What's new
What's new

Anilam 5000M ADvice needed before I lose my mind

Due to this new info the following ideas come to mind:

Did you unplug anything during the move?
When running code with z moves, have you tried shaking/jiggling the enclosures or cable bundles?
Have you made sure all ribbon cables and computer cards are snapped into slots etc?
Have you asked the previous owner of the machine about these issues? Did they have them as well? - If NO, then it may be move related.

I traced an intermittent error in my Fanuc to a dodgy contact block connection. It only happened when a very specific harmonic was hit when cutting. Must have sent enough vibration through the machine, but a little more vibration was ok and a little less ok too. Might explain why air cutting is ok and real cutting not?

I like the power supply theory too, but a move related problem seems easy to pinpoint if it was turned off working at old shop, and not working after move.
 
I chased a spindle drive alarm for a long time on my machine, no relation to yours, I finally figured out it was a bad cable in the chain track. It finally started happening in the same spot of machine travel is how I got to looking for that. New cable fixed it.
 
Hi thanks for the reply, Yes I have tried reseating all the cards, I've cleaned all the connections, and checked them thoroughly.

It's funny you mention the power supply, because, I have checked it and all voltages appear ok at the power supply, but I did notice recently when I went to fit a new screen that the power connector for the screen was only showing 10.49V (supposed to be 12V). I expect some voltage drop over the distance, but not that much, maybe the power supply is a little sad. that little revelation just gave me a little hope. I will probably remove the anilam controller form the machine this weekend and bring it home (somewhere clean) and give it a really thorough clean out, better than I can do in a dirty workshop. also the lighting is a lot better in my office at home to see anything out of place.
 
Hey Macgyver, you've giving me something else to look for.

I had a ton of trouble when I first got the machine, I couldn't reset the servos sometimes, and it turned out to be a damaged wire going to one of the limit switches, and I only found that because the limit switch itself had over-traveled and jammed in, I had to putt it apart un-stick it.

I know the Z-axis motor cable doesn't move with the Z-axis head, but maybe there's something else going on there, maybe there's a damaged wire that goes into the Z-axis that's causing the problem another way, maybe cause a relay to drop out or something. THe only reason I am thinking that is because, I can run Air cuts 30mm off the vice and no problems, but when I put it into a cut in a piece of stock in the vice it has problems, perhaps there's a dodgy wire that's shorting or breaking right at that particular height, and it might not even be related to the drives, but maybe it's cause a relay to drop out or something, and that's causing the problem elsewhere. I'll run it past my lecky mate on the weekend and see what he says.
 
Hi thanks for the reply, Yes I have tried reseating all the cards, I've cleaned all the connections, and checked them thoroughly.

It's funny you mention the power supply, because, I have checked it and all voltages appear ok at the power supply, but I did notice recently when I went to fit a new screen that the power connector for the screen was only showing 10.49V (supposed to be 12V). I expect some voltage drop over the distance, but not that much, maybe the power supply is a little sad. that little revelation just gave me a little hope. I will probably remove the anilam controller form the machine this weekend and bring it home (somewhere clean) and give it a really thorough clean out, better than I can do in a dirty workshop. also the lighting is a lot better in my office at home to see anything out of place.

Again, not sure how relevant this is to your specific control, but some controls are extremely fussy about function of their DC busses. I have known controls that won't even try to boot if the rails don't come up within a few ms of each other, or will lock up if the switching response is too slow or if there is too much overshoot etc. etc. Sometimes you can have good static voltages but the problem is still the power supply.

When suspicious about edge connectors a good method is to clean them with a soft pencil eraser, followed by a wipe down with some no-residue solvent of some description, followed by a very thin smear of contact grease.

If the power supply is a fairly standard affair, straight up replacing it with a new one is a very good start to troubleshooting. If it's something old, obsolete and unobtainium, a recap job can go a long way.

I still think this would be a good time to drop a short email to the guys mentioned, if nothing else just to ask them if you're going the correct route or if there's some known weak point that they can pinpoint immediately from a detailed description of the symptoms.

