Contest for the most insane member (They will be a Maho owner) - Page 3
Close
Login to Your Account
Page 3 of 3 FirstFirst 123
Results 41 to 56 of 56
  1. #41
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    For reference, my control has a EXM-2A "SSD" already, though it really is just battery backed SRAM that serves as it's main HDD. This little card attaches directly onto the EPC-8 and the BIOS is tweaked to be able to work with it from factory as long as the proper BIOS settings are changed. I'm curious if the newer Solid state HDD is much faster than the SRAM unit in my control now? My SRAM HDD now is as fast as RAM from it's time right? I suppose even then the entire system is slower due to the lower bus frequency? My SRAM HDD card has it's own lithium battery independent of the main PC CMOS battery.

    My I/O cards which have 80 I/O on each are each plugged into the VME bus. These are separate like the motion control boards.

    When the control first boots you can see it load the drivers for both those sets of boards so it is likely feasible they could be made to work with any sort of VME PC that is running the whole unit? There are versions of this exact control that do run off a conventional HDD so the ability to configure this style of VME computer to do so is already in the existing BIOS, getting the Heidenhain control software to actually recognize it is another thing entirely. Last time I starting looking into this I decided time may be better spent getting the machine finished and start running it for a while to determine if the control needs upgrading at all.

    It is true the newer controls are much faster but at the end of the day I will take a much more rigid machine that runs a little slower than something fast and flimsy any day. Right now I machine 4140 or harder almost exclusively and the vibration reduction from the rigidity will allow me to run faster and more aggressive than my one already pretty quick VMC I have. Sure the control is fast on that machine and so is it's spindle/rapids/toolchange, but the overall lower rigidity forces me to slow down the feeds and speeds to get good tool life and nice finishes. This big Deckel Maho will allow me to multiaxis drill in ways no other machine I've run could imagine. I've been using Kennametal KSEM drills lately with through spindle coolant and with a rigid setup the combination will be unreal. They really are amazing unless your part starts to deflect from the thrust force because the machine isn't rigid enough.

    Then they break, and the KSEM bodies are NOT cheap.

    Spindle Taper Upgrade:

    I am looking at converting the machine spindle to HSK100 though. Not only does it appear to be possible it doesn't even seem to be very difficult on my specific machine due to the design of the spindle and toolchanger. There exists already a manufacturer that builds a magazine chain for HSK that looks identical to the chain style my machine uses (I wouldn't be surprised if they are the OE supplier to DMG for the chain I have they look so similar). Also, the spindle is two main shafts internally, the spindle is driven from behind with a coupling so changing the taper is only a matter of changing the spindle cartridge to one that was designed to interface with the existing parts. Almost all the third party suppliers to the spindle internals also make HSK100 parts now so it really is a matter of preparing drawings and opening the wallet.

    Because the spindle taper change is not physically that invasive as compared to linear way replacement I'm once again going to get the machine fully functional before I consider going that route. I think to save on production costs at one point they stopped making a special housing for the 40 taper version and used the 50 taper housing for both horizontal spindle versions. They simply used a thicker walled cartridge to hold the 40 taper spindle assembly. This is important because I have more than enough room and rigidity to put in the HSK100 cartridge and yet I can keep the 12,000 rpm nature of the spindle intact, which the HSK100 design can handle (the SK50 spindle is designed for 6300 Rpm max, 664 Ft lb peak @ 44% Duty Cycle). I will only have 354 ft lb peak @ 25% but I don't really use much in the way of large diameter tooling much anyway. It is interesting to note that between 500 & 1000 Rpm because the optional 12000 RPM spindle can change the winding orientation from delta to wye it has more torque than the 50 taper does, as well as more torque right to it's max rpm. That is interesting, the 50 taper has a two speed gearbox and quite a bit more mechanical complexity to it too. Must be a much higher end motor on the 12k version. I don't even know the last time I ran a tool lower than 500 rpm? Only my touch probe

    I would much rather have the higher RPM limit as well as mechanical simplicity. The spindle taper change can probably be a thread in of itself when the time comes.

