Hurco KMB1x backlash and leadscrew strapping

Yep, still using one of these old dinosaurs even now. Or rather using it again. Had a couple I was using to pay the bills years ago and kept them when I went to work for the university. Well, the shops on campus are all locked down so I'm working from my home shop and that means dusting off the old Hurcos.

The good news is the machine works, mostly (I have to warm it up for a while before zeroing the table or it hangs while looking for Y after going to the limit switch) but the dimensions of the finished parts suggests it's past due to adjust the backlash and leadscrew compensation strapping.
I have the requisite 13.1" bar made up, I have the chart from the service manual giving the details of what straps give what compensation.
(also available on Buddy's site: http://hurcousa.com/strapping.php, thanks Buddy!)
What I don't have is a handle on the procedure. Does the machine use the strapping value present at any given moment or do I have to re-zero the table, reload the master, or reboot the whole machine before the changes take effect?
There are six blocks and the strapping sheet says the top row is leadscrew, bottom row is backlash. but not if they are in X,Y,Z, order or not?

While on the topic of antiques, I have a couple other questions:
Where is the control for the auto-lube, I see a electromechanical timer in the electronics chassis but I think that just runs the lube in the spindle housing?

Turning the manual feedrate override pots also affects the spindle speed. I'm pretty sure this isn't supposed to happen. Suggestions?

My Y axis has some palsy, particularly when cold. I tried tuning the servo but during the "Adjust (GAN) until the motor starts to oscillate, turn (GAN) CCW until oscillation stops,
then turn an additional 5 turns CCW." stage, the Y axis never really oscillated. How worried should I be?

thanks folks,

B or BX control? although it may not matter. On the backlash comp, I'll check my service manuals tomorrow for that. Since you have to pull a board to access the strapping, and you should never do that with the power on or the batteries connected, yes, you will have to reload the master and rehome the machine. If your rear Y axis rubber way cover is bulky. pull it up out of the way when homing as that can be a problem if it all bunched up between the back of the table and the column. Also you can check the home flag on the bottom rH side of the saddle- the bracket with the 1/4" diameter pin in it- make sure its adjusted upward against the limit switch rail- the little square opening in the bottom of the rail is the home switch and it wont trigger if the gap is too wide.

The autolube control is the timer mounted on the back wall of the RH cabinet- manually rotate the wheel and you should hear the air piston fire in the main lubricator.

Over ride knobs affecting the spindle speed- that's a new one on me- member here "Rooster" might have a thought- check the wiring schematic- I have it if you need one.

Y axis servo- they get neglected because of being in the knee- its a pain, but pull the motor out and blow out the brush dust and check the tach and power brushes.
Ya gotta pull the y axis encoder (easier on a BX machine) pop off the timing belt sprockets and unbolt the servo bracket. There are jacking screw holes in the cover, so its not too bad. Then pull out the brushes and blow out the motor to get the brush dust out. etc.

That shouldkeep you out of trouble for a few minutes- good luck

BX control.
I don't have to pull the board to get to the strapping, it's out in front of the stack.

I actually removed the Y way cover to put a new one on, so it's not the issue. I'll have a look at the flag though I think I checked that previously. It's weird in that it will work reliably if I use the machine in manual mode a bit before zeroing the table, or even just setting it to operate and letting it sit seems to do the trick.
Powering it on and not powering the servos up does not help...

OK, thanks. I thought that timer was the control. I hear something in the head but not from the lube pump down below, I'll check it out more carefully then.

Dan from Oakland said:

B or BX control? although it may not matter. On the backlash comp, I'll check my service manuals tomorrow for that. Since you have to pull a board to access the strapping, and you should never do that with the power on or the batteries connected, yes, you will have to reload the master and rehome the machine. If your rear Y axis rubber way cover is bulky. pull it up out of the way when homing as that can be a problem if it all bunched up between the back of the table and the column. Also you can check the home flag on the bottom rH side of the saddle- the bracket with the 1/4" diameter pin in it- make sure its adjusted upward against the limit switch rail- the little square opening in the bottom of the rail is the home switch and it wont trigger if the gap is too wide.

The autolube control is the timer mounted on the back wall of the RH cabinet- manually rotate the wheel and you should hear the air piston fire in the main lubricator.

Over ride knobs affecting the spindle speed- that's a new one on me- member here "Rooster" might have a thought- check the wiring schematic- I have it if you need one.

Y axis servo- they get neglected because of being in the knee- its a pain, but pull the motor out and blow out the brush dust and check the tach and power brushes.
Ya gotta pull the y axis encoder (easier on a BX machine) pop off the timing belt sprockets and unbolt the servo bracket. There are jacking screw holes in the cover, so its not too bad. Then pull out the brushes and blow out the motor to get the brush dust out. etc.

That shouldkeep you out of trouble for a few minutes- good luck

Click to expand...

Easy way t check the home switch- hold a steel block up on the switch housing as the saddle moves back and let the flag push it over the switch- hold the block in place after the slide stops moving and let the machine jog back out in x and Y after it finds the null marker on each encoder. If that's repeatable then its the flag gap.

