[Brian L]

I was attempting to do some tapping using a tapmatic style head and found that my spindle speed is off. Doing a 1/4-20 thread, running 650rpm (I thought) and so I was using a federate of F32.5 (650 x.050" pitch). Tap went into hole, pulled out, stopped spinning, had to wait for the tapping head to catch up with it, so obviously feed rate isn't fast enough. So... I get my hands on a digital tach and find S650 commanded is more like 742 or something.

Now, I can cheat my way thru it by just increasing my feed, but obviously I have something wrong. I haven't checked all speeds yet, but 1000, 1500 and 2000 all are running faster than commanded. So.... there must be some way of setting the VFD hz output to jive with the commanded rpm. Are there parameters in the control, or pots on the acroloop card, or programming in the VFD that need to be set in combination?

Machine is a 1992 Partner 4, think it originally had a Yaskawa VFD, which was changed out for a Safetronics GP5, so it has obviously been changed at some point.

I appreciate any suggestions....

[neilw20]

VFDs usually have a parameter, that can set the scaling of input command voltage versus output speed.
As your original VFD was replaced, it almost certainly would need careful setting of this gain to match the original. They will be different.

The accuracy of the DC voltage being fed to the VFD may also be a problem as many PWM converted signals, once they go through an opto device suffer PWM distortion due to the difference in switch on time versus switch off time, which is all in the order of micro seconds.
Unless an opto with speeds up around 10MB is used there can be some significant PWM distortion.
PWM distortion becomes more pronounced at lower speeds, as the narrow pulses approach the difference in switching times.

I think the stopping before reversing is normal? Anyone?
As long as the pullout feedrate is SLOW enough the tapmatic will do its work correctly.
If feedrate is too high, finally the tap speed will be too low and it will get ripped out.

Using a tapmatic device, the feedrate needs to be on the low side to make the tapmatic do its work. It will compensate for the low feedrate.

[Brian L]

I'll check into the VFD parameters. I've verified my 10V signal and it goes from 0V to roughly 10V as it should. Quick check of my rpms this morning shows S700 actually gets me 803 or so.... I checked some other rpms, but the paper is in the shop and I don't remember, but they all were higher and appear to be relatively proportional.

I increased my feed rate to 40ipm to coincide with the actual rpm and the tapping goes exactly as planned, down in and then pause and reverses on pull out. And no, you don't feed a tapmatic, or any clutch style tapping head slower than the pitch.

[allen mullis]

If it has a cent 1,5,6,or 7 control, you should have a parameter in "misc" after entering "setup", that has to do with spindle speeds, more specifically top end speeds in each range. Try adjusting the top end speed higher by a few hundred RPM in the range that you are working with and check with your tach, repeat until your commanded speed matches your actual speed. If it has an older control, I wouldn't have a clue. I would advise aginst messing with the acroloop or VFD, though you could, it is a whole lot easier to change a parameter, that doesn't require opening the panel. Good luck.

I'll check into the VFD parameters. I've verified my 10V signal and it goes from 0V to roughly 10V as it should. Quick check of my rpms this morning shows S700 actually gets me 803 or so.... I checked some other rpms, but the paper is in the shop and I don't remember, but they all were higher and appear to be relatively proportional.

I increased my feed rate to 40ipm to coincide with the actual rpm and the tapping goes exactly as planned, down in and then pause and reverses on pull out. And no, you don't feed a tapmatic, or any clutch style tapping head slower than the pitch.

[ZZZZ]

The actual spindle speed you will get from a programmed spindle speed depends on 3 things:
1. The analog signal output from the Controller Acroloop card (the X card).
2. The parameters in the spindle drive.
3. The RPM parameters listed in the MISC set-up parameters for each range.

You want to adjust the top two pots on the X acroloop card, the gain and offset pots, to give you 10.0 vdc between MISC-13 and MISC-14. Program a speed higher than the max RPM listed in the MISC parameters for the range you are in, run CW and CCW, adjust the pots as needed.

Now adjust the drive parameters to give you the speeds, ramps, etc., you want.

Then run the spindle at 10vdc analog in each belt range, tach the spindle in each, put that value in the MISC parameter for each range. This gets the control to 'scale' the output correctly for the range the machine is in, assuming that you tell the control which range you changed into. The Max RPM value will clamp the output at the RPM listed.

You will still probably not get the exact RPM programmed, but you should be close.

