Gearing down...

[snooper]

Hi all!

Just thought I'd ask: Why are most assemblies I've seen that use gears, belts & pulleys and chains & sprockets, using gear reduction? With my limited knowledge I find it curious, as in most other instances the stepper is coupled directly to the screw. Is there a disadvantage to going 1:1?

Also, can someone please tell me which is better: belt & pulley or chain & sprocket?

[Mariss Freimanis]

Power (the stuff that gets things done) is torque times RPM. It is wise to pick the point where the RPM times torque is at a maximum. That speed point oftentimes is much higher than the intended load's RPM so "gears, belts & pulleys and chains & sprockets" match the motor to the load.

Your'e in a sports car with a manual 5-speed gearbox. Do you want to be in 1st gear or 5th gear when the light turns green? Same thing.

Mariss

[mxtras]

I, too agree that gearing provides for a severe advantage over direct.

I prefer timing components - they are lighter and have far less (if any) slop than a chain.

Scott

[JerryFlyGuy]

Gearing is totally dependant on the application, if your building a mill, your going to want more Lb-force and accuracy, while speed and acceleration are less of a concern. However, if your building a plasma, the speed and acceleration are your primary concerns while the accuracy is less of a primary concern and Lb-force for cutting is nill, your still going to need some level of force to get the acceleration rates up high, but not for sustained cutting. Gearing is used to take a motor and modify the output to fit the level of performance you want. Therefore a Mill will typically have more reduction than a plasma cutter would. You basically need to sit down and figure out what you want and via some easy math figure out what level of reduction you need.

Jerry [using a 10:1 reduction w/ 2" pinion on my X axis..]

[NC Cams]

SIt down and FIGURE IT OUT??? Are you kidding??? (TIC)

I thought newbies were suposed to go to CNC ZOne.COM and ask somebody what to use for a motor/gear via an empirical SWAG.

Besides SWAG"s are much more entertaining.

[JerryFlyGuy]

ahhh.. SWAG's work great.. I garrentee, if you take a HMS [Huge mother Stepper ] over 2500 oz-in w/a min of 2000 steps/rev & w/ zero reduction and a HMD [Huge mother drive] and a HMPS [Huge mother Power Supply], you can use it on any ball screw, rack/pinion/ machine.. for either plasma or mill's or routers that are all home-built and you won't run out of power if you keep all feeds and speeds below 10ipm.. Acceleration may need to be played w/ to get the proper setting to not lose steps.

I GARRENTEEEE IT.. [and if I'm wrong.. oh well.. this Garrentee is worth as much as the paper I wrote it on..]

Hows that for a SWAG?

Jerry [Looking for some of those HM** products for my own machine.. , right after I figure out how to make this 125,000RPM home-built Gas turbine run for more than 3seconds on roller skate bearings..]

[Mariss Freimanis]

Here's a dynamometer speed-torque and speed-power readout I took for a NEMA-23 motor at 24VDC and 40VDC at 4A per phase. Note that power output ramps up with speed, then levels out. At 40VDC the power output ramps at the slope but reaches a higher level before leveling out.

I have marked in red the optimum speed the motor should be run at for 24VDC and for 40VDC. There is more power to be had above those speeds but it takes longer to get there.

The speeds are what the motor should run at when your mechanism is moving at its maximum design speed. Gear accordingly.

Mariss

[JerryFlyGuy]

Mariss, you should start a service of running Dealers motor's and publishing
their results for them, an un-bias, type thing.. Anyone who wants to get their motor tested can sent 1 motor and $150 to you and you get to keep the motor and publish the results for them.

This type of info is VERY invaluable! I wish I'd had this type of graphing for when I was specing my motor's!

Jerry [I wonder what the optimum speeds are for my motors...]

Edit: Can you explain a bit more why a person shouldn't just run their motor's at their torque peak? --End Edit..

[Mariss Freimanis]

Edit: Can you explain a bit more why a person shouldn't just run their motor's at their torque peak? --End Edit..

A rusted-in bolt has a peak torque of many 10s of thousands in-oz yet you don't think of it as much of a motor. Why? because its associated speed is zero.

The same here. Fundemental points:

1) It is power, not torque or speed, that gets work done. Power is Watts, HP, etc. It is not in-oz or RPM.

2) Power is torque times speed.

3) Peak power may not occur at peak torque. Peak power is when torque times RPM gives the biggest number.

4) You buy a motor to give you power. Run the motor where (RPM) it delivers maximum power. Anything above or below that speed and you are cheating yourself.

-------------------------

The dynamometer I built is set up to give me both speed-torque curves and speed-power curves. The speed-power curve is far more valuable because it shows where a particular motor driven with a particular power supply voltage should be biased-at to deliver maximum power (biased-at means at what RPM).

The dynamometer is not especially hard to build. The dyno load is a NEMA-34 servomotor (Kt = 53.2 in-oz/Amp) that is loaded by a seriously lobotomized G320 which acts as a transconductance amplifier in its zombie state. "Transconductance amplifier" simply means an analog control voltage input is proportional to a torque output from the motor. This torque output loads the test motor.

The servomotor is equipped with an encoder (500-line US Digital). A G100 generates the step pulses to run the test motor, outputs via one of its analog outputs a torque command to said lobotomized G320 and reads back the encoder.

An assembly routine I wrote for the G100 runs the motor up to 3,000 RPM, then applies an increasing torque load until the G100 lead-lag register shows a 1.8 degree test motor lag (the very point of impending stall). That 8-bit analog torque load value is sent back to the PC. The speed is then dropped (decelerated) by 100 RPM and the process repeats until zero speed is reached. It takes about 10 seconds to generate a 30-coordinate graph of the motor's speed-torque curve. The dyno load servomotor never gets a chance to overheat and skew the test results.

The motor speed-power curve is calculated by multiplying the 30 point measured torque table by RPM and then dividing it by a normalizing constant to get the answer in Watts. Simple computer programming.

The PITA thing is to enter these 30 coordinates into ACAD to cough-up a genuine graph. That's why the graph is for every 100 RPM from zero to 1,000 RPM , 500 RPM above 1,500 RPM.

Otherwise the dynamometer is very simple. Four 4" long threaded standoffs to couple the motors (dyno and test) face to face via the nema mounting flange holes and a helical coupling between the motor shafts. The rest is computer stuff.

Mariss

[JerryFlyGuy]

Well Mariss, I follow you for the most part, until you get to the computer stuff, and I understand whats happening at a base level there also. But I doubt I'd be able w/ my current knowledge level, be able to duplicate you installation. It's very interesting however.. I think via my torque graph's I can sit down and figure out how much power my steppers are going to deliver at what RPM, now that I better understand stepper power.

Thanks!

Jerry