[AJ_Mac2001]

I am jumping into the deep end with servo's and wanted to verify what I think I have absorbed in the forums.

Is this true: voltage is proportional to rpm, while amperage is proportional to delivered torque.

As to what voltage can be applied to a servo, I've read various posts suggested you can apply higher voltage than the servo's plate spec. Is there a hard and fast rule there? I am going to get 48V servo's... what is the max voltage I could continously supply, versus peak max?

Is there a way to physically test this (with a multimeter?) without doing damage. And incidentally, what damage occurs if you do go too high on the V's? [does smoke appear??]

Thanks!

[Al_The_Man]

You should first indicate wether you are talking Steppers or DC/AC servo's. As the technique's to get the most out of them are quite different.
Al

[AJ_Mac2001]

I'm referring to DC Servo's

[Al_The_Man]

I personally have found no advantage to supplying greater than the plate voltage to dc servo's, the most important thing to avoid smoke is not to exceed the max current rating of the motor, most amplifiers today have a current limit setting that should be set for the motor max value. In a CNC application with encoder feedback , use an amplifier that can be set to torque (current) mode, this allows the the use of DC servo's with no tach feedback (or just do not use the tach). The current mode produces a torque output from the motor proportional to the input reference signal.The advantage is that a movement of the motor shaft from the desired position causes a large correcting torque.
One thing to keep in mind with these and any AC/DC servo set up is that most amplifiers, unless they are of the older scr type, use a large internal or external DC power supply, which retains a certain amount of charge after the AC input supply has been removed e.g. in a e-stop condition, the retained charge allows the motors to move until the charge decays, so as well as removing the AC supply it is important to also disable the amps, most have an enable input that can be switched in a e-stop condition.
Al

[High Seas]

Al_The_Man
Good info on the servo and upping the voltage and current. Could you share similar information for doing STEPPER MOTORS too?
Thanks and :cheers: Jim

[Al_The_Man]

Jim, Cheers to you to! I probabally am not the one to give advice on steppers as I prefer DC/AC servo's. but I have dabbled a bit with steppers and the thing with steppers, as has been related here in quite a few posts, one of the down sides is that normally the torque decreases as rpm increases, This I believe occurs due to the impedance (AC resistance, if you will) increasing with rpm, sophisticated controllers will increase the voltage output as rpm increases, as a means to increase the torque at high rpm.
Others here with more knowledge of steppers may be able to step in and offer you some more insight as to what is out there.
Al

[AJ_Mac2001]

Thanks Jim for the info!

Now just to dumb this down a little to make sure I understand it. Too much amperage can damage the motors, but voltage is safe (er) to play with? I'll be going with Gecko's so I assume that will limit me to 20Amps from my PS (is that correct?). As far as volts go if the servo name plate reads 48V - is something going to blow if I put 60-70V to it?

The net result of this is that I want to squeeze more RPMs out of servo's that are spec'd to be 48V@1000rpm. I'm wondering how much more speed I can get (safely) without blowing the whole thing up.

Thanks

[IndHobby]

Since you said 1000 rpm and servo together I’m guessing you have Ametek brush servos. Those are low RPM servos that are designed to run at the slower RPM, the advantage is you do not need to reduce them to get a good working torque.

You’ll gain nothing by over-volting them, maybe a few more RPM’s but nothing worth the extra heat the motor will generate. The extra heat kills the permanent magnets pretty quickly, kill the magnets and it’s game over.

There is 2 ways to cook a servo,
Slowly- which basically kills the magnets.
Quickly-which means you smoke the winding, by the time you smell the smoke it’s already done. Also the case will be HOT but by then the armature is cooked.

Also, when getting the power supply.
Get an unregulated one. Not switching or regulated.
Go for 48V as is the motor plate
Current capacity will be the motors constant current of all the motors added up.

This little ditty may help in the future.

Steppers like it HOT, servos like it COLD,
Servos like to RUN, steppers like to HOLD.

If you want the quick rundown feel free to give me a call at 207-637-2523 anytime before 10PM EST. I won't try to sell anything


Thanks

[AJ_Mac2001]

Thanks Aaron for the insight... not to mention the little ditty

The servo's in question are 1100oz/in hathaways. The 1000rpm suits my final design well (ie. when I'm using 5tpi screws) but for a start I've gone cheap and used 1/2"-13tpi rod. So I'm trying to see if I can squeak out some extra speed by spinning the motors more quickly.

As an aside: The threaded rod is 6' long, supported by double bearings on each end, and slightly tensioned. Each double bearing supports the shaft over a 6" span, and really work wonders at reducing whip/vibration. With my 2400 rpm drill I can almost max it out with little/no vibration.


I may take you up on your phone offer (Thanks) once I've done a little more research on ebay power supplies so I don't sound like a complete ignoramus.

-AJ



Ciao

[IndHobby]

Threaded rod will more than likely cook the servo. Threaded rod is designed to bind, after a few hours the zink coating will be gone and the torque to turn the screw will go way up.

If you're going to go with leadscrew just go to use-enco.com they always have it on sale and it's pretty low priced.

It will save you some servos.

Feel free to call a big part of my job is not to make people feel like a complete ignoramus.

Industrial Hobbies...Not Industrial Headaches

Thanks

[Mariss Freimanis]

Hi,

I'm the guy that designed the G320s. Aaron is pretty much correct. If you averse to risk (as I am), run the motor at the rated voltage. You will be guaranteed a stable and reliable system. The official line is mfgs. rate their motors to certain values for a reason, respect it.