Hey Macgyver, you've giving me something else to look for.

I had a ton of trouble when I first got the machine, I couldn't reset the servos sometimes, and it turned out to be a damaged wire going to one of the limit switches, and I only found that because the limit switch itself had over-traveled and jammed in, I had to putt it apart un-stick it.

I know the Z-axis motor cable doesn't move with the Z-axis head, but maybe there's something else going on there, maybe there's a damaged wire that goes into the Z-axis that's causing the problem another way, maybe cause a relay to drop out or something. THe only reason I am thinking that is because, I can run Air cuts 30mm off the vice and no problems, but when I put it into a cut in a piece of stock in the vice it has problems, perhaps there's a dodgy wire that's shorting or breaking right at that particular height, and it might not even be related to the drives, but maybe it's cause a relay to drop out or something, and that's causing the problem elsewhere. I'll run it past my lecky mate on the weekend and see what he says.

Broken wires in cable chains are absolutely a thing that happen, and can have unpredictable effects. I had an encoder cable break in the chain once on a Fanuc machine causing the +5v to ground out, causing spurious and unhelpful drive alarms that lead me on a wild goose chase looking for completely unrelated faults. Like you said, it's only once I realised that it only happened at a specific point in the axis travel that I twigged it was a broken cable.

Maybe do another quick test to try your air cuts at the same Z height as the actual cuts and see what happens. You might have a ground fault pulling your 12v down for example.
 
Hi vmipackman,

I never unplugged anything, the machine had all the guarding and coolant tray, pumps, way covers, etc all removed and packed up on a pallet, but the machine for the most part was complete. I have already checked all the connections, I found a couple of wires going into contactors that weren't 100% secure or installed properly, I fixed them up, and I have traced and checked every wire, I took a notepad and a camera, and took a photo of every single connection and then drew the path and the pin numbers for every single wire in the cabinet, then I cam home and made a schematic of the full machine. I mainly did this for retrofitting purposes, but also to see if there were any problems. I found a few little issues, but nothing serious, I fixed them and moved on.

My guess is that the original owner has had trouble with this before, because the axis lag threshold was set at 0.2540 where the manual says it is set at 0.0047 from factory, so they have obviously encountered this issue before and tried some short-cut methods to work around it, I changed the threshold back to the factory setting and have not noticed anything worse, it appears to be exactly the same. The tool carousel was full of large drilling and thread milling tools, so I can only assume that they solved the problem, because if I'm having a z-lag error in a boring cycle or helical entry, then there is no way they'd be able to thread mill.

I am going to look at the power supply more closely this weekend, I have a few jobs on at home this weekend, but once they're all knocked over i'll head into the shop and see what I can find, I have a lecky coming to the shop on Sunday to sort out another thing, but i'll get him to have a look at a few other things on the machine while he's there. I'll post what I find.

I am feeling much happier about the whole thing today, it's really had me down in the dumps, I've been feeling like throwing in the towel, but I feel more positive now, I think if I can keep positive when I'm dealing with it, it will help me out enormously.
 
Thanks gregormarwik,

I was just saying before that I might replace the controller PSU regardless. The issue will be finding one with a 110/220 switch as the controller takes 110 in this machine (very unusual for Australia). the power supply looks fairly run of the mill, I'll check it out and see, I should be able to replace it no problems. I know some power supplies have a 110/220 switch.

I did already clean all the contact with contact cleaner, but I didn't put electrical grease on them as it was at home and I was in the shop, but I'm thinking to bring the controller home this weekend so I've got a nice clean place to putt it all apart, clean and grease every contact. I've got a little contraption that i use which is great for loading up power supplies, so I'll load it up and see what happens, but regardless, if I can find a power supply to replace it, I will just put a new one in regardless. power supplies are cheap in the grand scheme of things, if I have to I will do a recap, I've got a crap ton of caps here leftover from amp builds etc, so that can always be the fallback.