  2. #42
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Update:

    Machine castings are all bolted together, now I'm working on the electrical system. I am surprised by what I have found by going through the schematics... I have redrawn probably 25% of them in CAD just for ease of navigation. Anyway, unlike other industrial PLC controlled machines much of the logic used on this machine is not in the control. They have made a significant portion of the logic with physical external relays and other external devices. A great example of this is how the switches work for the doors and panels which prevent the machine from starting. All of this is controlled by normal relays and these PST3 units made by Piltz. If you don't like how "annoying" the door switches are on these machines it is INCREDIBLY EASY to change this. On a Fanuc you need software and to interface the control to modify the PLC ladder to change how this works. On this CNC 532 based machine, you only need a screwdriver.

    I'm not going to change the logic on my machine until I have run it for a while the way it was originally intended. If some of the switches are an annoyance it will be easy to change how they behave and not have to bypass them entirely. I cannot be sure if all Deckel Maho machines are like this, but I think the odds are very high that they are.

    Another circuit that is completely external is the E-Stop circuit that requires you to push and hold the button that starts the hydraulics until the machine is in a ready state. If you want to change how that behaves so it holds the button itself for a preset amount of time and then cuts out in the event of not-ready state, no problem. I remember early in this thread speaking about how I wanted the MIPS software for this control to be able to change how the door limit switches work or I would have to use an Arduino or something instead. Neither are needed.

    If anyone would like to talk to me about this further for their machine I'm sure I can help. It is too bad the ATC and APC logic is internal to the control otherwise I could put on a much more modern Heidenhain control without much effort needed on the control side. Still, it is much less work with the Indramat drives having all the specific servo tuning data right on them and much of the logic outside of the control. Installing and commissioning a newer control with all the old parameters on hand would likely only take a day or two while still keeping all the machine specific axis compensation data.

    Anyone have a blown up machine with a good iTNC Heidenhain on it? For now I'm just trucking along according to the original plan. A control upgrade will require very little wiring outside of the control enclosure so I don't need to do it now.

  3. Likes BugRobotics, Kaszub liked this post
  4. #43
    Join Date
    Oct 2005
    Country
    NORWAY
    Posts
    830
    Post Thanks / Like
    Likes (Given)
    294
    Likes (Received)
    170

    Default

    That is interesting stuff with the logic outside the control. At first I thought you wanted to buy a new Heidenhain control, and I though "That will be expensive!" , but a used new'ish one might be close enough I suppose.

  5. #44
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Hi Panza, replacing all the drives and control with servos would be really expensive but I'll bet it would be way faster than what I'm doing now.

    So I've been getting really familiar with the finer points of capacitor selection. I've spent a solid week dismantling and recording the values, types and sizes of every single capacitor on the entire machine. Every... Single... One.

    Some initial insight: Almost all the capacitors original manufacturers have been bought out, and the original types are discontinued. The idea that all new capacitors are automatically superior to the originals does not correlate with the detailed data in the spreadsheets I've been finding for all these original components.

    A couple examples:

    Siemens (Big DC bus caps) went into a joint venture with Matsushita way back and then EPCOS took it over. Then TDK took over them.

    Philips components were taken over by Beyschlag Centralab (BC) components which then were bought out by Vishay.

    An example of how a series is sometimes hard to cross reference:

    Philips (BC Components Vishay) offers these little caps that are all over the Indramat products:

    20180212_052909.jpg

    These are Philips 036 Low leakage caps. Almost no capacitor that is stocked with any of my local distributors has leakage as low as these units do (.006CV+3µA after 1 min, .001CV+3µA after 5 min). The Nichicon UKL units are very low (.002CV after 1 or 2 min depending on case size), have better ripple current ratings and better life ratings but I'm not an EE and I want to replace all the capacitors with units that are a close a possible to hopefully prevent complications. When I have looked into what others say on capacitor upgrading I keep hearing that increasing the voltage rating is no problem but I'm finding you get an instant increase of your leakage current unless the original wasn't a particularly low leakage unit to begin with and the new ones have less leakage factor. I would love to have all the circuits drawn out and learn how to calculate everything I need but I just don't have time to go to such extents.

    This question is for the pros: The UKL caps I mentioned before have really low impedance as compared to the originals. If a cap has really low impedance I don't know if this can cause higher inrush current that could damage the circuit, or what values could possibly negatively impact any circuits that are tuned somehow? How critical is low leakage in computer processing type circuits like the controlling side of the drives? I assume leakage will directly affect what voltage the cap will stay at in operation? Does it matter all that much with Electrolytics having such wide tolerances that change with age and temp?

    I'm going to replace the DC bus caps with the largest capacitance I can fit in the same spot with caps that have higher ripple ratings (sometimes much higher) and lower ESR. I don't know if I go too low on the ESR if it could cause a resonant voltage problem or something in an inverter DC BUS application?

