IH Conversion questions

[Ed from NY]

Hi everyone,

There is a good chance I would be getting a IH with the 3HP motor and convert it. Doing the conversion is a money thing and also I want to step up (pardon the pun) as a preparation for bigger projects.

I currently have a smaller CNC mill which I have taken apart and put together many times. So kinda I feel confident that although there is a lot for me to learn in doing a conversion from scratch I know more or less how the parts fit together. I would probably be using a USB box to do the real-time talking with the mill. I'm familiar with Mach3, but I'm not averse to EMC (though I have not used it). I would consider using flashcut, but I will need to run some number$ before making that jump.

As a design parameter I'm looking for 200 IPM in XY with a 200 lbs table load. On Z I would like to shoot for 150. Together with the 3HP motor that is about double of what I have now in each of those categories. I think the machine would still hold together in one piece no problem. I'm not ruling out steppers, but I would like to get some experience with servos.

So...

I have a few questions I was hoping you guys could help me with

1) What size ball screws are best to "fit" and still take advantage of the full travel of the machine in XYZ?

2) What kind of continuous torque motor would I need for reasonably quick (perhaps 1/2 sec) acceleration for the loads mentioned?

3) If anyone is running that spindle at "high" RPM's (3600-4800), how well is it holding up? Is it very noisy? Does it run true? What is the TIR?

Thanks a lot!!!

If it all works out it should be a fun!!! project

[ninefinger]

Hi Ed,

I have converted my mill which is an IH clone from Machine Tools Warehouse.

I used 50 oz-in continuous torque servos for my build (nema 23 frame) and I'm using a 4:1 reduction drive on the x and y axis, 5:1 on z. This results in quite healthy accelerations (70in/s^2) with my 50lb vice on the table, (ie under 0.05 seconds to 200ipm velocity), less accell and velocity on z (~50in/s^2 and ~100ipm due to variable gib tightness at top and bottom of travel)

I can achieve 200ipm (x-y) but its at the max for my setup. If you want to achieve these numbers regulary then what I've done is the bare minimum and its best to add some margin to it (but not too much if $ is a factor).

For the DC brushed servos the next step up is 100oz in I think. Another option might be looking at the kit from here DMM tech Pricing - guys in the router forum have used it and seem pleased. I would like to try but the max voltage is below my current power supply so I'd need a seperate one just to try...maybe later when I do my lathe. If you need to buy all your parts then 1 stop shopping might make good sense...

Mike

[arizonavideo]

I have finished my CNC convertion of my IH mill and I ended up using large servos much like the largest Kelling KL34-180-90.

I get lots of top speed but you just cant use it with the slow spindle RPM.

I just upgraded to a 4hp 3,600 RPM motor and it works fine but high range makes even more of a racket. 2nd high is fairly smooth for a RPM of 1,800

I did a quick video of the table doing a test run at 300 IPM.

[nomedia="http://www.youtube.com/watch?v=M299KWBsuBs&feature=player_detailpage"]YouTube - IH table speed test 300 IPM.MP4[/nomedia]

I got lucky and found the only 7.5 TPI ballscrew I have ever seen. This let me do a direct drive setup for the table.

I can get 380 IPM but I just don't see the need. I set the max speed at 125IMP.

When dry cutting aluminum with just air blast at 1800 RPM and a 1/2 carbide 4 flute I could only cut a outside profile at .2" d at 60 IPM with a 40% engagement before I started clogging the bit.

I have done some cutting with larger bits but not at max yet. Either way I don't think at 3,600 RPM I will ever cut more than around 100 IPM.

I have a few other videos of the mill on youtube and I will now that the conversion is done I will add a few more.

I'm sure the smaller servos would be fine for the table but you may want the larger one for the head. My head is still slow and I am going to change the gearing to get better acceleration at low speed.

[Ed from NY]

Great info guys!

I use 3000-3200 all the time with steel and a 1/2" 4-flute, that gives me the 400 FPM (yeah, ear protection ). So I'm looking to go higher for smaller tools and softer materials, but I suppose we are already close to what can be done without getting too costly.

Did you try a higher RPM (the full 3600) with a 1/2" 3-flute mill and energetic coolant? Try that with the same 60 IPM feed and 40%. It might work. If it does you should be able to go 0.6-0.7" DOC with that big motor of yours

Nice videos BTW.

I just got a IMV100 Graymills. It is a 3-phase 1HP coolant pump rated at 50 GMP (over 60 ft of head) so I'm really exited about hooking that baby up

[jetflatline]

Just curious -- is the need for travel speed simply a matter of increasing production rates or are there other reasons?

[Ed from NY]

Just curious -- is the need for travel speed simply a matter of increasing production rates or are there other reasons?

