What's new
What's new

Resurrection of a lobotomised Maho 600T

Yes, it would make sense to straighten it before evaluation or starting reconditioning. I was just having some thoughts about how to do that. Aboms video shows how to do it very well. Personally, I could use the quill (with DRO) on my toy mill as the press.

Now, I have no intention of reconditioning the gib or other slide surfaces at this moment. But it's interesting to me to know a little bit about how much wear is where. In case I ever want to give a shot at improving the accuracy in the future. Prior to starting dismantling of the machine, I cranked the table from side side side by hand. One side measured 0.01 mm low and the other 0.05 mm low compared to the center of the table reference. So that gives max 0.05 mm error over 600 mm travel, which is way good enough for me.
A bit more progress. I started with the wiring of the cabinet. The DC power supplies and the emergency stop / reset circuit are wired up, but not much else.

Today I also put the cross slide back on the saddle(?).

I replaced all the copper washers for the lubrication lines. While doing that, I noticed that there appears to be a sheared bolt on the saddle (top right on the photo below). I don't recall seeing or at least not paying any mind to that before. I don't know what it's for right now, and it doesn't appear to be shown in the spare part catalogue. My guess is that it's nothing important. And if it is, I'm sure it will come back to bite me in the ass later. One thought was a mechanical end stop, but that can't be it. If it would protrude by more than a couple of millimeters, the total travel length would be shorter than what is stated in the manual.
Below Adam Booth doing that. He deemed scraping not necessary for his purposes.
That's a hack job, its an embarrassment.

Your gibs may not be perfectly parallel so flip the gib over and measure the scraped surface as it bears on your precision parallels
I cleaned up the x-axis ballscrew and end cover and installed them on the machine.

Just for my own interest, I measured the vertical play in the x-axis slide by in turn lifting each end of the slide by a car jack. This is corresponding to the jib I measured a bit on a few posts back. I pumped the jack until the machine base started lifting, so quite a bit higher force than the machine should ever see in normal operation. The play was measured at the end of each side of the saddle/knee slide. On the non-driving end, I measured approximately 16 µm. Given that the manual says 5 µm is normal/allowed, this is quite well in line with the measured 10 µm wear on the gib. On the driving end, I measured 24 µm play. In that end the gib appeared to have less wear, so not sure why the play in more there. In any case, I'm not terribly concerned about this amount of wear/play. And I think almost all of it is in the gib, so it should not be a terribly big job to recondition if I ever want to do that. Provided I learn a bit of scraping first...

I mounted the servo and have moved the axis under power. With some initial tuning of the servo and LinuxCNC PID loop, I was able to get down to a maximum measured error of 15 µm during a linear move at 1000 mm/min and acceleration of 500 mm/s². The maximum error is at the start of the move and most likely because of stick/slip before the slide starts to move. At the end of the move, the maximum spike was around 5 µm. I believe this is decent performance, and hopefully I might still be able to improve on it.

We discussed earlier regarding the emergency stop and potentially delaying the power off of the servo drives to allow the drive to decelerate with power. I have done some testing, and it appears that the delay in the contactors and capacity of the servo drives DC bus capacitors make a stop by simply opening the contactors quite slow. I put one electrical end stop closer to the middle of travel. The travel past the end stop with 3000 mm/min was as follows:
Hardware only stop: 16 mm before stopping
Hardware + input to software e-stop: 4 mm before stopping
So there is no need to delay power-off of the drives. On the other hand, simultaneously triggering a software e-stop significantly reduces the stopping distance, and will be required to avoid hitting the physical end stops at high travel speeds.

I believe it will be several weeks until the next progress update, because I will again be travelling for quite some time.
Deckel CNC’s have a ramp down cam (physical switch) that slows the rapids when the slide approaches the end of travel.
So you can’t rapid to the end of travel.
There is also a physical switch for the end of travel, plus a switch for overtravel (E-stop) all 3 axis have these three switches in each direction.
Cheers Ross
That is some interesting information, thank you. But not something I will use. There are software limits in LinuxCNC (and all other CNC controllers I would presume), which will be the first layer of protection against overtravel. Potentially, it takes the acceleration (deceleration) into account to prevent the machine from exceeding the soft limits even at high feeds / rapids.

After all is installed and set up properly, the only reason I can think of for reaching the limit switches is if the linear scale loses pulses in one direction of travel. I can think of two ways to safeguard against this in the controller:
1. I have reference marks in both ends of the linear scale. After homing, I believe I would be able to check that the position feedback is roughly what it should be at the reference mark. And with me, I of course mean the controller.
2. Since I use step + direction pulses, the real position should "stay in sync" with how many pulses have been sent, give or take 1 mm or so maximum. So if how many step pulses (in each direction) has been sent does not line up with the position feedback, it could generate an e-stop.
After a long time (mainly due to a lot of travelling), here is the next progress report.

I pulled the ram off the machine for inspection and cleaning.
This is the most wear I have found so far. It's not great. Towards the front of the machine, the flaking marks are all gone and the surface is scratched along the way of travel.
On the ram, the flaking marks are still visible in all locations. But only just barely in the locations with most wear. The angled part of the dovetail slides are in similar or slightly better condition, but difficult to get any decent photos of those.
The lubrication ports have spring loaded ball check valves for the ramp. There was a lot of crud in those. That type of check valve was not present in the lubrication ports on other parts of the machine. I wonder why Maho put them on the ram.

While it would be nice have the sliding surfaces reconditioned at least to some degree, it's nothing I'm up for right now. So I cleaned everything, stoned the surfaced (with precision ground flat stones) and put everything back together.

I also installed the servo motor and brake assembly for the vertical axis, after some smaller modifications. The main parts were already made by the previous owner. I polished the fit between the servo motor and timing pulley shaft, because it was pretty much a light press fit when I received it. And I made a hole so the key set screw can be tightened. The brake cable was damaged close to the brake, so I turned a sleeve for that and epoxied it in place.

I have been working more on the electrical side, and the control cabinet backplate/components starts being close to finished. But it appears I reached the photo limit for this post, so I can include that in the next update instead.

The following steps should be starting to put everything back together with all the covers, auxiliary equipment and so on.

Are those box ways bolted onto the main casting?
Interesting, never seen that on a Maho before, seen some machines with bolted on dovetail sides on top, but bolted boxway rails are new to me.
Good catch!
They probably are. Take also a look at how oddly they are integrated into the main casting.I can understand box way cost and production time is greatly reduced if done separately, moreover there are various options for materials and heat treatment.I remember some Weiler lathes with box ways that are also bolted.

Yes, for the vertical axis the box ways are bolted on. In the photo linked by DeadMahoDude, the bolts are the ones where the heads are painted green.