Retrofitting a 1986 Maho MH400E

Thanks for all the input guys.

That's terrible to hear about your Renishaw probe Andy. I only have a Haimer 3d Taster and have so far managed not to break any tips.

Re: the Philips control panel and pendant, reading this thread I just thought that the panel was probably the most work-intensive element of the entire retrofit. The barebones alternative is a keyboard and a touchscreen, though I was hoping to use an Eizo LCD panel I replaced the old CRT with, and that's not a touch screen.

As for the pendant, my personal pick for most useful feature that's not on the Philips is setting zero points with a button for each axis. I saw that on a hobby Stepcraft machine, though that pendant was separately purchased by the user. When it comes to using the jog buttons you can set them to continuous, which handily moves the axis by simply pressing a direction button, or to incremental, which is very handy for finding workpiece edges (e.g. on a Haimer taster you can see each press of the button move the dial 0,01 or 0,1 mm so you can very easily get it to where you want it).

Maybe it's just that I've gotten used to doing things this way but I find spinning the jog wheel for longer distances kind of annoying, especially on slower speeds (which the Maho doesn't have a high speed, at a whopping 3m/min X rapids and 2,5m/min Y and Z rapids). On the other hand, precisely moving only a 0,01 or a 0,001 seems finnicky with a wheel since they're kind of designed for fluid, continuous operation. On the Maho I press the button once and there's no risk of me pressing it more than that.

In the end, I'll just go with a simple solution that works first and then move on to making improvements incrementally. We'll be able to share our findings and get things just right in the unspecified future

Anyway, on the MH600E I find that the table is large enough that, more often than not, I can't reach the control panel easily so some kind of pendant is a must.

EDIT: guys, this is a very dumb question, but now that we have a connection with Mark, how exactly do I send a personal message? I can't find any help feature in the forum controls :blush:

You are correct that there is no pressing need to make a user control module. You can control LinuxCNC with just a regular keyboard and mouse, or touch screen. I am using wireless logitech combined keyboard/scratchpad at present, with the intension of not needing a keyboard once everything is finished and configured.

I suggest you download the LinuxCNC liveCD to try out Gmoccapy. The way that Norbert has implemented the touching off soft keys and offset's is extremely intuitive and easy to use, but I don't think there are pins defined to hook into those functions, so you'd have to get into the code to put those functions out onto your pendant.
linuxcnc.org/docs/html/gui/gmoccapy.html#_jog_hal_pins

With LinuxCNC you can switch between continuous jogging, and your user defined increments.So if you set increment 1mm per click, your jog wheel can move you across the table very rapidly.

If this forum has no private message system, you can also PM in the CNCecke, with you email so I can send the draft manual. I have the same user name there.
Mark

Thanks Mark for sending me your notes and diagrams!

I just wanted to share a thought I have after reading your file. I think that the reason Maho machines were so slow was because of the very high torque generated by the gearbox spindle. My machine has fine-pitched ball screws - 5mm in Y and Z, and 6mm in X (all are 50mm diameter). All axes have a 1:4 toothed pulley transmission and the Indramat motors are rated at 2000 rpm. This tanslates directly to the available rapid speeds:

X 2000*6 / 4 = 3000 mm/min
Y and Z 2000*5 / 4 = 2500 mm/min

You can see that later Maho models that had VFD spindles had twice those speeds. Likewise, Deckel didn't use gearboxes in the FPX NC series and their machines from that period had around 6 m/min per axis rapids; their VFD spindles had up to 6000 rpm IIRC which brought the torque down also (with the same overall design that was also used by Mikron and others).

Today, new tool room CNC machines with slide ways have around 10m/min rapids (at least those produced here in Poland). So in my opinion these differences between models are about torque and table load, rather than slide ways or machine castings (though those are obviously super important overall). The MH600E documents list an Md const. of 400 Nm and allow up to 400kg on the table.

If you changed the spindle to a modern high speed, direct drive model with a torque of 80Nm or less, I'm pretty sure the speeds could be cranked up without harm to the machine, especially with faster servos, like 3000 rpm (though the spindle would likely blow up a home fuse with its high Amperage).

Just some random tangential thoughts.

If you look at Stig over Ruud's retrofit, he ripped out everything and used modern servos, with much increased maximum speeds and acceleration.


I am a bit suprised at how lame the acceleration of the Indramat installation is. There is a obviously a lot of inertia in those large drive components, but there is also a lot of power to push it around. How did MAHO increase the rapids on the later machines? courser thread ballscrews, or higher speed motors?

If you want to do HSM tool pathes, you probably run a significant risk of galling the sliding ways of the MAHO. A linear rail machine is probably more suitable for such work.

Mark

RotarySMP wrote: If you want to do HSM tool pathes, you probably run a significant risk of galling the sliding ways of the MAHO. A linear rail machine is probably more suitable for such work.

I don't know the actual limits for sliding ways, but consider:
That did 200kph with crank pins at about 1/3 wheel radius.
So the peak speed of the cross-head was 1000 m/min....

Wonder what the MTBR was

Maybe I am overreacting to seeing too many threads with way galling like this FP2:
forum.zerspanungsbude.net/viewtopic.php?...ilit=fresser#p429909

Mark