[pbrokaw]

Fadal 4020 y axis runaway issues:

Model number 906-1 (4020)
Serial number 9603353 (1996 machine)
Glentek servos and amps.
Machine had somewhat excessive reversal spikes according to the renishaw ballbar program… suspected a bad resolver and changed it out. From then on during power up the machine would runaway for a couple inches of travel and fault out due to over current. Then changed out whole servo for a rebuilt thinking something went wrong with the resolver that was rebuilt for us. No change.
Tried another servo… No change.
Swapped X for the Y axis amps just to check...no change
Swapped axis controller cards…no change.
Sealtight was old and brittle and had to be flexed to remove servo… thinking a wire was broken I replaced sealtight and pulled a new tach wire,checked over the resolver connector and wires…(seemed good). Motor cables seemed good also. Have measured 1.75vac on the lower brass bullet connector on the axis controller card. Have tried many different scenarios of the wiring combo’s for the motor leads to tach leads (i.e. red to black vs. black to black, black to clear vs. red to clear, you get the picture) problem still persists.

Any ideas from the techs on the forum would be very helpful.

Thanks

Perry Brokaw
Plastic Ingenuity Maint Dept.

[denmar]

Quick Question: after cold start the machine and respond Yes to the last home position,
does it shows any A or B axis Position?

[Joe S.]

Check your power supply to see if the voltages are in spec and the same. A new PS fixed mine, new on the net $125 or installed by tech $900 . . . ask how I know that price!

[pbrokaw]

denmar:

I cannot even get to the cold starting at all. faults out long before that stage of the game...

On the power supply... you talking the 5 and 12 volt +/- supply in the side cabinate at the top,or are you fererring to the axis drive high DC voltage one.

voltages recorded right now:

127.8 vdc for the axis drives(taken with the Y axis red power lead disconncted)

12.003 and 4.98 vdc for the computer power supply.

Perry

[Al_The_Man]

Try using a ±10vdc battery box in place of the analogue input and get control this way before going any further.
This tests the velocity loop drive/tach.
Al.

[pbrokaw]

Al:

Where does a guy connect in the -10/+10 voltage to? The axis control card or the amp itself...Sorry for the ignorance on this but in my 20 years of maint on these beasts here at Plastic Ingenuity we have never had a problem like this,nor the need to tap external voltages into a machine...

Thanks

Perry Brokaw

[Al_The_Man]

Hopefully you have schematics for the machine?
I am not sure what model of Glentek's you have but there is usually a 12 pin connector on the edge of the card, what you need to to is find out which two are the analogue input from the controller and intercept the two connections and feed the ±10vdc into these two, this tests the drive and motor without the influence of the CNC control.
If you can show a pic of the drive card I may be able to identify it.
Al.

[pbrokaw]

Al:

here is a pic of the amps...

[pbrokaw]

Al:

can't seem to post the pic, go to my flickr page:

Fadal axis amp drives Glentek | Flickr - Photo Sharing!

Sorry about that.

Perry

[Al_The_Man]

Looks like it might be 2 & 3 what is the board marked at these terminals?
Al.

[pbrokaw]

Al:

2= signal in

3= signal comm

Perry

[Al_The_Man]

Thats the two.
You can either use a single 9v battery and a 5k or 10k pot, or series two 9v batteries and connect the outer ends to the outer ends of the pot, the centre tap of the batteries go to T3 and the slider of the pot to T2.
This will give you bi-directional control with the slider centre being off.
Al.

[pbrokaw]

Al:

Thanks!! I will go make this up and try it out.... I will let you know my findings.

Perry

[pbrokaw]

Al:

all right I went and performed the voltage test using 2 9 volts as a voltage devider. I could run the axis pos and neg at a speed proportioal to the pot setting. Just not very fast as the alarm would pop up(axis does not respond to NC controller) as I did have the resolver disconnected at the gold bullet on the axis contoller card.

Perry

[Al_The_Man]

You could unplug the resolvers if this causes a control error, if you can control the axis without runaway, this would confirm the drive/motor/tach is working normally.
Also just jumpering the two terminals the motors should remain stationary and resist turning by hand.
If the above checks out then the problem is in the resolver/controller side.
Al.

[pbrokaw]

Al:

I Test ran axis motor direct with a 15v volt drill battery to motor leads.measured 1.87 vdc coming out of the tach feedback wires right at the amp chassis. The tach is indeed functioning properly.

Perry

[FanFan]

Hello. Don't mean to butt in but I have a couple of questions. The original problem was detected with a ballball setup. At that time was the Y axis running smoothly or was it running rough? Spikes were mentioned. Was that seen with a scope or with something else?