If the question is in relation to cutting conditions then the answer is that to cut at 100 IPM you would need either a lot of power with a big bit or a lot of RPMs with a small bit. That is because the recommended inches per tooth get bigger with bigger endmills and smaller with smaller ones. So with one rotation of a bigger bit you can "walk" more.

Here are some numbers and you can work out the calculations. These examples are for 3-flute SGS S-Carb endmills on Aluminum.

1/8" diameter -> 0.0015" recommended feed rate per tooth
1/4" -> 0.0030"
1/2" -> 0.0060"
3/4" -> 0.0080"

Feed rate per tooth for plastics are twice as large, so... I guess you could try it if you have a mill like Arizona's

Try cutting Delrin with a 1/2" 3-flute (check your endmill's speeds and feeds recommendation) at 3600 RPM, 130 IPM, 0.1" WOC and .5" DOC. I do not have power numbers for plastics, but it should be very doable on a 3HP machine. Make sure to use compressed air for cooling or things will get gobbed up pretty quickly... and if that does not work, let us know what does so that we can learn

Good luck!

[arizonavideo]

Some time ago I watched all the IH milling videos and he says they were running at 2500 RPM.

Their machine has a 3,600 rpm motor and a VFD. They can get 5k RPM from the motor or more with the VFD.

I never understood why they did not run around 4K RPM but I think I have a better idea now.

With the motor at 4500 RPM and the gearbox in second you get about 2500 RPM. High range is almost twice the RPM but my machine makes a massive amount of gear noise in high range with the 3600 motor. I don't know how long it will last but it is really loud. If I was running all day long I might skip high range too.

Do other IH owners with 3,600 RPM motors have a massive amount of noise in high range?

[Runner4404spd]

no. my gear box is relatively the same in all the gears and i run the high gear all the time. and i have a 3hp motor on there with a vfd. i can push just over 5500 rpm and the machine hauls. my biggest issue is not having flood coolant to cool the cutter when cutting aluminum. but i can take a .5 doc in cold rolled at about 4000 rpm with a .05 woc and the machine just hums along. i get my coding via high speed tool paths from onecnc.

[Ed from NY]

I really liked your idea, Arizona, of direct driving X and Y. You say that the biggest Keling's at 1.6 N.m may be able to drive a 7.5 TPI screw well. But in order to direct drive a 5 TPI screw I guess you would need 50% more juice

BTW: I found two levels of products:

One level would be like a Keling DC motor + Gecko and the cost is about $250 per axis. The motor runs at 90VDC and the Gecko tops of at 80VDC, so perhaps you want to go for a Viper? (add an extra $120 for $370 per axis). That would give 1.6 N.m at constant torque and a max of 3200 RPM.

The other level would be something like what Direct Automation provides. A 230VAC drive with a Brushless AC motor. The cost is about $1200 per axis. That would give 2.39 N.m at constant torque and a max of 5000 RPM. Big difference in price... 4-5 times as much.

[arizonavideo]

ED:

The difference in torque from 5 TPI to 7.5 is 50%. My servos are a bit special in being higher voltage, low speed high torque motors. This setup just happened to work out for the table but I still ran a 2 to 1 belt for the Y.

For a standard servo of the KL-34-150-90 type I would run a 3.5 to 1 or 4 to 1 ratio.

Runner: How loud is your gear box in general? I have to wear ear plugs in high gear. Did you change any bearings?

You only get so much power out of a Gecko drive. I was playing around with the numbers and if you adjust the gearing for different servos their was vary little difference with any of the servos. You can set up for really high max speeds but you will pay the price in low speed and directions changes. You will get good performance with the X and Y with any servo and proper gearing to 150 IPM or so. The head is a bit more picky if you under gear it it will fault all the time. The idea of a larger drive has crossed my mind but I just won another electro craft servo for $19.00 so I'm just going to go to a lower gearing and change the servo.

[Runner4404spd]

my gear head isn't that loud to the point i have to wear ear plugs, i did however replace all the shaft bearings that hold the gears as well as when i put the nmtb 30 spindle in place i replaced the spindle bearings with angular contact bearings.

[Ed from NY]

One idea I'm playing with is to go with something like Direct Automation, but have only one set of drivers for 3-4 machines. Put the computer the controller and the drivers in a cart and wheel it to the machine. Hook it up and use it. If it turns out that I need to run 2 out of 4 machines at the same time then add 2nd cart. Not sure of the pitfalls, but that would allow for the conversion to be cheaper in that you only need to get the motors and some cabling. If I do this I would start off with 2 machines and 1 cart and adding a 3rd machine later down in the year. The 2nd machine is a 2500 lbs lathe and the 3rd machine would be a VMC rebuild. That would allow me also to go with something like FlashCut for the controller as it could be shared. Is that a crazy idea?