I am sooo hoping i test the power supply and it breaks down under load, it is nearly 20 years old after all, and it looks original, so i'm feeling excited about testing it, hoping, paying it's the problem.

having said that, I am also leaning towards the broken wire theory too, lets hope its power supply, and not a broken or chaffed wire, but the broken wire thing does make a lot of sense I just need to find out which circuit is turning off or shorting to narrow down the search. I might set a work offset away from the vice and run the program at a different work offset in X but keeping Y and Z the same and see if I can replicate it without cutting any material. then maybe so the same thing and only alter Y.

Thank you so much for the encouragement. I have really needed it this week.
 
Crossfire,
Sounds like you have done a lot of looking for the problem. And that's good.

Get Scientific Method with it. Testing Testing
You mentioned air cuts above the vise are OK, and material cutting lower in the vise NOT. That's changing two things, material cutting and z height.

This is where you start (In the absence of having debug tools or looking at diagnostic bits.)

Get to a place where stuff works. Then change one thing, and only one thing. If you break it, step back to where stuff worked and try one other change. Run it with no material, then again with material. Only one change at a time. If you can isolate the problem down to a single variable you will save a lot of guessing. Once you get a recipe that works, try something new like move your vise to a corner of the table and put it 6" taller test. Write a program that uses full travel of the table and z, no spindle running, and just get stuff moving.

You started by thinking it was a software problem, but then air cutting would not work in that case. You are looking at PS and voltages, but again if air cutting OK then how does that jive? Could be when spindle is under load its drawing down supply voltage, idk and not worth guessing at but the fact that you say there is a recipe (air cutting) where everything works, and another where it doesn't means you need to keep going back to what works and do testing.
 
To confirm mine I ran a long jumper wire to test run it before I ordered a new cable. I was happy that was all it was, but it was a chore to replace as it was 47ft long and ran through the whole machine back to the cabinet. I have since replaced two others in the same chain and need to do one more. Not unexpected on a 1992 machine.

If you or your buddy have wire to run a jumper it may be a quick test.
 
Thanks alphonso

I won't bother with replacing parts of the Anilam, they're too hard to find and too expensive when you do find them. It's cheaper to replace the whole controller.

I'm hoping that I don't have to, but I am prepared for the worst. I have been talking to a couple of different companies, and there's good ones and cheap ones, the cheap one I can get away for around $2k AUD, but it won't do a lot of stuff, like rigid tapping for example. Then you have the expensive (or actually mid range) option which is gonna cost me around $12-15k AUD, but it's a full professional system that will do more than what I need to do, If I had the spare cash now, I would do that anyhow, regardless of the outcome, but I don't have the money, and if I did, I need a CNC lathe, so my next priority, once this machine is running and not giving me headaches is to find a CNC Lathe. I nearly bought one last year, but the guy decided not to sell it in the end, but he might still sell it at the end of this financial year, fingers crossed he does because it would be the perfect addition to my shop, and I wouldn't need any more machines for years to come.

I looked around previously and found some Anilam 5000M gear for sale, but it was over half the cost of the complete centroid retrofit, that would be dumb. I don't know what it is about Anilam products being so expensive, I emailed a guy a while back about the manual panel keypad, he wanted $495 USD and then said i had to buy it off his mate in Australia, who then wanted to add another $150 USD to the price, that came to almost $900 AUD for a keypad the size of a credit card, SO I reverse engineered it, got a couple of simple PCB's built with tactile switches as a temporary option (5 complete keypads for $10 including postage) then I ordered a proper professionally built and printed membrane keypad for $63 AUD including postage. I haven't even installed the membrane keypad yet, as the temp PCB one I made works perfectly, until it plays up, I'll leave it there. Same thing with the function keypad, I knew when i looked at the machine that the F10 key didn't work (or you had to push it so hard that the panel behind it was bending, and you could see that's what the previous owner had been doing, so the first thing I did was make a keypad to replace it. works perfectly.