    I saw a photo of a newer Indramat board that appeared to have Nippon Chemi-con caps in the place of the 036 Philips units, but I can't make out what series they are. Fanucs are pretty rock solid and they use Chemi-con caps so I like to use them when I can. If you have low quality caps, you have low reliability electronic product.

    Any thoughts on this would be quite appreciated! This is a ton of work but when finished I want all the electrical to be problem free for a couple decades if possible.

    I am particularly impressed with the Nippon Chemicon KZN caps. The seemed to be filled with magic and still have a wide range and great stock locally. The ripple current capability and life expectancy is a whole order of magnitude better than most others I've been going through.

  6. Likes Panza liked this post
  7. #45
    Join Date
    Dec 2011
    Location
    Ohio-USA
    Posts
    17
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    0

    Default

    I never have - nor will I ever be a Pro
    Don't remember any big deal with water migrating thru tin plated IC leads. Do remember problems when some bean counter decided to use aluminum leads internally from chip to external leads. Thought those ICs had all self destructed within few years of manufacture.
    Looked briefly at a Indramat DC supply and noticed it did not have any soft start feature but did have healthy diodes. Figured they were counting on the diode surge rating for inrush to the filter caps. Seemed as if that was the typical way euro design was done. ie. no pre charge - no soft start ,just close the switch and shoot the juice to it .
    Also noticed Indramat sold a add on filter module that was just filter caps connected to the 300 volt DC buss. They even sold a add on AC line filter ,which I thought was their way of inserting extra AC source impedance for start up inrush. To me a pre charge for the DC buss is not a big deal.
    As to the circuit boards and their frequency response ---- The limited stuff I have seen,if it was running at 10 Kc there would be racing stripes attached.. Those boards were pretty much hand drawn and very high frequency Bd. layout wasn't normal. Just look at some of the early microwave stuff from that era .

    Wonder if "Mr Carlson's Lab" on u-tube might be useful to you.. He has video showing the repair of a servo amplifier . Plus he is from your side of the border.

    Wish to Thank You for bringing us along on this-your MAHO journey .
    Hope you keep posting

  8. #46
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Thank you for your post. I thought this whole project might more "likes" let alone feedback so it is nice when it comes.

    I ranted for a bit about the electrical stuff in another thread under the general CNC machining section but I think I'll bring it all back here though from here on. I wanted some quick answers and this particular area of the forum (or the dumb name I titled this thread) doesn't bring as much activity.

    I also saw that Indramat offered that bus capacitance upgrade. I want to add it myself since I'm replacing caps anyway, I get mixed reviews as to whether or not this will mess with the drive... I'm starting to think if I ran a decent "reactor" to the drive it may help a little with inrush? Fanuc drives all use reactors leading to them but then again this may be more for noise suppression than try to reduce inrush current. Maybe it is for both? I don't know enough about how the input side of the drive works to know if an inductor, and thus drop in available energy when the load is high, will cause a momentary low voltage condition that will mess with the drive feedback electronics. I assume the low voltage automatically translates to less inrush because the overall current drops?

    Anyway... I'm going to change how this machine is wired on the mains side. That part of the electrical is all in CAD now and frankly I'm finding the machine on 380V 3 phase to be a pain that I don't need.

    With some rewiring I can run the mains on 230V 3 phase, which is what I have. The autotransformer that came from the 380V 3 phase mains and powers the 230V 3 phase inputs of the Indramat servo drives can be rewired to supply the spindle drive instead and the mains can go straight to the Servo drive power supply. This gets tricky with some of the smaller transformers in the cabinet that create the 110V single phase for example but I can just pull them and use some transformers I have from Fanucs that have 220-230V as an input. Most if not all motors can be wired to run either so I only have to change out the gauge of wiring in a few select spots and reset some current limiting breakers (or change them if they are now undersized).

    I have two phase perfects that are rated at 20Hp each but they are the older technology and aren't supposed to be connected together like the newer ones can. The people at Phase technologies told me this has to do with regenerative energy being fed back into the grid and the input side not being able to share the load properly.

