[jsearles]

Just left my first help request and got stellar replies. Here's number 2.

Just finished building my first table. Drivers came in defaulted to 25000 pulses per cycle/revolution. Plus the steppers are belted 2 to 1 to increase output. (I found the belted xyz table part on eBay intact and just added a router).

The problem is that I think the high resolution is slowing down the table a lot. I can't get Mach3 to go above 19 ipm because of such a high resolution (effectively 77,000).

I need to confirm that this would be a problem as I have described it, because I am told that I need 80ipm and above to cut aluminum properly. 19 is a far cry from where I was at 4.1 ipm, but I need to go much higher by reducing the driver preset resolution from 25,000 to say, maybe 5k or 10k.

I originally left it at 25k because it was incredibly smooth. When jogging, I couldn't even hear the steppers. Seems that may not be so good.

Plelase confirm or correct my understanding of how this works.

[louieatienza]

I think more info on your setup would be helpful. Are you using belt reduction in conjuction with rack and pinion? Leadscrews or ballscrews?

Yes there is a tradeoff between resolution and accuracy with steppers. You may have to use a lower microstep setting on your drive. Or you may need a faster computer, or even an external pulse generator like a SmoothStepper. You may have too small a power supply.

It may also be the timing pulleys used for the belt reduction; you could easily swap one pulley to get a different gear ratio. This may require swapping the belts as well.

As for aluminum your feedrate is ultimately determined by the spindle speed. There is no "correct" speed, though aluminum machines nicely at faster speeds, provided your machine can handle it.

I try to achieve a .0027" chipload when cutting aluminum with a 1/4" endmill. This is a tad lower than the recommended chipload (.003"-.006"). At about 12krpm that comes to about 32ipm per flute. In fact I (and a few others here) use single edge spiral-O-flute bits in order to maintain a decent chipload. At 19ipm you'd have to be at about 7000rpm with a single-edge spiral-O, and about 3500 with a 2-flute bit. Once you know what speed to use, it's a matter of figuring out how much depth per pass your machine can handle; I try to acheve 1/2 the bit diameter for pocketing and 1/4 bit diameter for profiling, due to my machine ridgidity (or lack of).

The above is only a guideline, you may have to tweak a little to get it just right. Basically if the aluminum chips are not hot to touch, you're not doing it right. The basic formula is

feedrate = chipload * number of flutes * spindle speed

and you can derive any one by knowing the other three.

[dbsharp]

If your going with something smaller than 1/8th or 1/10th micro stepping your getting meaningless resolution.

[jsearles]

Hey Louie,
u sound like a pro. Your explanation was very helpful because what I am looking for is perspective. I guess I'll learn all of the formulas one day, but my current sense says that I should be using low flute count bits or the one u mentioned above (I'll try to find one tomorrow).

I also need to add some sort of speed control to get down to the 7-12,000 range that I have been hearing. I'll also play around with the depth per pass.

Thanks again.

[jsearles]

dbsharp,

thanks, I'll take that into consideration. I am sure my resolution settings are wacko. Thanks again.

[Walky]

I also need to add some sort of speed control to get down to the 7-12,000 range that I have been hearing. I'll also play around with the depth per pass.

This is perhaps the best router-oriented speed controller out there:
www.SuperPID.com - Super-PID Closed-loop Router Speed Controller

You can find lots of information about it here:
http://www.cnczone.com/forums/diy-cn...ter_speed.html

[Christian Knüll]

It is usually no problem to just try different settings and see what happens.
I'd try 1600 steps per revolution - this should still run very smooth. Resolution is unlikely to be a problem - most stepper motors aren't able to resolve 25000 microsteps properly anyway.

[louieatienza]

The only potential problem I see from the higher microstep setting (other than speed) is the potential loss of torque... unless you have a drivev like a GeckoDrive that morphs to full step at speed.....

[louieatienza]

Hey Louie,
u sound like a pro. Your explanation was very helpful because what I am looking for is perspective. I guess I'll learn all of the formulas one day, but my current sense says that I should be using low flute count bits or the one u mentioned above (I'll try to find one tomorrow).

I also need to add some sort of speed control to get down to the 7-12,000 range that I have been hearing. I'll also play around with the depth per pass.

Thanks again.

Well, I'm not a pro, though I do have a few clients that I machine parts for on a regular basis. One client has me machine 1/4" to 1/2" aluminum plate, and I've learned a lot the past couple years both by trial and error, and picking the brains of those here and around the Web. Another client routinely sends me 3/8" and 1/2" Garolite to machine; this is the fiberglass reinforced fire retardant stuff that is pretty brutal on tooling. And yet another client has me cutting guitar parts out.

The SuperPID (which I do own as well) is probably the best investment you can make for precise router speed control, with the benefit of having considerably more power at lower speeds than the stock speed control. You'll also appreciate the quiter operation; my ShopVvac easily drowns out the router.

There will always be some trial and error, as not every machine can do the same thing or have the same capacity. But I have found that for the most part the tool manufacturer's chipload charts work well; I use their lower number. Keep it simple. Once you know the chopload, and you spindle RPM, you have your feedrate. Then just adjust your depth of cut to match your machine. The feedrate is directly proportional to the number of flutes and spindle speed, so if you increase or decrease either one, you end up increasing or decreasing the feedrate as well. Do tests, going .01" doc more at a time till the cut goes bad, then back off about 10%.

[dbsharp]

it can also limit your top speed if your computer cannot produce enough steps. If your using 1/256 micro step on your drivers, and your computer can only produce 25k signal, you only see 30 rpm on the stepper motors.

[jsearles]

DB, as i have been playing with tuning the motors since I started this thread, it has been easier to see the effect of a high resolution setting. As you alluded to, less speed and less torque. That may be the entire problem and reason for my frequent stalls.

Thanks

[dbsharp]

I found that I got higher speeds with my stepper motors connected at the half windings. The top end torque was better. If you want heres a good read, Stepper Motor Inductance / Torque

[jsearles]

DB, thanks for the info. I'll give that article a read as soon as I get a chance. Right now, I'm waiting on my Super PID to arrive and I've started on another build.