[zukes3140]

I have just purchased a Bridgeport EZ-TRACK DX2 S/N 8251-E and I am trying to cut a circle 10 in. in Dia. I have installed the X&Y backlash comps and I cannot cut a good circle. I also bought new ball screws and nuts. I ploted a polar chart of the circle every 15 degrees. On the X-axis Plus and X-axis Minus the circle looks good until it gets to the y-Axis at 105 degrees Plus, and 270 degrees Minus. The circle deviates about .005 from the perfect circle. At 105 degres the circle deviates plus .003 from the perfict circle, then jumps to minus .002 at 90 degrees. At 285 degrees the circle will deviate, plus .001 from the true circle and jumps to minus .004 at 270 degrees. Talk to me like a six year old; because, this is my first CNC Machine.

[machintek]

There are a few things to check. From the simple to the complex.
Do you lock your knee? The series 2 has a bit more play in the knee and as the mass of the table changes center of gravity, it can rock.
You have said you set your backlash. Usually, I do this in a dynamic state. I make the backlash bigger by a few thou because I can measure this a lot easier at the points of the compass (sweep the circle with a dial indicator). If you do not have enough backlash, you actually get a flat at the compass points which is hard to measure. By seeing how big the jump is, you can deduct that value from the backlash you have dialed in. Since you changed the ball screws, the pitch error comp must be set to zero unless you know how the screws were ground.
Are you DOS literate? Do you know how to attach a full size keyboard? The next step would be to get to a rott directory program called PFM.exe. You can edit parameters here, run a program and watch position error AND watch the DAC command (digital to analog). Typically, you run an axis back and forth watching the DAC command and position error and adjust the drive card so both directions are balanced. I am assuming that the motors are in good shape, belts have the correct tension, gibs have the same play the entire length of the circle you are trying to mill. I installed 8249E in March of 1997 so your machine is probably 8 years old. Most machines are used in the center of travel and the way surfaces get "dished." As you get ouside of the usual travel area and it gets tighter, it may cause drag and a misposition. If it is overshooting, it is typically a drive tuning issue. This usually occurs closer to the 45 degree on a circle. Start looking at some of this and give us an update.

George

[zukes3140]

Thank you very much for your reply to my problem with the EZ-TRAK. I am going to bear all my ignornance, but what is a rott directory program, and what does PFM.exe. mean? I removed the saddle and checked the ways with a straight edge and blue. I used granite straight edge and indicator to varify the straightness of ways, guiding side. gib side, and keepers, they are within .0002 parallel and straight. I assembled the saddle to knee, table to saddle, installed the gibs, and pushed them by hand, they are smooth when I push them and there are no tight spots. I am confident this is not the problem. Moreover, the people I bought it from indicated that the mill ways had been rescraped. Thus, I believe the problem lies somewhere in the control. I have a Maintenance Manual, but it is very confusing to read the instructions on how to check the voltages at the control. Is there a Service Manual that explain in precise detail the keystrokes that I should try? I have obtained a keyboard, but I am needing more detailed information before I go this direction. I am going to learn this with your help and guidance. zukes3140

[machintek]

You have done well Luke Skywalker! If I may suggest my version of a bible: DOS for dummys. I apologise. Rott needs to be "root" directory, which is the first level directory of the hard drive. PFM.exe is in the root directory. First thing to check but I do not believe the problem is here. I copied this from another post.
At the main screen (where you can select Do event, etc) press the "*" key. In the top left hand corner, it will ask for a code. Type in "11235" floowed by the enter key twice. These are the machine parameter. The rest are in a BMDCPRMS.SYS file. The file you are looking at is really the sys.bez file in the root directory.
For a series 1:
Machine series: 1 Series 2
X travel: 30.0 32.0
Y travel: 12.0 14.5
X offset: DNA
Y offset: DNA
X backlash: always a positive value. Never saw anything greater than .0035
Y backlash: same as above.
X lead error comp: positive or negative. Never saw more than .0035 here.
Y lead error: same as above.