    I've been looking at how they are made as compared to the spindle drives since I'm so deep in this now and the suggestion came up of connecting the DC buses of the Phase perfects. I'm thinking the control circuit doesn't need to be redundant if the bus is common so if it is connected such that only one board is feeding both sets of IGBT's and both inputs and outputs are connected the same way it should be able to handle the extra energy. The other option which is far less risky and I'm more likely to run with is to run only the spindle drive off one phase perfect, since the spindle already has it's own complete power supply, bus and drive. At first that seems like a pain to newer Fanuc and Siemens with a common bus to all drives including spindle but in this specific case it may be a blessing... the spindle of this machine can draw more power than everything else at max combined. Increase the capacitance to the drive, run it's own PP and now I have a really solid inrush damper when you get back to my single phase main input. Phase perfects can take big current for 4 seconds and two of them can take double, since the spindle is exclusive from the drives I can run at max rapids all the while bring the spindle down from max RPM without worry of tripping the converters.

    I'll send more photos once all electrical parts are ordered and I'm back to working in the shop.

  9. Likes Kaszub liked this post
  10. #47
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    I'm back in the shop and here is another quick update.

    HBOB - Indramat drives come in so many flavors I wonder which type you were looking at? The units that do not have internal protection for inrush are supposed to be attached to a large inductor on the input terminals, at least the units I see in the drives power supply manual I'm looking at.

    For my machine, the KDV has several levels of protection for inrush. It has a current limiting DC bus choke which allows me to add capacitance up to 8mF without any further modifications.

    On the Fanuc Alpha system on my other machine I decided to dig into this further. I want to take care of as much electrical as I can within reason while I'm in this mindset. This system is designed in such a way that if you add capacitance to the bus it is through a module that monitors the existing bus voltage. Once it gets to a certain value it then permits the additional caps to start charging. In this way it reduces inrush to a couple big steps instead of a giant metal melting moment. Once they are charged all should be as before just with more of a reservoir for braking than before. Having said all this, the power supply does have an input reactor to help with inrush too.

    What I love about the drives and spindle setup on my MC800H is they don't use complete regenerative braking, but it does have regenerative braking. The braking bleeding resistor has the capability to take in 40kW of power just for the servos! If the power goes out the system is designed to come to a stop with internal bleeders while using the internal bus power to back the tool away from the part before the machine fully stops. On the Fanuc these are all add on options that are in the form of several modules.

    Back to inspection and cleaning:

    The before photo: Bottom of the cabinets, covered with nasty black stuff from the cabinet doors being left open...

    bottom_before.jpg

    And now. I've pulled all the output cables from the terminals and cleaned in the cable trays. Somehow a couple of them had a bunch of oil in the bottom! I have found many insulation breakages and even a couple badly damaged wires. A few ferrules that inserted wrong in the terminals as well as a bunch of modifications that have been done over the years to bypass certain functions or maybe just keep the machine going? I've already repaired all these changes. If I need to take out certain functions of the machine I will not be doing it that way, I will do it much like the OEM would have.

    bottom_after.jpg

    It is coming along well.

    On the issue of the capacitors, I decided to use the life rating of the original caps to determine which I was going to replace. Many of the really large units were hard to find replacements for and had lifetime ratings that were at least 4 times what the lowest rated caps had. I will preform the bus caps and check them with the new ESR meter to make sure all is well. If any are bad they will be changed as well.

    After all the discussion on my other thread in regards to the drives as well as the recommendation to "Mr Carlson's Lab" I decided to check him out as well as EEVBlog. Those guys are very helpful! I feel like my knowledge is right at the point where EEVBlog is easy to take in and not too dumbed down, while not being too complex either. Both those guys are easy to learn from. I decided to start mapping out some of the circuits that the caps are used in. After reviewing the datasheets on many of the IC's in use it became much clearer!

    Those low leakage caps are not only low leakage, they are pretty low ESR for their time with decent impedance values. For some of the applications in this machine they were being used as output caps for regulating modules and the datasheets on those laid out the most important factors to consider with the capacitors being used with them. Lower ESR is more important in that application within limits which they define, so I went that way with the replacements.

  11. Likes BugRobotics, Kaszub, Panza liked this post
  12. #48
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    As I'm redesigning the mains side of the MC800H, I have been reading the appropriate documentation from all the units that attach to them to make sure all fuses and breakers are set as the manufacturers intended for these components.

    I came across something for the Indramat drive power supplies that restricts the inrush to the bus caps, there is a signal that comes out of the drive when the bus is fully charged, you can use this to default a pair of contactors to reroute the power through some power resistors to restrict current at power up. After the bus is charged the signal changes and the power is switched to direct removing the resistors from the circuit. This may be what was intended for the supply that HBOB was looking at. It applies to the KDV 3.1, TVM 2.1 and TVM 2.4.