The unofficial line is marketing has a lot to do with those ratings.

3 things determine the maximum voltage of a brush motor. In order of significance they are:

1) Commutator arcing: At rated speeds, (voltages) there will be very little commutator arcing. As speed increases, this arcing will increase slightly until the arc-over speed is reached. This happens fairly abrubtly, when it does though, neither the motor or the drive have long to live. Brushes slide over the commutator much like the tires of a car stay in contact with a bumpy road. Above a certain speed that no longer is true. Your car goes airborne like something out of a Clint Eastwood Dirty Harry movie and so do the brushes. That is when arc-over happens. It kills the motor the same way it would kill your car.

2) Brush life: Brushes should last between 2,000 to 10,000 hours (1 to 5 years at 8 hrs a day, 5 days a week). The faster you go the faster the brushes wear. Once they start arcing, the arcing process itself drasticaly reduces brush life by electrically eroding them. Think about it: those pretty blue sparks have to be working on something to make themselves seen. If you can live with 100 to 500 hr brush life, double the rated speed of the motor and put a good load on it.

3) Armature destruction: This is really hard to do. A motor designed to turn a 1,000 RPM may have its windings torn off by centrifugal forces at much higher speeds. It usually takes at least a 10 times overspeed to do that (running a 12V motor at 120V). Usually arc-over will bring things to a smoking halt before that speed is ever reached.

Let's move on to 2 other things:

Paradoxically, using a much higher supply voltage than the motor is rated for is not a problem if, (a) you don't ask the motor to go any faster than it's rated for, and (b), you have set the current limit to a safe (rated) value for that motor. The G320 is a PWM switching type drive. The voltage the motor "sees" is the averaged PWM value (modulation % times the supply voltage).

This means if you are asking the motor to turn at its rated speed, the voltage the motor will "see" will be its rated voltage, not the much higher supply voltage.

The second part of this is motor heating. Motor heating is entirely dependent on torque load applied, not the speed the motor is turning at. This means a motor turning at 150% of its rated speed will get no hotter than a motor turning at rated speed. Heat = torque squared. Nothing else.

2) I don't think you don't have an 1100 in-oz motor. That more than likely is the BIG number people selling things want to impress you with. There is a much smaller number (10 to 20%) that is the maximum continuous rated torque. It will be between 110 to 220 in-oz. That is the number you have to pay attention to.

Your motor is 1,000 RPM at 48VDC. That means its Kv is 48V/KRPM. Kt is always Kv times 1.351, so your Kt is about 64.8 in-oz per amp. To get 1100 in-oz the motor would have to draw 16.97 Amps (1100 / 64.8 ). The force (in lbs) applied to your 13 turns per inch (TPI) screw would be: pi * 13 * 1,100 / 8 or 5,613 lbs. (2 1/2 tons). That would be enough to turn your screw into a pretzel.

Unless I'm missing something, your actual working torque won't exceed 220 in-oz.

Mariss

[AJ_Mac2001]

Thanks Aaron for the note on the limited life of threaded rod. My machine evolution will lead to better parts! Does anyone else out there have experience with threaded rod binding... or a life estimate for it? I have to make do with it for now until I can afford to move up.

Mariss, Thanks for the detailed explanation. Instead of going airborne like Dirty Harry, I think I'll go for the "Driving miss Daisy" approach and take it easy. Eventually I'll use appropriately spec'd leadscrew to get to the speeds I want.

I don't really know what my "1100's" true continuous torque is... but judging from the experience of another Zoner with the same serovs they should pack enough punch. He had them coupled with radiator hose and they had enough torque to shred it when the gantry hit a hard stop. I gather from the seller they should draw around 5A so I guess that puts it on what you said they'd rate "continuously". 5x65~325oz/in

Here's to making chips, not pretzels

[HuFlungDung]

AJ,

You'll just have to figure a way to lubricate the threads, in the midst of all the cutting dust The zinc coating is not meant to increase the life of the threaded rod anyways, it is merely for corrosion resistance.

Use a long nut, to help distribute the load over more of the threads. Monitor the temperature of your motors.

[AJ_Mac2001]

Roger on all that HuFlungDung.

All the nuts are tapped pieces of Delrin ... 2" long, so hopefully that is easier on the threads. Would you recommend periodic WD-40 sessions... or something else??

Of course with my dust collector sytem, this will be the first dustless CNC (ha ha - cough cough, choke choke).

I'm sure I'll be babying it along and keeping a hand on everything in the beginning (Sophisticated temperature monitoring!) - but the meltdown will come the moment I figure everything is OK and turn away to answer the phone.

Cheers,
AJ

[HuFlungDung]

Hmm, what to recommend. Maybe some kind of spray on "dry lube" with graphite and/or teflon? Plastic nuts should not require much lubrication.

Some threaded rod is a bit rough to start out with, and might even be rough enough to cause excessive wear on the nuts. This is just a guess, though. Before putting the machine into service, a person could lap the threads in a bit, using a plain steel nut and some abrasive paste. Run the nut up and down the screw a few times, using something light duty like your portable drill to drive the screw for this hand work. Wash all the abrasive off thoroughly before using the plastic nuts. Don't introduce abrasive into the plastic nuts though, because it will just stick inside and wear your screws out very quickly.

[AJ_Mac2001]

I'll give that a whirl. As always, thanks for all the advice folks.
AJ