I am not familiar with the B/BX controls, but on the slightly less ancient Max 2 Hurcos (with analogue servos) there is another thing that can cause the symptoms op is seeing - if the balance is off slightly, the axis can run up to the limit switch, but not actually back away from it. Because this part of the homing cycle happens in open loop it relies on the servo balance being pretty close. If this is the case it will just sit there indefinitely after it hits the switch. Working after it warms up is a pretty big clue that it's reference drift in the amplifier that's the cause.

Balance is easy to adjust with a volt meter as I recall. The balance pot on the amp gets tweaked until the voltage to the motor when jogging at 25 IPM is the same within a few mV in the plus and minus direction. One indication of the balance being off that I have noticed over the years is a flickering of the position display in the least sig digit.

Also after setting a home position and going back to the operate screen, the position display is not exactly zero in the last digit. Something easy to check.

Looks like the strapping pads are X-Y-Z from left to right.

The federate override pots applied voltage and grounds are all in parallel. Wonder if one pot has a slight short in it and is pulling the other two down with it. Have not had time to trace the wiring from the terminal block in the pendant to the control.

See attached pics.

Dan from Oakland said:

Looks like the strapping pads are X-Y-Z from left to right.

The federate override pots applied voltage and grounds are all in parallel. Wonder if one pot has a slight short in it and is pulling the other two down with it. Have not had time to trace the wiring from the terminal block in the pendant to the control.

Click to expand...

Thanks very much for confirming that, that will save me pulling my hair if the adjustments don't do what I think they should. Or at least about that one question. :-)
Interesting about the pots. Slight short or it could be the power supply is really soft there for whatever reason (bad filter cap?) so the pot in normal but the voltage sags as the value changes. It seemed like both the xy and z pots both affected speed but I should check that for verification. I seem to recall not noticing a change in feeds as I changed spindle though.

Dan from Oakland said:

Y axis servo- they get neglected because of being in the knee- its a pain, but pull the motor out and blow out the brush dust and check the tach and power brushes.
Ya gotta pull the y axis encoder (easier on a BX machine) pop off the timing belt sprockets and unbolt the servo bracket. There are jacking screw holes in the cover, so its not too bad. Then pull out the brushes and blow out the motor to get the brush dust out. etc.

Click to expand...

Guilty as charged, I've had the machine for something like 15 years and I don't know if I've even seen the Y servo... until this morning. Pulled it out and blew out the crud. I noticed thick oil on the tach brushes and the power brushes seem a bit worn. There was still some spring tension and the blocks are about .4" long. I have no idea what the service limits are on these? I believe I have a spare motor I pulled when I got a chance to scrap a machine, I'll see if that's around and how the brushes look.

BTW, do you happen to recognize this oddball retention-stud?

Oakland... you still there? I'm just down in Santa Cruz.

Should have mentioned this earlier- one of the problems with the Y servo sitting under the ball screw is it gets oil all over the servo and yes it can get inside and coat the commutator. Go to the hardware store and buy a piece of 4" stove pipe or some sheet metal and wrap it around the motor on the top side and secure it with a big hose clamp as an oil shield. The later machines had a shield on the motor for this reason.

That pull stud looks like the ones for a Hurco MB1 or MB2 bed mill. Those had a tool changer and Cat 40 spindles.

Folks poke fun at these old beasts with the baby blue paint and all and they might be slow by current standards but the darn things run forever.

Dan from Oakland said:

The autolube control is the timer mounted on the back wall of the RH cabinet- manually rotate the wheel and you should hear the air piston fire in the main lubricator.

...Then pull out the brushes and blow out the motor to get the brush dust out. etc.

Click to expand...

The auto-lube system is spread out all over the machine: lube pump low in the back, timer on the back wall of the RH box, and the regulator and solenoid is for whatever reason in the head of the machine. I had a kinked air-line so that's all good now. Thanks.

I found my spare motor and the brushes in that were worn to under .25", so I'm feeling OK about my .4" of brush.
I should find a source for new ones to have on-hand, but at least I can put the machine back together and feel good about it.

Dan from Oakland said:

Should have mentioned this earlier- one of the problems with the Y servo sitting under the ball screw is it gets oil all over the servo and yes it can get inside and coat the commutator. Go to the hardware store and buy a piece of 4" stove pipe or some sheet metal and wrap it around the motor on the top side and secure it with a big hose clamp as an oil shield. The later machines had a shield on the motor for this reason.

That pull stud looks like the ones for a Hurco MB1 or MB2 bed mill. Those had a tool changer and Cat 40 spindles.

Folks poke fun at these old beasts with the baby blue paint and all and they might be slow by current standards but the darn things run forever.

Click to expand...

My motor has the shield and the power brushes were spotless, the lube was just on the tach brushes.
I hosed the motor out with copious quantities of electrical cleaner and put it all back together.
Amazing! I don't remember that machine ever running so smoothly on the Y axis.

Indeed, it is off my machine with ATC. I guess the question I should have asked is where I can get more!
I've got four and it's a twelve position changer so at least eight would be nice and more if I could find them for a reasonable price.

