Retrofitting a 1986 Maho MH400E

Bernd wrote: Hey y’all

So i bought today a Maho 500C, with a problem on the control, i atached a picture with the screen,wich shows the error.
greetings from Southtirol

Bernd

Servus Bernd,

Welcome to the wonder world of MAHO ownership
The error message seems to refer to the drives or motor tachos. Does your 500C also have an Indramat 3TRM2 amplifier unit? Do you have the Indramat manual? I found a pdf of it somewhere on the web, so if you need it, PM me your email address so I can send it to you.

There are fuses on the Indramats PSU card, so that might be a good place to start looking.
Mark

gernoff wrote: Mark;
As you can see your work on this project is inspiring others! Thank you. I will eventually be doing the same. I've just started a thread on converting my Deckel FP5NC and will try to include the levels of detail you have. My DC BBC driven spindle motor died and I am installing a VFD controlled motor in its place. Initially I hope to get it running with all my other existing hardware so it can handle shifting etc.

Good on you. There is a fair bit of negative information on the web on how hard it is to retrofit a Deckel or Maho due to the complexity of these machines. I was seriously considering doing a CNC conversion of a Rong Fu 45 clone like this excellent example:
www.cnczone.com/forums/rf-45-clone-mill/...g-8.html#post2005116
When the MAHO 400E showed up for the same price as the RF-45, I started thinking through the real comparison:
MAHO:
- Engineered by experts
- initial build with german precision
- high precision bearings and BT40 spindles
- rugged gearbox to match torque to tooling
- huge rigid ballscrews
- huge rigid axis motors
- highly accurate linear encoders
- integrated lube system
- industrial quality wiring and relays
- hydraulic tool release
- comes with a matching enclosure

RF-45:
- You end up dumping everything but the castings, and they are a chinese lottery.
- You have to make/buy and integrate everything on the list above yourself.

I think it is a whole lot easier to retrofit a Deckel or Maho with LinuxCNC than it would be to do a conversion of a manual machine, and the result should be much better.

gernoff wrote: The work done here on the shifting .hal is going to help immensely during an actual LinuxCNC conversion.

Sergey is the one to thank, he did the real work, I just fed him with half knowledge.

gernoff wrote: In you spare time........ I am sure the group here would appreciate a synopsis of how you went about the reverse engineering of all the relay based logic chains these machines have. Sort of a 'getting started' guide. You could call it "Deckel / Maho LinuxCNC Conversion for Dummies"

I actually started doing exactly that from day one. It is of course unfinished at this point, but if you want a copy, please PM me. Since I put the gearbox info into a wiki on Github, I have been wondering whether it would not be better to convert my "How to" to a Wiki for the whole machine.
Mark

cgroth wrote: Maybe you already wrote it but did you swap the scales of different axis (with the drives off!) to check the EXE? Christian

Yes, I have already worked through all permutations of encoders and exe's. The three axis of EXE work fine.

The X scale works perfectly.
The Y scale was dropping about 20% steps. I cleaned the scale and it worked perfectly for a while. Now it is dropping 50-80% of steps.
The Z scale gives off no signal.

I was out at a work function yesterday, but will get back to trouble shooting the Y and Z scales tonight.
Y - will be cleaned again, this time including the condensor and sensor.
Z - I'll finish removing the head from the scale, and see if the bulb works, and go from there.
Mark

Mark, did you grab the old ls403 docu from the web? It states 120 mA @ 5 V for the bulb, so that can be checked directly. As far as I understand, the sensor itself does not contain any active electronics hence it seems to be a phototransistor or similar device, that delivers a current once exposed to light. If you have a proper Voltmeter or oszilloscope you might be able to plot the current using a 1 k resistor

Yes, I did save that info. Thanks, that is a good suggestion. I have a little osci and a halfway decent voltmeter.

I don't think the interpolation of the integrated EXE's I have is adjustable. There are no jumpers on the boards. I have no manual for them though. They output the LS403 scales inputs with micron resolution.
Mark

Should be quite easy to check though. Just measure the current the head is drawing, and if there is no significant current going you can conclude there is a problem with the lightbulb.
The output is in the micro ampere range (the standard used here is 11 uA), so the best way to look for that is to use a scope or other sensitive instrument to see if there are changes when the head is moved. Possibly best done by measuring while in circuit (so they are loaded with the correct impedance).
And the scales are analogue, so two 90 degree phase shifted sine-curves should be what comes out. So a scope is definately the best instrument for checking this.

Anders