Same thing with the internal screen, it was dead when I bought the machine, they had an external monitor hooked up, it was giving me anxiety looking at this monstrosity sitting on top of the control panel, I found a heap of Anilam screens for sale all around the $1200-1500 USD mark. I have replaced panels before in other machines, so I knew that the panel would be generic, SO i yanked the panel, got a part number, got online, ordered a new panel for $62 AUD plus $11 shipping. Then when i plugged in the new screen and reassembled it I damaged the screen controller, completely my fault, contacted an Australian company and got a new one all setup and sent to me for $200 AUD, so now I have the screen sorted for under $300 AUD, a long way from the $2k+ that it would have cost to buy one in. And over half of that costs was because of me being in a hurry and not taking enough care.
 
Good job getting the screen working again for 300$ versus 2k. Sounds like you traded your time for money out of pocket. Sometimes it’s a no-brainer, sometimes not.

Question if I may. If every hr you spent working through the screen issue cost 100$, how much would the screen have cost?
 
Hi vmipacman,

Sorry for the late reply on this one, I've been busy doing other stuff and I've spoken to my fair share of experts and It's pretty much come down to a problem in the computer.

Since my last post I have done the following:

1. Pulled apart the computer again, checked every wire and every connection, I found one broken wire, but It's entirely possible (almost certain) I cut it while cutting a zip tie to release it, I soldered the wire back up and kept going. I checked, and double checked everything.

2. checked out the power supply, and all voltages according to the manual, all voltages are good. I considered recapping the power supply, but it appeared to have been recapped not too long ago, one cap had a date of 2018, so I just did an inspection and everything seemed fine.

3. after reassembly, the main keypad is no longer working, I checked everything and cannot find a reason for this, an external keypad works fine.

4. I set up 5 different work offsets and ran the same code 5 times and got the same result. in single-step it would work through the code, in Auto it would just shut the machine down and say it lost connection to the node or something like that. didn't matter which work offset I used, it just kept doing it.

5. I made a cable to connect my laptop to the drives and tried tuning the drives, but kept getting errors, only due to one of the settings in the drive dive for the drive enable signal having to come from the controller every time. I just saved the settings and left it at that while I emailed some experts about it.

6. I emailed a guy who was able to really help me out. He explained that the lag error and a few of the other alarm codes I was getting were encoder issues, but he knew the drives inside and out and said that it's usually never an encoder issue, usually a configuration issue. He had me check a couple of parameters and I found that all 4 drives were configured incorrectly for the motors, but he said not to change them until I do some more trouble shooting, as he's seen this before where the CNC controller requires them to be set that way. The motors have incremental encoders and the drives were set for absolute encoders, He's getting back to me about this, but he thinks that it's not causing the problem as the drives are doing everything they're supposed to.

6. I had a field service tech come have a look, he's had 20+ years of experience with CNC's and has seen a couple of Anilam controllers, though not many, and he ran through a few things, taught me a lot about stuff, and after he had finished he pretty much said ditch the controller. He reckons that he's seen too many people spend way too much time and money chasing a problem that can be solved by throwing it in the bin. He was convinced it's hardware in the controller unit, and he reckons it's not worth chasing up or trying to fix, as the cost would be more than a Centroid unit. I told him I was thinking of using a Masso unit in the short term and he reckons they're no good. well he didn't say no good, he just said that for what I want, They're not really suitable, they're more aimed at the hobby level guys.

7. I went to send an email to the guy that someone earlier suggested, the Anilam expert (I won't mention his name on here), and while typing his email address into my address bar I realized the address was already present, turns out he's the guy that tried to rip me off with the manual keypad. So I decided not to contact him, I'd rather throw the controller away than to deal with a rip-off merchant. My dignity is worth more than the cost of a new controller.

So where am I at now... well I am waiting for 2 people to get back to me, the machine is not running in any sufficient manner, and I am no Further ahead than I was before.

1. All drives are working correctly and there are no faults in the drives or the wiring for them.
2. I now have no keyboard working on the console, but an external keyboard works fine
3. I'm still getting the same response, works in manual mode and can do facing and contouring but no helical interpolation
4. the power supply is working correctly, I'm getting all the right voltages in all the right places.
5. So I'm ready to make a the call on Monday when these other people get back to me, it's very likely that I am going to replace the controller. I cant afford to go with a Centroid unit right now, especially since a really good lathe has come up for sale that I really want in my shop, SO I am going with a Masso for the short term. It wont do 100% of what I need it to, but it will get the machine running for under $2K. Then I will come up with a plan to do a full Centroid conversion down the track once I get another mill.
 
