Powermaster Down..HELP!!

[crow36]

Dear all, this is my first time here and currently m having a problem with my mazak powermaster 1600. The controller is Meldas L1. When the machine's running on auto mode, it will stop moving at G1 X...F... or at any point with G1 code. If its in dry run mode, the machine will complete the entire cycle.
There is nothing wrong when i change G1 to G0.

I'm suspecting the spindle feedback to nc side not complete. I don't have any wiring diagram for this as the original machine comes with DC motor but has been retrofit with a Fuji Electric inverter n AC motor.

Anyone has any idea how i can troubleshoot the problem?

Thanks in advance.

[chippy1]

Crow,

You are exactly right in reference to the spindle feedback being the problem. You need to locate the spindle encoder which is most likely driven by a belt with this old of an machine. The cable for this should be connect to the main mother board which should be a LX1 PCB with a smaller PCB mounted on top of it. Be careful not to mess with the smaller PCB to much. This is your main s/w card. It also houses the parameters and your programs too. There is a battery wire connected to it too. The cable your looking for should be labled CNA1. CNA1 is located on the bottom left hand side. Should be a 20 pin honda connector. This is if your LX1 is mounted length wise left to right. This main pcb is about 24 inched long by 12 inches wide.

There are test points pin 8 and pin 9 which are for PA & PB respectively. Pin 7 is PC. 1 pulse/rev for threading. You will need to hook a scope up to read these points by sign wave. One pulse/rev. for PC maybe hard to see.

Hope this helps.

Good Luck.

[Leelund]

Is it the axis that stops feeding or is it the spindle that stops? Lee

[crow36]

Chippy: Thanks for the info you've given. I have tried to plug out the CNA1 cable from the pcb board n the encoder side only to find out the original encoder is no longer being used.
After pluging both end out, i'm still able to rotate the spindle via MDI command. No alarm appeared.
The DC spindle motor has been converted to Fuji Electric AC Motor n Frenic 5000VG7S Inverter unit. I'm trying to figure out which is the cable sending signal back to NC unit regarding the SPN RUN ON or something like that. Also any idea whats the diagnostic number on the Mazak side with regards to this SPN RUN ON?
Thanks

[crow36]

Lee: Sorry for not mentioning it on my earlier post. The Axis is the one that stops moving when it reads G1 code. The spindle remains turning and no alarm appeared on the screen.
Thanks

[chippy1]

Crow,

Not sure I understand your configuration now. Is the inverter sending some type of encoder feedback to the CNC through CNA1. No matter what you have to have the encoder feedback going through CNA1 in order to feed using G01. I would think that they would still being using an encoder to follow the speed of the inverter. Please explain the configuration a little deeper.


Regards,

Chippy

[crow36]

Chippy:
I've tried to remove the CNA1 cable from the LX51B PCB board and also on the encoder side. At that particular moment i could spin the spindle via MDI command.
To my surprise no alarm appeared. On other machine i think there should be an alarm like "SPN SPEED MISMATCH" or something like that right?

So i assume the original encoder has been bypassed. I'll be checking the cables today and then maybe ask the MITSUBISHI guys to come over to either replace the encoder or the LX51 pcb.
Any more ideas?
Thanks alot.

Cheers

[chippy1]

Crow,

You said that you had a L1 control. Typically this control has a LX1 card. However you described it has a LX51 card. Is this a Mazak machine. If so, is the CNC a T1? Typically the LX51 is used on a T1 control not a L1 control. However, there could be different BOM's for the L1 I guess.

Did you measure the test points on the card that I mentioned in a different thread to see if you had signal coming back from the encoder? You will need an O'scope to do this.

Do you know somebody from Mits that you normally talk too? If so, what's his name?

Regarding your assumption for spindle speed mismatch. This control is not that intelligent. That came in the late 80's with the digital age. The CNC is not measuring the PLG feedback from the spindle, yet in this case the inverter. So there is no reference for the CNC to compare with the encoder feedback, thus no error.

Regards,

Chippy

Regards,

Chippy

[crow36]

Chippy,

This is indeed a Mazak Powermaster1600 with Meldas L1 controller. The NC side have LX51B, LX59B and some other board m not sure whats the name. It is sure not a T1 controller as you have mention.

For measuring PA,PB n PC, i could not do it because i do not have a scope.Can i do it with a multimeter or such? I've check the entire length of the CNA1 cable from the encoder till the connector where it connects to the LX51B and is all good.

I usually speak to the technical guys from Mitsubishi Singapore.

Regards
crow

[chippy1]

Crow,

What is the model # of the spindle encoder that you have on this machine? You really can't measure the signal without a scope, because it is a 5v square wave which PA & PB are 90degree out of phase. If it were a sign wave then we could probably mease it with a multimeter and get the RMS value from the p/p value. The encoder that I think it should be is a RFA1024-22-1. 1024 pules per/rev.

Regards,

Chippy

[crow36]

Chippy,

The encoder model is OSE1024-3-15. I am getting a loan from the Mitsubishi guys 1 unit of the encoder to try it out. Hopefully it does the work.

Thanks for the help and will update you soon.

Regards
crow

[Leelund]

Maybe I missed something somewhere while following this thread but I've got ask, why, if the spindle runs properly and the axis feed is the problem, would you think there is something wrong with the spindle controller or feedback to it? Does the crt show spindle rpms when the spindle is running? If so, your encoder is probably fine. Everything is fine in G0 mode. Is everything fine in manual jog? Will the axis feed by itself with a G1 X....F.... MDI command? Is it only X axis or does it do the same thing in Z? It almost sounds as if you have the feedrate override set to zero. Maybe a problem with the feedrate override function? Sorry for more questions than answers but it's killin' me.
Thanks, Lee