Also did the runaway problem showed up right after the resolver was changed? Was the resolver rebuilted or was it replaced wth a known good one? The resolver probably has 3 pairs of wires. I think they are called sine and cosine and the resultant. Wonder if the wiring got swapped around?

Good luck with your problem.

FanFan

[pbrokaw]

It ran okay but the circlatrity was about 1 1/2 thousands. The machine in question is used for milling snytacktic foam. Here at Plastic Ingenuity as a CNC wears out it's duties change. I was hoping this would have been replaced about 2 years ago when the head casting broke out around the ball nut for the Z... We had a 4020 Fadal that had caught fire(literally) and the insurance co gave us the check and allowed us to scrap it as we saw fit. we yanked the head casting,servos,boards,pendant,monitor. sold the rest for scrap metal.

I have tried different servo's,resolvers etc...

Here,read this. These are the notes so far:

Fadal 4020 y axis runaway issues:
Model number 906-1 (4020)
Serial number 9603353 (1996 machine)
Glentek servos and amps.
Machine had somewhat excessive reversal spikes according to the renishaw ballbar program… suspected a bad resolver and changed it out. From then on during power up the machine would runaway for a couple inches of travel and fault out due to over current. Then changed out whole servo for a rebuilt thinking something went wrong with the resolver that was rebuilt for us. No change.
Tried another servo… No change.
Swapped X for the Y axis amps just to check...no change
Swapped axis controller cards…no change.
Seal tight was old and brittle and had to be flexed to remove servo… thinking a wire was broken I replaced seal tight and pulled a new tach wire, checked over the resolver connector and wires… (Seemed good). Motor cables seemed good also. Have measured 1.75vac on the lower brass bullet connector on the axis controller card. Have tried many different scenarios of the wiring combo’s for the motor leads to tach leads (i.e. red to black vs. black to black, black to clear vs. red to clear, you get the picture) problem still persists.
Called David at VMC electric in MT. was told to disconnect the control feedback at the axis control card via the gold bullet connector on the card. And to jumper J8 (diagnostic safety on power up). After doing this the axis will runaway and bounce off the hard stops and over current out.
With the tach leads hooked up color matched and the gold connector undone it would take off and fault. Reconnected resolver and now it’s faulty with no movement.
At this point I went and performed the voltage test using 2pcs of 9 volt batteries as a voltage divider. I could run the axis positive and negative at a speed proportional to the pot setting. Just not very fast as the alarm would pop up (axis does not respond to NC controller) as I did have the resolver disconnected at the gold bullet on the axis controller card.
Test ran axis motor direct with a 15v volt drill battery to motor leads. Measured 1.87 vdc coming out of the tach feedback wires right at the amp chassis.
Hooked motor and tach back up to axis amp. Had control lead to axis controller card (upper gold bullet disconnected. Upon e-stop clear and jog reset the motor jumped hard and faulted out.
Found a wiring combo between tach and motor that allowed for VERY slow rotation of the motor with the resolver input and the control input unplugged at the axis controller card. Balanced out the amp using the BAL pot on amp. Got rotation to stop. Reinstalled the bullet connectors thinking the problem was going to be resolved very shortly. Upon estop and jog reset the motor would sit there until you jogged it a couple clicks. The motor took off and faulted out (over current). Swapped the tach wires for the heck of it… same results except the rapid rotation was very jerky.

[Al_The_Man]

The ±10vdc test will test the drive, motor and tach, the resolver and CNC controller will be out of the picture, if you disconnect the resolver during the test, it should not cause a fault in the CNC control.
This set up is used to set what is called the velocity loop and is used to initially tune the drives, if the tach is connected wrong or defective, this will cause a run away.
If you can control the motor fwd and rev and also test the drive with the two analogue terminals jumpered which should cause the motor to freeze or creep if the balance is slightly off.
If all the above work, then this eliminates the motor, drive and tach as a problem.
Al.

[Marty_Escarcega]

The ±10vdc test will test the drive, motor and tach, the resolver and CNC controller will be out of the picture, if you disconnect the resolver during the test, it should not cause a fault in the CNC control.
This set up is used to set what is called the velocity loop and is used to initially tune the drives, if the tach is connected wrong or defective, this will cause a run away.
If you can control the motor fwd and rev and also test the drive with the two analogue terminals jumpered which should cause the motor to freeze or creep if the balance is slightly off.
If all the above work, then this eliminates the motor, drive and tach as a problem.
Al.

Al, I'm a little confused, if you have the 2 terminals on the servo amp, you are applying 10VDC to the two, swapping the polarity will reverse the servo. The servo motor will respond proportionally to the voltage applied.

But you say jumper them, how can you jumper them? would that not short your power source? (I have a bench power supply I was planning on using)

What am I missing?
Thanks Al
Marty