[Mzones]

BrianL



First determine the actual full speed of the spindle.

This is a generic approach as I'm not familiar with your hardware or software

Verify that you have 10.0 Volts AT 100 percent PWM.

Set up a pulley ratio of 1-1 0-100 in your setup.
Start spindle at 100 percent PWM .(S100)
Your tachometer will Show RPM at 100 percent PWM.
Change the 0-100 in your setup.to the RPM that the tach indicated at 100 percent PWM
This should get you calibrated.


Mike

[Brian L]

ZZZZ, thanks, this is the process I need. A couple of questions, the two top pots, I assume we are talking R1 and R9? Which is which (gain vs. offset) and what does each one do? I don't quite understand why I would twiddle two pots to get the voltage adjusted. Also, how close to 10V is good? If my meter reads 9.98 is that close enough? I don't know what sort of tolerances to expect "electrically", and how stable it will be.

Drive parameters, ok, I haven't checked the actual unit in the machine yet, but I do have the original parameters from when the machine was built, for the Yaskawa VFD. I'm seeing a few different things, the Maximum Frequency is showing 190.0, Minimum Output Frequency of 001.5, but then I see Frequency Command Upper Limit of 100, Lower limit of 000. So am I correct in assuming (very dangerous for me and electricity) that my hz output on the VFD is from 1.5 to 190hz?

If I can get past the hurdles above, I think I understand the rest of the process.

[neilw20]

The two pots. I think one will set the top speed (gain), and the other the lowest speed(offset).
At low speeds, the tops speed pot will have little effect, but at the top speed, you might see more effect from the low end(offset) one.
Once you are happy with it, check mid range, and you might need to change the offset a bit and then the top end again.
The low end might already have an offset above zero, and the VFD setting will probably (maybe) not run to zero.
Pick four or 5 points through the range and plot on a graph your results, then you will see the effects clearly.
The gain will affect the slope of the graph, and the offset will move the plot up and down for the whole range.
With a bit of luck you will end up with a straight line, but if it has a curve, then you need to position it as best fit for your most commonly used speeds.
1.5 to 190 I think. Lower limit might need to be set to 1.5. This will allow you to bend the line in the graph, with a little luck.
right at the bottom end the torque might become useless, so once you get it nice and linear, you can push up the bottom limit in the VFD, at the expense of commanding say 25 rpm, and getting 50 but the linearity will probably remain the same.

[Brian L]

OK, spent some time today on this.... checked my speeds and voltages, results as follows:

S1000 1125 2.073V
S2000 2232 4.13V
S3000 3325 6.17V
S4000 4414 8.22V
S5000 5344 10.25V

I changed the speed parameter in the control to 5344 instead of 5000 and checked the rpms again and it brought them closer, like 1000 was 1040 or so. I tried adjusting the pots on the Acroloop card and got nervous as I wasn't seeing any change at all. Going to try to attach a picture of my spare acroloop card to see if I am looking at the correct pots. Also, how sensitive are these pots? Will 1/4 turn of the screw make a difference, or you need to turn them 5 turns to see a change?

OK, picture showed up, pots are upper left hand side of the image, a blue and grey one. I also have two pots lower on the card.... shown on the upper right side.... wasn't sure if those were the ones I should be fooling with....

Oh, one other thing, at 10V for the signal, this VFD is putting out 200hz, not 190 like the original one was, so I'm suspecting a change at the VFD might help to. Have to read up on the manual before I attempt that.

Attachment 221734

[ZZZZ]

Brian, I'm sorry, I screwed up.

The pots I told you to adjust are the Axis Gain (R1) and the Axis Offset (R9) pots. All I had you do was mess up your Following Error adjustments.

The pot you need to adjust is the very bottom pot on the X card. This is the spindle analog adjustment pot. On both of the cards I checked and on your spare card, that pot is the blue one of the pair between the P7 and P8 socket.



In MDI, I commanded a speed higher than the range I was in, and I adjusted the R15 pot to get as near to 10.000 volts as I can.

I wrote a program to set the Spindle Speed to 5% steps from 0 rpm up to the max rpm for each range, pausing for 5 seconds between steps so I could get a voltage reading. I then ran the program several times and noted the readings.

It was interesting to see how accurate each test was to each other; they were almost exactly the same numbers. I compared the values I got to the calculated values I expected to get and my worst error was only 3.1% higher at the lowest speeds but most of the errors were less than 1/2%. NeilW20 has probably described the low speed error better than I can.

