BMC-20 Making Progress, More Q's

[fasto]

Well, with some help from the forum folks and quite a bit of reverse engineering, I have fixed the Fanuc A06B-6050-H401 triple servo drive in this machine. A Fanuc isolation amp (a little potted module), is faulty, and a Hitachi dual op-amp had a leaky non-inverting input. The op-amp is replaced, but I don't have another Fanuc A76-0300-0077 isolation amp yet. When that is replaced, the drive will be back in order. I also have hand-drawn schematics for most of the servo amp, so should it fail again I have somewhere to start looking!

While I was troubleshooting the servo drive, I took the servo motors off all 3 axes of the machine as I didn't want a runaway and I had to turn the motors by hand. Now, I realize that I've probably messed up the Z axis index pulse, and the toolchange height is going to be off.
Question I read that there is a secret screen that can be used to reset this - how do I go about setting this?

Speaking of secret screens, I also read about backing up the leadscrew map to floppy. I have floppy drives, but I don't know if I have a memory-based leadscrew map or a dip-switch based leadscrew map. The memory card has four dip switches on it. Also, this control was upgraded from V7.xx to V8.xx. I wrote down the minor numbers but left that paper at the office , have to get it next week. I can also get pics & numbers of the various cards if that helps.
Question Do I need to back up the leadscrew map, and if so, how?

Next, the graphics screen is dead. The driver board has a shorted horizontal output transistor and perhaps a shorted flyback transformer. Newark Electronics will be delivering a supply of h.o.t. transistors Monday and we'll see if that's the issue. Hopefully the flyback is OK, because I haven't been able to source a replacement. We will see.

The machine hasn't yet moved under power, but I have jogged all the axis servo motors off the machine, run the spindle - the bearings sound good, and oriented the spindle. I still have some checking to do, with the axis limit switches and the tool changer limit switches.

Which is my last question, is there a way to have the control show what it thinks the limit switch status is, or are they all shown on the manual-> tools & diagnostics-> diagnostics screen? I don't remember seeing the axis limit switches there. I ask, because there is something funny with the limit switches. Someone has added a limit switch on the toolchanger whose purpose is not obvious to me, and a second limit switch on the Z+ axis, and tied these into the E-Stop circuit. This leads me to believe that there are more issues, perhaps intermittent, yet to be discovered.

I did order some retention knobs and new bezels and filters for the EAO series 14 pushbuttons, so we are getting close!

Thanks,
Aaron

[bloke]

Sorry about stepping in late, Aaron. I got sent away working for most of this week. Seeing as you are progressing with the drive, do you still need the manual and card for it?
After re-mounting the motors, you'll need to calibrate the axes. After calibrating, you'll need to check that the axes jog back 0.1" (2.5mm) to the calibrate limit switch from the axis zero. If it doesn't, undo the motor and move the belt around a couple of teeth until it does. If you don't do this, it's possible that the axis could calibrate 5mm out of position. Make sure to do all three axes.

Once set up, the toolchanger height can be set by going to the toolchanger diagnostics page and jogging the Z axis and moving the magazine right/left into a tool in the spindle until the magazine engages with no deflection. Once done, press "enter" "488" "enter". This brings up the "store t/c height" softkey - hit this and the position will be stored. Exit out when done.

To save the leadscrew map, go to storage functions and select current storage as drive A. Put a disk in drive A (double sided, double density or if you're lucky, a high density disk with the hole covered with tape) and format it.
Calibrate the axes and go to the toolchanger diagnostics page and press "enter" "642" "enter". This brings up the leadscrew map page. Select the save option and it should write it to the disk. To re-load the file, hit the load option. Exit out when done.

[fasto]

Originally Posted by bloke Seeing as you are progressing with the drive, do you still need the manual and card for it?

Hi Cliff, I haven't been able to find one of the faulty Fanuc hybrid modules, so I am still interested in both the card and the manual. I exchanged the faulty module between "L" and "M" to verify my diagnosis, so I sill have a bad axis - it's just a different axis.