    I spoke with the manufacturer of the mains suppression filter I have which is an inductor with X and Y caps on both sides of it as well as resistors to drain the caps on the output side. They say I can get away with the lower voltage as long as I don't exceed the rated current and the current overload ratings of the unit happen to match the Phase perfect exactly (4x for 4 seconds/startup). They aren't sure if it will filter as well as before but as far I can tell this is more for preventing the noise from getting back into the mains from the switching of the drives power supplies and the Phase perfect does this already anyway.

  13. #49
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    I'm at a bit of a crossroads here with the Deckel Maho electrical system design in regards to safety.

    It uses a mains filter that leaks up to 31.8mA of current back through ground while in operation, if two phases are interrupted it could leak back 6 times that. It calls for an isolation transformer in the schematic that has the secondary neutral attached to ground. All of this seems potentially dangerous? If you simply powered this machine up what is to stop the filter leakage current from deciding to pass through the operator at any random time? It even says on the unit that if you don't hook up a really good ground to the machine before you power it up you are very likely to be shocked.

    Other machines I have do not use a mains filter with caps that leak to ground, instead they use line reactors and a switch on the door that can detect ground leakage and shut the machine down. One time I had coolant get into a servo and the main switch tripped. That could have been a dangerous situation and since then I have really appreciated that way of designing the machine.

    With the MC800H as is, this type of switch would be tripping the moment the machine was started! My CNC equipment now runs straight from the Phase perfect. The fusing is on the input side of the PP via a fusible disconnect. I looked at the documentation for the PP and it can only sense a ground fault on the generated leg as far as I can tell, so you can't rely on that to help you. The PP does filter the noise so it doesn't feed back into the single phase line so the function of the suppression filter is already being performed, at least as far as the input mains power is concerned.

    The Deckel Maho does not have reactors between drive power supplies in the cabinet like my Fanuc controls do. I do think adding them to the inputs of the spindle and servo drives would help the rest of the machine inside have less noisy power? I would love to hear what others think about this.

    If I don't need isolation transformers I really don't want them because of the cost and the energy they waste. I would need two because I cannot connect the outputs of the two phase perfects. Phase perfects do not act as isolation transformers, the question remains as to whether or not isolation transformers will add much safety when they have the wye neutral bonded anyway? I've run machines directly off of PP units directly from the panel for a long time but that doesn't mean it is the safe way to do it.

    At this point I'm thinking of ditching that big 80A mains filter and adding the same type of mains switch used on my Fanuc machines. I'll put in reactors before the drives internally to help with filtering and and trust the PP will take care of the rest, worst case I'll add a single phase filter right at the main panel, this way any leakage to ground will be going directly out to earth. Hopefully this approach will make the rest of the equipment in the cabinet get cleaner power at the same time?

  14. Likes rotarySMP liked this post
  15. #50
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Working on the automatic pallet changer:

    Before image during disassembly:

    20171011_145137.jpg

    And progress as of yesterday evening:

    20180307_174339.jpg

    Found some more broken parts, not affecting function but I'm going to repair it anyway. Once assembled many of the components on this assembly are very difficult if not impossible to access.

  16. #51
    Join Date
    Oct 2005
    Country
    NORWAY
    Posts
    830
    Post Thanks / Like
    Likes (Given)
    294
    Likes (Received)
    170

    Default

    Looks like it is coming along nicely!
    It is kind of hard to see what part of the machine when you are working on, is it possible to take some pictures that gets the whole thing in the picture ? What sort of things are broken in general on the machine?

  17. Likes Nerv liked this post
  18. #52
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Hi Panza,

    No worries, yesterday was tough getting that big tool changer plate into position with the cramped area it is in was very challenging:

    20180313_235314.jpg

    I'm not quite finished cleaning that plate so please excuse the mess.

    Here is what the side looked like before:

    20171005_180055.jpg

    And now, with the ATC plate mounted again...

    20180313_235327.jpg

    I only use brass to clean the ballscrews and linear ways so a little bit of staining and pitting is visible. I would much rather keep as much accuracy as I can than make it pretty with abrasives.

    The roller ways move extremely smoothly though with almost no effort at all, this is surprising when you are pushing 6000+ lbs with a single hand.