The graph of commanded speed to voltage output was very linear so I can predict quite accurately what the voltage will be for any given programmed rpm. I tried adjusting the voltage at a lower rpm and it come out perfect for that rpm, but the error for the rest got worse (2%-3% average).

We need to remember that the control doesn't determine the actual rpm, that is the spindle drive's job. The control just provides the analog signal and a run/stop signal to the drive and then the drive's parameters take over.

Once again, I am really sorry I sent you down the wrong path. I need to rely more on my documentation and less on my 'old farts' memory.

[Brian L]

Hi ZZZZ,

Not to worry, I was very careful to bring the upper pots to the exact same positions they were before I changed them, and I had only gone one full turn on one and back to the original position. I kind of suspicioned that I was fooling with the X axis following as the other two axis cards have just those same pots at R1 and R9.... kind of made me think that the spindle pots would be the "extra" pots on the board, as the X is the only one with the extra components. I will try to play with the voltage tomorrow and see what happens.

I did do some adjusting today, checked the VFD parameters and it was set to an odd max frequency, 202.5hz, so I reset that parameter to 190hz, which is what my original parameters for the Yaskawa drive were. This resulted in a better correlation of the actual RPM vs. the programmed RPM. At 5000 my actual was only 5030, which was pretty close. The lower RPM was actually a bit worse, commanded 700 and got closer to 750, maybe it was 748... not sure now, all the numbers are starting to run together.

My voltage at the Max rpm call is running like 10.25V, so I suspect if I can dial that down closer to 10V I'm guessing all my rpm ranges will fall within a pretty small amount. The way I understand this, my max RPM is 5000 and Volts is 10V, so I should be getting 2V per 1000 rpm, and my checks of 1000, 2000, 3000, 4000 and 5000 bear that out pretty closely... again, all just a little high, so if I can turn the pot down a bit everything should fall right into place.

All of this because I wanted to so some tapping.... I should have just sprung for a $60 thread mill and been done with it.

[Mzones]

BrianL


You need to establish what the exact speed that the spindle motor rotates, at 100 percent PWM!
This number is needed to setup any and all additional pulley ratios !
You need to use this number in the spindle setup file.!
The number you use in the spindle setup needs to match the actual spindle speed at 100percent PWM input!

If you don't understand that, maybe having better lighting would help

See post #5 on this thread.this should take less than two minutes.

Mike

[ZZZZ]

Mzones, maybe you could explain more completely to help me understand what you mean by "100 percent PWM".

[neilw20]

Brian,

The numbers you measured, voltage wise, were quite linear.
At max speed, set the VFD so that you get the correct speed. -- you 190hz setting.
Then at the slow end, change the VFD minimum speed setting to get the bottom end correct.
Try that and see how it goes. With the voltage readings you got, it points to the VFD being the linearity problem.

[ZZZZ]

The control's job is to provide 0 to 10 vdc to the VFD. Get the control to 10 volts. Set the max frequency in the VFD to give you whatever max rpm you want. Put the actual rpm reading from the tach in the range parameter. The control can now scale the less-than-max rpm more accurately. (I've not seen a drive set higher than 180hz on a Milltronics machine.)

[allen mullis]

The possibility of messing up the acroloop settings are the reason for my suggesting that your changes should focus on the parameters in the spindle portion of the control, One would generally assume that the drive parameters are somewhere close to what they should be, which was verified by your measurements. That being said, getting your rpm within a tolerable range should be good enough for what you are doing, the VFD doesn't know whether the spindle is cold or warm or any of the other variations that can cause slight differences in the actual RPM, all it knows is "when I get this voltage, I output this frequency". I think if you had played with that parameter a little longer you could have got it within 10 RPM, as it is now you may have altered the signal going to your X drive, turning it back to the same place is not usually good enough due to the slop in the pot screw and the mechanics of the pot. ZZZZ is correct with respect to the HZ. settings, I just got done doing a drive swap on my 16, the top end is 180 HZ, and the bottom end is 1.5 HZ if I remember correctly. Since you verified your signal volts, and the hz parameters in the drive, the rest should be adjusting the parameter in the control to scale the rpm in sync with what the drive is doing, at least thats what the manual says. I have the parameters for a safetronics and Yaskawa drive, they are the same thing, if you pop the cover on your drive you should see yaskawa logo on the corner of the top board. It sounds like everything is working, so after checking the min. and max. hz settings, I would probably leave the drive alone. Maybe ZZZZ can help you with the procedure to set/check the resistance on the X pot, it has been a long time since I have had to do that, and it was on a machine with AC drives, so the things you have to unhook are different im sure. Good luck

