speed motors are more efficient (and therefore compact) - hence why cordless drills use relatively small DC motors which operate in the 1000's of rpm.

dsc
03-02-2014, 04:01 PM
Ok, I can see where the max power calculation kicks in, the only issue is that running at 600RPM, the ratio stepdown would be massive, with 6.66 you are looking at huge pulleys or simply a 6.67:1 planetary gearbox, which wouldn't fit due to the size of the motor. I will try and look at less powereful motors, maybe some Nema 23s or shorter Nema 34 and match a gearbox to those.

Where did the 37.1Nm figure come from? at 600RPM, you only have 2.7Nm torque delivered by the motor, so over a 6.66:1 ratio, there's 18Nm on the output. Or am I missing something?

Regards,
T.

Jonathan
03-02-2014, 04:11 PM
Yes, the optimum ratio is a bit unwieldy. Not difficult with two belt reduction, if you've got the space. You imply that the requirement is 188W, plus a bit to account for the inefficiency of the gearing. If that's a planetry gearbox then you'll need to add quite a lot, but if it's timing belts then 200W is probably sufficient. You're probably not going to get that much power from a smaller stepper motor, but by all means check.


Where did the 37.1Nm figure come from? at 600RPM, you only have 2.7Nm torque delivered by the motor, so over a 6.66:1 ratio, there's 18Nm on the output. Or am I missing something?

No, you're correct. I made silly mistake and corrected it but you were too quick!