I will follow you steps to reset the servo motors.
Thanks,
Aaron

[fasto]

Some success - Newark delivered my H.O.T transistors after sending them on a scenic tour of the western US. And, I found a single-axis Fanuc drive board on ebay in the US, advertised as operational. Thus, I should have all the parts needed to make this a go.
Spent last night rebuilding the auto way lube pump, now I just have to replace all the metering units! They're very gunked up and don't flow anything at all. This is not surprising, because there was a half-inch deep supply of sludge in the bottom of the way lube tank. Should have all this sorted by the weekend!

[fasto]

OK, fixed the oiling system for the moment by soaking the metering units in kerosene (paraffin for our British friends), probably need all new metering units though. The bad news is that the Turcite is worn off on the back X-axis bottom side . Oh, and the flex line from the machine to the saddle blew out when I got the oiler working as it was badly cut at some time in the past, and misrouted to boot. Also I must have mounted the oil reservoir float upside down, as the low way oil alarm comes on when the tank is full. Fix that tomorrow.

My ebay Fanuc single axis servo card came today, removed the isolation amp and put it on my triple axis card. This got my servo drive working, all three axis at the same time even! Remounted all three axis motors, we're ready to go!

Power up, press calibrate axes, Z goes up to the limit switch then slowly descends... and descends... and descends... and after descending about four inches errors out with "No Z Axis Index Pulse Found".

Well, I hooked up the meter and the index pulse is there, and it repeats every 0.1968 inches (or something like that), presumably once per rev. Evidently the index pulse isn't getting to the control. Probably a bad optocoupler on the Z/S board if I had to guess. Of course I could get lucky and have it be only a bad cable, but I am expecting trouble with the Z/S board. I will get into this tomorrow, too.

PS, don't mind me talking to myself here, sometimes it's helpful just to 'write' down this blah-blah. And, I am gaining a very detailed understanding of the BMC-20. I'm even hoping to make some chips some day!

[fasto]

Well, I spent four hours cleaning solidified way oil out of the saddle internals to get at the 5 metering units for the Y axis ways & ball nut. The good news is that I have a relatively clean area to put things back together in, the bad news is that all five metering units are totally plugged up with gloppy crud. I expect the lines are also plugged up, but I can't get to the other ends without taking off the table. I expect that the metering units in the table are also plugged up, and they are impossible to get to without taking the table off and flipping it over.

Anyone ever take the table off one of these machines? Can the X axis ball nut be removed with the table in place, or does the X ballscrew come out with the table? It's not clear from the drawings in the books.

I checked the X and Y axis index pulses, they're working OK. I found a neat thing on the control, too: When the control is running, it displays the X/Y/Z position in the top left, like so:

X   39.8876
Y   39.7786
Z  -39.6577

The numbers are about what my machine reads before it's calibrated. When the control sees the index marker, it puts a decimal point after the axis that has the index active, in this case the X axis:

X.  39.8876
Y   39.7786
Z  -39.6577

So, I confirmed that the control isn't seeing the Z axis index marker as the decimal point never shows up. I checked all the cables which are all good, of course. I started the reverse engineering process on the Z/S board, and have the index marker circuit pretty well figured out. Need to test it tomorrow. The Hurco boards are only two layers and are very easy to trace out, and use mainly regular 74 series logic which is easy to get. Much easier to trace the wiring than the Fanuc servo drive board, which is 4 layers with internal traces.

[bloke]

Slide the table to the left side by winding the screw.
Undo the screws holding the motor box to the bearing housing and remove the box.
Slacken the four M4 screws on the ballscrew locknut then undo the locknut.
Wind the ballscrew out of the bearing housing then remove the housing.
Undo the oil pipe to the ballnut.
Undo the four screws holding the ballsrew in then pull the screw out.
Remove the wipers and gib strips.
Remove the keep plates under the table (don't drop the little ball-bearings between the gib and the adjusting screws).
Lift the little fella off.