[sportybob]

The easy solution (for machines with out ridgid tapping) is so set your spindle speed at a reasonable tapping speed (say 100 rpm, in low range) (newer machines can tap at much higher rates) then adjust the Max spindle frequency parameter ( NO-02 on one yakawa (GPD-333) drive and E1-04 on another (GPD-515) check your drive manual for the correct parameter) on the VFD drive untill you get your "100 rpm" (you can use xxx.xx) and get it REALLY close +/- 1 rpm.) doing this will get your best "tapping" results. You will notice that when setting a spindle spped of 5000 rpm you will only get maybe 4950 rpm. at 5K 50 rpm does not mean much. but you will be DEAD on at your tapping speed, which is more important.

While on the subject of spindle speeds, for those that have ridgid tapping and have a CSM spindle motors, when the spindle speed display starts "jumping" around, it is an indication that the upper motor bearing is bad. The bad bearing causes a "wobbel" in the upper motor shaft and causes the resolver output (sin / cos) to change. Meanwhile the resolver signals are feed to the spindle drive and is converted to an "encoder" signal that the control reads and the control continually tries to update the encoder signal hence the "jumping spindle speeds. solution: change the upper bearing before it takes out the resolver.

Sportybob

[Mzones]

ZZZZ


If you have one dollar in you hand, 100 percent of that dollar would be one dollar
If you had one dollar in you hand, then spent 50 percent of that dollar, you would
have spent fifty cents.left in your hand.
If you had one dollar in you hand, then spent 100 percent of that dollar, you hav
nothing left in your hand

The 100 percent PWM refers to the duty cycle Of the PWM signal
The 100 percent PWM means it full on
The 100 percent PWM means it at its' max.


Now get a breath of fresh air.

Follow the four simple steps I listed in post #5
That will give you that magic number that you are looking for.
That number is used in the spindle setup, and is the 100 percent PWM

The number you use in the spindle setup needs to match the actual spindle speed at 100percent PWM input!

These simple steps will do that , and take only about two minutes.





Mike

[Fastest1]

Isnt there a linearity.dat file In Mach 3 that needs to be deleted during spindle speed set ups?

[Brian L]

Wow, the post is going to hell and a hand basket...

First, Mzones, you don't have a clue, there is no pulley ratio, and you obviously have no clue about my specific situation. If you had bothered to read the posts before you offensively stuck your nose into it, you would see I have more than 100% of PWM, I have 10.25 volts and I'm in the process of trying to figure out how to adjust that. Try turning on the lights on your end and get a breath of fresh air.

Allen, I double checked the following error and I'm perfect at .050" when running 100ipm, a process I have done hundreds of times back trying to resolve a following error problem. Turned out to be bad voltage regulator that was pumping out -24V instead of -15V on my power supply. So that was an easy double check. The only parameters I have in the control are the speed range numbers, 600 for speed range 0, 5000 for speed range 1. The documentation I have from Milltronics, based on my serial number, shows the parameters settings of the VFD and it says 190hz, so that's why I went with that. I can change it to 180hz and see if that makes any difference.

Fastest1, we aren't dealing with Mach 3 here, so that doesn't apply.

SportyBob, thank you, I can try that, I'm not using low range though, running open belt, tapping aluminum at 700-1000 rpms or so. I work from a wheelchair and reaching up to swap the machine into back gear isn't going to happen.

Now, ZZZZ, back to what I really want/need to know.... I'm still getting 10.25V, I have turned the R15 pot with no effect whatsoever. I even went so far as to swap out the X axis Acroloop card with my spare card and fire it up with that and still absolutely no effect. So my question is, where is the acroloop card getting the 10V from? Do we have an errant voltage from a power supply somewhere that I could adjust? Or am I doing something wrong with my attempts to adjust... I MDI M03S5000 and it turns on the spindle, running at about 5030 by the way, and I turn the pot while it's running with no results. I tried stopping the spindle and the signal stays at 10.25V as that was the last commanded, and tried more adjusting and still nothing. Could there be some other reason that pots are having no effect on the voltage?