Good job getting the screen working again for 300$ versus 2k. Sounds like you traded your time for money out of pocket. Sometimes it’s a no-brainer, sometimes not.

Question if I may. If every hr you spent working through the screen issue cost 100$, how much would the screen have cost?

Realistically not much. I maybe spent 2 hours total, spread out over a couple of weeks. the worst part of the whole deal was a waiting for parts.

Pull the old screen out, maybe 10 minutes. half an hour if you really take your time.

Find the part number, do a few internet searches, find the best part at the right price, and compare lead times, finally decide and order one. Maybe half an hour.

Receive the new part, check it out and install it, maybe 15 minutes. I actually too a bi longer because I replaced all the foam around the glass screen protector too, that actually took a while, I had the foam tape already, but the time to clean the glass, cut the tape to size and fit it, probably took half an hour.

finding out you rushed putting the screen back in the enclosure and didn't route the cable properly, damaging the board, took about a minute, as I pretty much had a gut feeling.

emailing and finding the replacement driver board, maybe 5 minutes a day over 3-4 days.

receiving and installing the new driver board, reassemble and setup, maybe half an hour.

The thing is, I could have done it all a lot quicker had I not damaged the board on reassembly, and that was 100% my fault for being in a hurry trying to meet a self imposed deadline. If someone came to me with the same machine and same issue, I could easily get the whole job done in half an hour Of physical labour, and half an hour of administration, ordering parts etc. so if I charged $100/hr, I could make profit doing the whole job for $200.
 
Good job getting the screen working again for 300$ versus 2k. Sounds like you traded your time for money out of pocket. Sometimes it’s a no-brainer, sometimes not.

Question if I may. If every hr you spent working through the screen issue cost 100$, how much would the screen have cost?

The other thing to note about the time spent on this: My shop is on the outskirts of the industrial estate and backs onto a residential estate, so I can't make any industrial noise before 6am or after 7pm, so this gives me quite a few hours in the evening to do stuff that doesn't require a machine or compressor running. and most of this stuff I tend to do at home over a beer. With the screen, I literally just pulled it out of the machine and bought it home, I then pulled the enclosure apart at home over a beer. To me this is where you save the money, everything I can do towards my business that doesn't cost me down time in the shop is a bonus. If I'm in the shop doing this stuff I'm losing money, If I'm sitting at home with a beer in my hand, I'm not losing money (except for the cost of the beer)...

I've just signed a 1 year extension on my lease, after that I plan on moving shop to a larger shop, and one where I can work around the clock. it's ridiculous paying full rent and only being able to run half the day... And If I buy this lathe I'm looking at, then I'm gonna have no room to do anything in the shop, so I'm gonna have to move next year no matter what.
 
Hi Everyone, I haven't been back in a while, to be honest I completely forgot about this thread, so I thought I'd share an update.

I tracked down someone that was very familiar with Anilam controllers and he gave me a quote to fly up and take a look but it was going to cost me over $5k for him just to take a look, and he was in lockdown anyway so he wouldn't be able to fly up for what turned out to me months. He gave me no guarantees, but said he knew what the problem was and was confident he could fix it. In the end though, I bit the bullet and bought a Centroid Oak controller and did a self-install. The machine has been running very well with the new system, and almost paid for the retrofit on the first project alone. The machine has been running nearly all day, every day, and I've done some very long days on it running from 4:30 am till midnight without a hiccup (or a noise complaint).

There are no problems with any of the drives, and all periphery and auxiliary components are functioning correctly, further proving that the problem was in the Anilam controller itself.

The entire retrofit cost me a fraction over $10k AUD, with only one unexpected cost which was paying someone to write a custom PLC program for me because I didn't have time to learn to do it myself.

I would like to thank everyone who contributed on this thread. I really do appreciate all of your input.

Thanks

Jason
 








 
Back
Top