    Pallet changer again, but from further back:

    20180309_031800.jpg

    Please mind the clutter again, better a productive mess than tidy idleness right? I was using the pallet changing surface as a table as I cleaned the unit. This was a few days ago, the whole unit is pretty much ready to go now except a little electrical. All the wiring was inspected and cleaned as well, ultrasonically cleaned the limit switches, etc etc. I'm holding off mounting it to the machine until I'm finished with the trunnion hydraulics, coolant hoses, wiring and cleanup.

    I've been in touch with Heidenhain about putting a new control on the machine. The good news is the newest TNC640 has modules that can be inserted into the control that allow for analog servo drive control. Although the Indramat drives for my machine had optional serial communication, mine do not have it and are +/-10V analog both axis and spindle. The feedback is already Heidenhain scales on all axis so I can reuse all my drives and servos and most importantly, the spindle.

    If my drives had been digital (and not Siemens or Heidenhain) I would not be able to use the new control, so this is good. We are working together to get the internal programming portion of the job minimized. This machine has a bunch of extra functions I would like to retain so collaboration between Heidenhain and Deckel Maho is important to reduce the amount of time this will take to do properly. I will have to do some modifications to the control housings on both the main station as well as the back cabinet as well.

  19. Likes Kaszub, Panza liked this post
  20. #53
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Just a quick update:

    Here is a quick summary of the process of effort escalation that happened on the MC800H over the last several weeks.

    As you all know I went through a bit of the electrical and the pallet changer. After reading through the process of dialing the machine axis to the trunnion I decided I would make sure everything was straight before entertaining the idea of reattaching and aligning the pallet changer.

    Step 1. Check pallet (Table) perpendicularity to it's axis of rotation. Result: It is way too far out.

    Step 1.1 Take off pallet and inspect pallet carrier. Result: Alignment rings sitting at the wrong heights, various troublesome signs.

    Step 1.1.1 Take apart entire pallet carrier assembly including pallet release components and actuator, optical rotary encoder, pallet carrier clamping elements, etc etc.

    At this point I had a nice flat scraped surface which I could use to at least confirm the trunnion itself was properly aligned to the X axis. Result: Way too far out as well.

    Step 1.1.1.1 After inspection of the trunnion there was a little corrosion that was difficult to access. Result: Remove entire trunnion, building custom tools for unloading the gearbox bias as well as the optical rotary encoder, translate from German the special process to pull the bearing to facilitate removal...

    And although I'm still not really even done what I started to do, at this point I'm facing this:

    20180405_032949.jpg

    So after some cleaning...

    20180405_162833.jpg

    ... the trunnion was ready to be put back together. On this machine, the trunnion bearing attachment is non adjustable on the driven side. Also, this is newer version of the original design, they removed the clamping elements from the servo driven side in favor of heavily preloading the gearbox and using the servo brake on that side instead. This would be fine though as the gear reduction is so high and the preload so high that it basically can't move at all once the brake is engaged anyway. The only thing I don't like about this version rather than the original is that the gears themselves are taking the machining vibrations instead of clamps. You can tell by the shape of the castings that there were brakes on that side at one time. I have older diagrams which show that design as well.

    Anyway, the side opposite the drive gear can be adjusted up and down, front to back on EITHER SIDE of the bearing. So you can move the trunnion if you determine the trunnion isn't parallel to the tilting axis but the bearing itself is in the right position relative to the X axis, or you can move the bearing if the tilting axis is off.

    This adjustment took me 4 days. Every single time you adjust the bearing or trunnion it affects the other three measurements. I was aiming for at least as good as DMG recommended so I had to get better than .02mm (.00078") across 600mm (23.6"). To change a value you have to set up dials to catch any motion in either direction and have to torque the bolts to 200 ft lbs before you take your final measured values. Every check involves sweeping the machine full stroke across X with the indicators and because the machine doesn't have servos on it yet, everything was done by hand, or with overhead cranes (lifting the trunnion to 90)

    I wound up at a smidge better than .0001" over 600mm.

    Still not even back to getting the table dialed (The original goal) now comes some other surprises...

  21. Likes Peter S, BugRobotics liked this post
  22. #54
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    The pallet carrier has some issue of alignment that can't be properly diagnosed until it is properly reattached and everything is both spotless and properly torqued as well as lubricated below it.