It's the same for the Y axis. Should take about an hour to get each off.

[fasto]

Thanks, Cliff, you've been spot on with all your help! I don't expect to need to take the Y axis off, because there isn't anything under it. The oil lines are either atop the Y axis under the table which I can get to without additional disassembly, or attached to the bottom of the table. For the moment, I'm anointing the ways with way oil using a brush.

I did have some more success and might be making chips tomorrow!

I pulled the Z/S card to check the Z index marker circuits, which looked OK, so I brought out a bunch of wires for testing with the oscilloscope. Put the card back in, and it worked, now the control sees the Z index marker just fine. Beats me. These intermittent problems can be a real bear to deal with.

Anyway, now the machine calibrates! Hooray! Adjusting the belts on the Z axis got me to 0.104" inches from the limit switch to the index marker which I think is close enough to the recommended distance. I powered it on & off and calibrated it at least 10 times with success each time, just like it's supposed to work. I still have to touch up the positions on the other 2 axes.

Now, the toolchanger is another story...
I powered up and shut off the machine quite a few times today. On one of those times, as soon as I pressed "Power" to turn on the 24V, the toolchanger went to the right and crashed into the spindle casting wrecking the tool holding springs for slot #16. Mind, I hadn't enabled the servos yet, or calibrated the machine, so the control should not have been calling for the toolchanger to move. Yet, the "move right" solenoid valve was energized, as was its relay (I believe the move LEFT and move UP solenoids should be on when no toolchange is being commanded, to keep the toolchanger at home). I shut it down and checked the personality board and cabling for any trouble finding none.

I jumped out the air pressure switch so I could run 30PSI air for testing (just enough to move the toolchanger). I ran numerous left - right - up - down operations from the ATC diagnostics screen with no trace of trouble.

I did find out what the extra 2 limit switches are: If the Z axis isn't almost fully UP (extra switch #1) and the toolchanger moves off the second LEFT limit switch (extra switch #2), it E-stops the machine. Presumably the previous owners had trouble with the toolchanger as well and fitted these switches to prevent a crash?
I will probably fit another limit switch (extra switch #3), to disable the "move right" solenoid if the Z axis isn't almost fully up.

The toolchanger also couldn't possibly have been used as it was adjusted. The shock absorber/end stop for "move right" was installed in such a way as to prevent the toolchanger from moving right far enough to make the "right" limit switch and also engage the tool in the spindle. I supppose the previous owner decided to abandon the toolchanger and change tools by hand. I'm not going to stand for that, of course!

I adjusted the shock absorber, and reset the Z target height for toolchanges. Upped the air to 80 PSI (can't unclamp the spindle with less, those bellevilles are strong). Ran a bunch of automatic toolchanges with no trouble at all. As I said, I'll probably install a limit switch to disable "move right" if the Z axis is out of position in case there's another control glitch.

Now, the only toolchanger issue is that the splash shield mounted "rotate carousel" button doesn't work correctly. Pressing it just gives the Geneva mechanism motor a brief movement, not a complete rotation to move to the next tool. The soft keys in the ATC diagostics screens work great.

If there is another toolchanger glitch, I may just replace all the involved TTL logic on the personality board. Presumably the control's 8086 CPU reads the logic on the personality board directly as I don't recall another CPU on it (the axis drive cards have a separate 8032 CPU for each axis). The Multibus backplane is working correctly or the entire control would probably not work, so there must be an intermittent within the personality board. I suppose that a single bit error on the Multibus could result in both intermittent Z axis index marker and odd toolchange operations, but I think that's unlikely given that the control runs properly. A single bit bus error, even an intermittent one, would probably cause memory corruption and make the control crash or hang (given that there's no ROM on main CPU card and all the program code is loaded from the memory card as the program executes).

We are getting close, I need to add to my supply of CAT 40 tooling now!