Yep, I've been mocked for my Hurcos but I bought them for a song and they have mostly given me excellent service. Most guys buy a CNC mill and it has to make money, the Hurco is slow and cumbersome but it costs more in electricity to run than anything else. So I guess I can't say if they are great for heavy production service but they do excel in those roles where it might be hard to cost-justify a CNC because it won't be running all-day every-day.

thanks very much for the help,

Darn, seems I spoke too soon. The Y axis works great and I'm making good progress on my backlash settings but the table home is still an issue. It sometimes works, but not regularly.
Machine hits the Y limit and then backs out some distance from 0-3/8" or so and just hangs.

I'm pretty sure the limit switch is working OK. I can drive the axis manually and it always sees the limit, the rod is very close to the switch, etc.

Further observation suggests the fail to home has flipped, it works fine when it's cold and fails once the machine has warmed up.
I suppose I'd better re-tune the servo anyway after doing the cleanout on the brushes, etc.

teletech said:

Darn, seems I spoke too soon. The Y axis works great and I'm making good progress on my backlash settings but the table home is still an issue. It sometimes works, but not regularly.
Machine hits the Y limit and then backs out some distance from 0-3/8" or so and just hangs.

I'm pretty sure the limit switch is working OK. I can drive the axis manually and it always sees the limit, the rod is very close to the switch, etc.

Further observation suggests the fail to home has flipped, it works fine when it's cold and fails once the machine has warmed up.
I suppose I'd better re-tune the servo anyway after doing the cleanout on the brushes, etc.

Click to expand...

This is kind of unusual behaviour, at least it would on the later Hurcos that I am familiar with.

Normally if the marker pulse is missing it will just keep creeping along looking for it until it hits the opposite limit switch and faults out - there is no "fail to home" on those.

If your machine actually stops trying after some specific distance, it may just be that the limit switch has shifted slightly and is (just right on the cusp of being) too far from the marker pulse. That's speculation though.

Another possibility inferred by the warm/cold behaviour, is a failing opto or line receiver in the marker pulse (encoder Z channel) circuit, or a cold solder joint on one of those. I'd try and rule out mechanical causes first though - dirt on the encoder, limit switch position, broken cable etc.

A really good careful cleaning of the encoder with air and contact cleaner and thus far it has worked all morning without a glitch.
Repeated reboots during my backlash/leadscrew comp measurement and setting really makes one notice startup issues.

Dan from Oakland said:

Since you have to pull a board to access the strapping, and you should never do that with the power on or the batteries connected, yes, you will have to reload the master and rehome the machine.

Click to expand...

My interface board is out in the front of the stack, so I can get to the headers with the machine on.
I've also confirmed that if you change the strapping it does affect the machine without a reboot. I can't say it always does the right thing though. It seems like bits get set but not reset or perhaps it just updates the values when you table zero or switch from enter to operate more and back, because it sometimes/often leaves the table zero values hanging in an odd state. Still, if you run your program, it does change the amount the table moves which gets you close enough that you can just do a final reboot as a confirmation.

I really wish the source code were out there so we could look at how the machine really works on the software level. Of course given how tightly the executive has to be packed to do what it does it would be hard to read anyway.

Gregormarwick might be on to something with the home switch- you had to pull the belt when you pulled the Y servo out and that certainly changed the relationship between the marker pulse location and the home switch trigger point. Ya might want to go shift the Y belt on the sprocket a wee bit one way or the other. Since we normally don't set up the home position to a T slot CL or any physical table feature, you don't care where home actually happens but in this case it might solve your issue. Also, just loosening the encoder coulpling and rotating it on the screw would do the same thing without messing with the belt itself. You just want to change the relationship of the encoder and that switch.

If I remember correctly, the homing cycle moves to the limit, finds the marker pulse, moves an inch and zeros. I think the marker pulse appears every .200" of travel. If you pay close attention to the display for the axis that is homing you might get a clue what is going on. When the marker pulse is read a small dot appears next to the axis readout on the display. It's possible that the dot will remain after homing, if not, you can sometimes jog slightly until it appears. Another thing to pay close attention to is the display value when it hits the limit and also right before it resets to zero. It should repeat the same values each time you re-home unless there is a problem. If the marker pulse happens to be too close to the limit it may not register the pulse. It might be a good idea to verify the voltage to the encoders, if it's low that can cause you to miss a pulse too.

Sorry all, I dropped the ball on updates. I'll claim pandemic.
I did manage to get the machine in question to consistently find zero. I don't remember if it was a servo board that was getting ready to fail or if it was just a matter of tweaking the pots on the encoder on the axis, but I got it to run. It tool some time but I also got the backlash and wear strapping dialed in so I was getting consistent parts and my squares were square.
The machine was running so well, I decided to splurge and buy one of the upgrade CPU boards from Buddy over at AMTS-INC. Really I should just upgrade to a newer style controller, but this does keep doing what it does pretty reliably and I'm late enough in my career that learning a bunch of new stuff isn't the best investment when I'm already short on time.