    The process of getting the carrier itself back on involves a large bearing that remains attached to the trunnion during the process but must be aligned properly and uses special ground thrust washers on each bolt hole but you can't insert the bolts until it is assembled, and the thrust washers much go on prior to assembly. These washers can fall off or move at random but can only have one bolt inserted at a time through a small inspection hole. This a process of patience and mental torment that I'll just sum up as "Thankfully completed".

    So next is putting the braking elements back on. This should be simple enough, little units that have some springs in them that "let go" when hydraulic pressure is applied. Sounds very robust and pretty fail proof, I made a simple image in my head of how this likely looks and proceeded to start cleaning and testing the units before reassembly.

    But I was wrong, this is German.

    I applied pressure to one of the clamping elements that behaved a little different than the other five when I was taking it apart. I thought, oh, it might be a little sticky or something from lack of use, I'll just take it apart and have a peek.

    Well these units are made by OTT Jakob and just taking them apart without damaging them was challenging. I wound up using a big three jaw chuck with softjaws to grab the body hard enough to get the caps off without damaging the bodies. I was pleased to see the design was very clever in that the hydraulics were only in the cap, and the rest of the body did the clamping. So far so good, but then I pulled the clamping side apart:

    20180415_180419.jpg

    Those belleville washers are now a whole bunch of broken rusted chunks. There is a slot in the body of the clamping element that allows hydraulic fluid to escape in the event of seal seepage or outright failure. This prevents it from pumping hydraulic fluid into the area the clamping shaft extends into and it also prevents the entire unit from getting pressurized and possibly exploding. Either way as a bonus it also allows coolant and moisture into the unit to cause destruction and mayhem.

    The German part of this that I didn't take a photo of is how the release actually works on this unit. The hydraulics actually push on ground pins that go through the plate that "pulls" on the clamp shoe. Instead of the hydraulic piston pushing against the belleville stack it pushes against a hardened ring with a chamfer on it which pushes against a "ring" of precision hardened bearing balls that then push against a wedge which then applies force to the spring washers to release the brake. This allows mechanical advantage to the piston so the unit can clamp harder than the hydraulic piston can actually push. When you look at OTT Jakob's product lines and patents you can see they had a real fancy for making designs like this, the spindle drawbar in this machine also uses a variation of this idea, also made by OTT Jakob.

    So far I've taken apart three of the 6 clamping units and they all have failed springs. If any of you have equipment with these style of clamping units it would be easy to assume they are maintenance free. What is most unfortunate is this is one of the few systems on the machine which doesn't have actual mechanical feedback to notify the machine if they are not clamping properly. It does know if the hydraulics have been applied to them to release the brake but that doesn't mean the brake actually will grab again once the pressure is released. In this case, at least one of the units wasn't clamping at all when the pressure was released, and at least a couple more were only applying a fraction of the original clamping force.

    Stay tuned!
    Last edited by Nerv; 04-17-2018 at 01:56 AM.

  23. Likes Peter S, BugRobotics liked this post
  24. #55
    Join Date
    Jun 2015
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    216
    Post Thanks / Like
    Likes (Given)
    589
    Likes (Received)
    82

    Default

    Thoroughly enjoying this thread. I love the analysis and pics of the damaged/dirty assemblies. Excited to hear about the new Heidenhain control. Thanks for sharing!

  25. Likes Nerv liked this post
  26. #56
    Join Date
    Jun 2016
    Country
    CANADA
    State/Province
    Alberta
    Posts
    86
    Post Thanks / Like
    Likes (Given)
    20
    Likes (Received)
    44

    Default

    Thanks for your positive feedback. It is great to hear that members are still following this project and find this of interest.

    I just bought a DMU50V so all the experience gained thus far will help keeping that machine running well too. Same Philips VME based control computer, only Siemens digital drives instead of the Indramat stuff in the MC800H, and "Millplus" instead of "532" though time will tell how different they really are. I think they are really similar.

    We are holding back on the control upgrade for now until we can prove out the whole MC800H as it was originally built. It may not be needed, I am considering changing the machine spindle to HSK 100 which earlier in this thread I covered a little. The cartridge for the spindle is separate from the motor which is directly behind it. This will make replacement much less expensive and easier to implement. Time will tell if the rigidity limits of the Cat 40 will necessitate this upgrade, I think long term it may be the best upgrade for the money of all the mods I can do to the MC800H.

    More to follow soon!

  27. Likes craigd liked this post

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •