Alpha 7.2 motors or PK299-F4.5A motors?

[richards]

It's time for a quick sanity check before modifying my PRT-Alpha. Basically, I have two choices. (1). Buy the 7.2 Alpha motor upgrade from Shopbot. (2). Install Oriental PK299-F4.5A motors and Gecko G203v stepper drivers.

From the surface, it looks like a no-brainer. Shopbot is offering an upgrade at the world's greatest price, about $250 per motor when the old motors and drivers are returned. Those new Alpha motors with gearbox do more than I would ever expect my PRT-Alpha to do. With those new motors and new 30-tooth spur gears, I would have excellent resolution, with about 0.000655 inches per step. (1.5 spur gear diameter * 3.14 / 7.2 gear ratio / 1000 steps per revolution). Brady and others have reported cut speeds up to 12-ips and jog speeds up to the limit of the gear box. (I'm sorry, but no matter how fast the machine will jog, I feel safer jogging at a maximum of 12 to 15 ips. A geared motor has just too much torque and can cause too much damage if it gets away from you.) The only fly in the ointment is the replacement cost of the stepper and stepper driver. I haven't checked with Shopbot, but Oriental Motor sells the motor and driver for about $2,500 per one-motor set. Stepper motors and stepper drivers, especially those from Oriental Motors seem to last forever, so that's a fear that I might actually outlive.

On the other hand, if I install Oriental Motor's PK299-F4.5A stepper motors and Gecko G203v stepper drivers at about $380 per motor, driver and spur gear, what would I gain and what would I lose? Actually, in my case there would be almost no price difference if I raided my spare parts cabinet. I would end up buying, at the most, three stepper motors and four spur gears. Everything else is on hand. Some would say that I'm cheating by not including the cost of the the G203v drivers and the motors that I have on hand. That's true, but I would end up retaining all four AS911 motors and drivers, which cost $1,100 each if I bought them new from Oriental Motor. So, I think I'm being fair. Secondly, by gearing the motors 3:1 or 3.6:1 with a belt-drive transmission, I would end up with similar per-step resolution of 0.000785 inches per step for the 3:1 gear box or 0.000655 inches per step with the 3.6:1 gear box. The main factor that makes a 3.6:1 gearbox give the same resolution as the 7.2:1 gearbox is the number of steps per shaft revolution. A Gecko driven motor requires 2,000 pulses per revolution and an Oriental Motor Alpha motor can be set to 500, 1,000, 5,000 or 10,000 steps per revolution, with 1,000 being the default. So 3.6 X 2,000 = 7.2 X 1,000.

The real difference is holding torque when an axis is in reduced current mode, which automatically happens about one second after an axis stops. The G203v stepper driver automatically reduces current to about 50% of the normal current. The Alpha stepper driver can be adjusted to give more or less than 50%, depending on the amount of heat that can be tolerated. Now, this is the crux of my dilema. The PK299-F4.5A motor puts out 880 oz*in of holding torque. With a 3:1 gear box, that is multiplied to 3 X 880 or 2,640 oz*in of torque. Even at 50% when the motor is in reduced current mode, the motor has 1,320 oz*in or 82 lb*in of holding torque. The figures are even more impressive with the 3.6:1 gear box.

Cutting speed is about the same. The amber light comes on at 8.5-ips with a 25.6V power supply. If I used a 35V power supply, the amber light would come on at 11.6-ips. The amber light means that the stepper drive has reached full output. The motor will turn much faster than 8.5-ips, but torque drops off rapidly as the speed increases. According to Oriental Motor's torque charts, when the PK299-F4.5A motor is running at 325 RPM, which is the speed needed to move an axis at 8.5-ips with a 3:1 gearbox, the motor is still putting out more than 800 oz*in of torque. Multiply 800 oz*in by 3:1 and that motor cuts with 2,400 oz*in or 150 lb*in of torque. Can anyone say "impressive"? Also, the motor has no trouble jogging at 20-ips, which is too fast for me.

So, my choice is whether to replace the four orignal Oriental Motor Alpha stepper drivers and motors with three PK299-F4.5A motors and 3.x:1 belt-drive gearbox on the X and Y axes, and one PK296A2A-SG3.6 motor on the Z-axis, or whether to buy the upgrade from Shopbot and get four new 7.2:1 motors.

Right now, with the original alpha motors, 20-tooth gears, and 500 steps per revolution, the machine is running very smooth with almost no chatter. Even though resolution is about 0.0063 inches per step or about 10 X worse than either upgrade would give, my cuts look good. What I am seeing is what I call torque induced chatter. When I do have chatter, it usually happens after an axis goes into standby mode. I'm thinking that replacing my current drives and stepper drivers with either the 7.2:1 Alpha upgrade or the PK299-F4.5A motor upgrade would multiply the holding torque during standby by a factor for 2X or more. Does anyone have an opinion?

[evan]

I don't know if this will make a difference to you but you actually only get three new 7.2:1 motors, drives and cables; both of the X axis and the Y axis motors. You keep the Z axis motor and driver you have, and send back the old X and Y motors, drives and cables. All I can say is I am pleased that I made the upgrade to the 7.2:1 drivers and motors, and I did see a definite improvement in cuts and of course speed. I don’t know if the extra torque would improve the cut, especially if speed is not the issue. More power is always tempting… luckily for me I don’t have this decision to make.

[richards]

Even,
Thanks for the heads-up on the number of drives. I just assumed that they offered three drives for about $750 and four drives for about $1,000 with trade-in. Even though I do very little 3-D work, I'm really leaning towards upgrading all the axes.

The main goal is to reduce chatter to the point that it will be a surprise when it happens. After installing 3:1 belt-drives on my PRT-Alpha, most of the chatter disappeared, but, on occassion there is still a little roughness that can't be easily explained. Whether it's worth $800 - $1,000 to chase that smallish amount of chatter away is my main delemma. Sandpaper doesn't cost much and the time it takes to rub off the rough spots doesn't amount to much, but, it breaks the flow when each piece has to be carefully inspected for something that 'might' be there. What would be nice is to have totally consistent results so that if any part of the cut was rough, all parts of the cut would be rough. And, of course, if any part of the cut was smooth, all parts of the cut would be smooth.

[stevem]

Mike, I’m running Gecko drives with geared steppers. The only chatter that I get can be directly traced to the backlash in the gearboxes (.012”), radial runout in the PC router (.008) and the flexibility of the machine frame, especially in the Z axis.

[Brady Watson]

Let me start by saying that I run machines that have Geckos & machines that have the Alpha motors. Hands down, I (that means me personally) find the Alpha components to be superior to the Gecko offerings in every way. Of course there is a wide difference in price. My Gecko machine runs reliably and I haven't had any problems with the current reduction, although I have experienced a few drives that failed, but they were kindly replaced under warranty. The Alpha 7.2:1 with 30T pinions is a real workhorse. 2 weeks ago I aggressively cut several 3D parts non-stop except to swap out material and tooling - 170hrs straight (24/7). I have NEVER experienced the tool going into 'Alpha-mode' with the 7.2 motors. There is so much torque that it is really tough to break the lock on the motors - even when I lean my 210# body into the Y-car to budge it. Pinion wear is greatly reduced. I cut about 75,000 parts before replacing my 30T pinions (1st year). Before this my PRT would eat pinions every 2 or 3 months. For $750 the Alpha upgrade is an absolute no-brainer. The Oriental stuff is among the best out there in my experience. I think that it would cost you more to put together your own Gecko/PRT style setup than it would to just get the upgrade from SB. Either setup is good - I like the Alpha over the Gecko & the price is a sweet deal to boot.

-B

[richards]

Thanks guys for the quick responses. The safest and easiest path would be to get the upgrade from Shopbot. I'll see if I can come up with a test cut that causes consistant chatter on my PRT-Alpha and then have someone that has already installed the upgrade run the part on their machine. If I can do that and if the test comes out as expected, then I'll have to call Shopbot and see if they sell a four-motor upgrade so that I can replace the Z-axis motor also.

Also, there is an apparent signal termination or noise issue when I connect a G203v stepper driver to the PRT-Alpha controller card. Without going into details about the problem, it looks like it might take a little reverse engineering and a lot of head scratching before I get it solved. (So far, I thought that I had it solved two different times, but it popped up again this afternoon, right after I posted.) One thing that I really don't like to do is chase down subtle problems on a production machine. So, the safe and secure path points towards installing the upgrade.

[richards]

I just got the quote from Shopbot for the upgrade of all four motors. Although I'm sure that the total price is still a bargain, it was almost three times more than anticipated ($2,809 vs $1,000), so I'll go the Oriental Motor PK299-F4.5A route when it comes time to change out the motors.

[Brady Watson]

The 7.2s on just the X&Y are adequate. It is doubtful that you will need the extra torque on the Z - keep your existing Z. The Oriental drivers are leaps and bounds better than a Gecko.

-B

[richards]

Brady,
It's a little difficult to know what to do. I've just spent a few hours cutting 12" test circles in MDF with a 1/4-inch upcut spiral, 2-flute cutter at various speeds to try to determine the extent of my chatter problem (with a 3:1 belt-drive gearbox). It's all in an effort to determine whether the 7.2:1 motors could improve the cut.

Here's the program:

JZ, 0.250
MS, 3.0, 0.5
'
' CCW CUT
JZ, 0.050
MZ, 0.0
CP, 12.125, 7.0, 6.5, O, -1, 0, 0, -0.250, 3, 1, 1, 4, 1, 0
JZ, 0.250
'
' FINISH PASS
JZ, 0.050
MZ, 0.0
CP, 12.0, 7.0, 6.5, O, -1, 0, 0, -0.765, 1, 1, 1, 4, 1, 0
JZ, 0.250

On the test, the move speeds were varied from 3-ips to 5-ips to 7-ips. The spindle was run from 12,000 RPM to 13,000 RPM to 14,000 RPM. The X center point was adjusted to get a fresh cut on each circle.

As you can see, the rough cut was ramped down at 0.250 inch per pass with a bottom cleanup pass. The cleanup pass was a single ramped punge with a bottom cleanup pass. The cleanup pass only took 1/16" from the radius of the 12" part.

The chipload calculator showed that 7-ips at 14,000 RPM was the best combination for a chipload of 0.015. I was a little surprised to see that the fastest cut was also the best cut. The 3-ips cut looked like a cut made BEFORE I had installed the 3:1 gearbox - lots of chatter. The 5-ips cut was nearly as good as the 7-ips cut.

What I'm seeing is NOT tooling marks. They are chatter marks. (On Tuesday I ran a roundover cutter too fast with a handheld router on some knock-down tables I needed for Thanksgiving - and got the most beautiful tooling marks spaced every 0.050 inch. It would be a real pleasure to see tooling marks instead of chatter marks.)

Last week I had a decent job cutting 31-1/2 circles from MDF sheets - along with a few small parts per sheet. The total CNC time, including loading, cutting, table clean-off for ten sheets was just over two hours. Sanding took more than eleven hours. Sanding is what I'm trying to minimize.

So, with the clock running out on the upgrade offer, I'm still somewhat undecided.

[Brady Watson]

Personally, I would jump on the upgrade. I cut everything here & if you have good hold-down and the correct bit, edge quality is very good. It takes a $250,000 machine to better what I am getting on my machine. I never listen to what chipload the manufacturer says the bit is supposed to run within - In my experience it is almost ALWAYS wrong. Speed is one consideration, but I always do testing to find out what yields the best cut quality, often on the fly. I rarely care about speed. Customers come to me for quality.

The other thing that very few people do is modify the VR to match the part being cut. Since every job is different, I modify the VR to match the conditions of the file. It isn't hard to do and most will find it necessary to do if they demand the best cut quality and cut a variety of materials and 2D/3D. The settings for cutting thick AL plate are very different from cutting MDF. Just as when cutting a 3D file with a 1/2" ball they are very different than when cutting with a 1/32" or 1/64" ball. Additionally, the Small Circle Definition setting in VU will affect edge quality on circles. I have mine set to .005". Read up on it and see how it works on your machine.

-B