On the back of your CRT housing, you will find 3 keyboard plugs, 2 of which are not used. These are actually a jumper. Move the existing wire to the top one and plug in your full size beneath it.
Boot control; press 9 for utilities; you should have a black screen with white letters (if not you have the newer software , so now press 4 for comm); Now do a CONTROL C ( which means press down and hold down the control key and then pree the C key, now release both). At the bootom of the screen, you should see "terminate batch job?" Press Y for yes, you may need to also press the enter key. ou are now in DOS. To make sure you are in the root directory type CD\ and enter.
Now type PFM.exe and enter. Use the arrow down to highlight TUNER; press ENTER.
You will now see across the top your axis position, position error, DAC command, home switch state. Bemeath that are commands that are self explanatory.
You may have to write a small program to run the X and Y axis at a 45 and tune the drives using position error and DAC. This is done using the offset and gain pots on the drives. Your maintenance manual may have a description of the drive boards and the pot location. If I can dig up a procedure, I will try and post it.

George

[zukes3140]

I thought I would provide more information on what I have done to the Bridgeport EZ-TRAK. I have to admitt that I am a little intimadated in adjusting the program in DOS. Thus, I set up a standard and (Starrett Reference Bar) and checked lead error of the screw in X=axis, a distance of 22in. The difference between the read out in 22in was +.004. However, I had to change lead error -.007 to get +.001 in 22in. I checked Y-axis with a 10 in. jo block and it was off only -.0005 The Parameter settings now read: [email protected]; Y-Backlash @.0015; Z-Backlash @000; X-Lead [email protected]; Y-Lead [email protected] I positioned a ring gage on the table and traced the I.D. with a Blake CO-AX indicator so that I could keep the indicator point at 90Deg. to the direction of travel.One thing stood out while making these readings. At the Y-Axis, in the 6:00 o'clock and the 12:00 o'clock position, the Y-axis handle would stop turning about 5 or 10 sec. But the x-Axis handle kept on moving at a normal speed. I will attempt to send an attachment of the indicator ring gage readings. In addition, I programed a 5in. square to be cut in a piece of AL. The square measured 4.9995 in the X-Axis; however, it measured 4.993 in the Y-Axis. I hope this will help you in your diagnosis. When I have more confidense I will then start cranking on the electrical gismo knobs. Zukes 3140

[zukes3140]

I hope this goes through.Zukes3140 Sorry The Polar Chart is to large to up load.

[zukes3140]

I hope this goes through.Zukes3140 Sorry The Polar Chart is to large to up load.

[machintek]

Humour me here. Try this. I usually increase the backlash by .002 for both axis and cut a three inch cicle with a spring cut (finish cut) in aluminum.
I then sweep the circle with a dial indicator. I find that dynamic testing is more accurate than static testing. Also it is easier to measure a step in a cut at the points of the compass caused by backlash overcompensation than it is to measure a flat on a circle. The step is the amount of backlash that needs to be subtracted from the value being used. Make sure the knee is locked! If the step is different at opposite ends of an axis then I would be inspecting the ball screw/ball nut. The machine may have done a lot of repetitive work without the vise being moved around.

George

[zukes3140]

I was successful with my first attempt to get into the sys.bez file and look at the control parameters. When I did the CONTROL C, nothing happens. I do not get a response of any kind. Should I remain in the sys.bez file, or exit this file before I do the CONTROL C? Please advise.

[machintek]

I am sorry but I may have confused you. There are two ways to get to those machine parameters. By re-reading your posts, it is obvious that you have entered the file through the "front door." That is by the "*" and "11235." There is another way to look at the file. In DOS, the same file is called the SYS.BEZ file. But it lacks the descriptions that the front door has. The control C is used to get to dos from the utilities page.
I usually measure backlash comp in three places on each axis. Both ends and in the center. This gives me a good idea of what shape that axis is in. The pitch error comp or lead error comp is a fudge factor to achieve better accuracy. No ball screw is ground perfectly. This corrects for this inaccuracy. Most high end CNCs devide the ball screw into short increments and have a correction table for each increment. This means that the EZTRAK can have some variation as you go along the length of the screw. Again I state that you may have different results trying to figure out static backlash and dynamic backlash. This mostly due to "Stiction". The difference in the coefficient of staticfriction to sliding friction. I also pump the lube oiler a few times manually to make sure there is enough oil in the table.

George

[NC Cams]

We had a similar problem with our EZTRAK even after doing all the comp and backlash checks.

Turns out that the tach circuitry in the motors was defective. Also, the encoders were dirty due to brush dust goo.

If the motors are SEM, Clarkson near Chicago will service them as they are the factory service center.

We'll surely have to retune the servos to the amps when we get them back.