[RustedOut]

They got no lines, they got no wheel and they say they got no led. They got a variable ppr and they fit multiple shaft sizes and have multiple mounting options. For less money than the HEDS with the EM1 encoder module, actually less money than the EM1's themselves. I was just going to get the EM1's, but decided to save some money with the others. Here's a link to them, AMT102-V.

And they fit to the old HEDS cable, run from 5v and have the same 1000 count ppr I have been running (and many more) ppr's. They should fit into the original IH housings and I saved about $10.00 apiece. I'll let you know how they work out.

If the flats I ground on my shafts is a problem I may be backing up and getting the EM1's to modify my HEDS. I'll know that right away.

Bob

[Cruiser]

From what I saw on the video, I think the base adapter would just average out any deformation from the flats and otherwise not care. I be a bit curious about possible contaminations, the bug spit celluloids are just too finikitastrofic (word designed especially for this application) ! If they are not that finiki then they might just be the ticket.

[RustedOut]

Ok, now I'm learning a another thing or two about encoders. The new AMT102-V's showed up yesterday and the shipping was fairly quick. Had it in my head I'd just put them in place where I removed the old HEDS encoders. That didn't work out as the mounting screws for the HEDS are a size larger than the space and holes in the AMT102's. The AMT's really need 4-40's to fit the holes and clear the encoder itself.

So, I had a choice to make. Mount them on the motor or drill and tap for them on the plate where IH had them. There are already holes to use on the motors, so I was leaning that direction. The deal was sealed when I figured out there was no bushing for a 3/8" shaft. I thought I read that somewhere, but I'm wrong again. My motors all have 1/4" on the back end and there's a 1/4" bushing. But I didn't have the little 4-40 screws needed to mount the bases.

Picked up the screws at the local Ace this morning and set to work to get the Z encoder working. The old cable does plug into the new encoder, but you can't snap the encoder to the base using it. The connector is too thick. So now I'm needing cables specific to the AMT's. Looking like next week is as soon as this thing will be going again. Now the EM1's are looking like a better replacement if you already got the HEDS, or at least an easier swap.

These are nicely made though with cast metal housings. The encoder snaps to the base and locks the cable in place. Installing them was easier than the old HEDS, imo. I know I'll have to redo the encoder info in Mach, but I got a few days now to get to that.

Bob

[Cruiser]

I don't think I caught where you ended up mounting them, it sounds like you modified the housing so they mount to end of ball screw, which I think is the BEST place for them.
How do these units look for handling some of the possible contaminants that exist in the milling environment ?
oh and do they function on a reluctance Rather than optical influence ?

[RustedOut]

I got the new Z encoder on the end of my Keiling motor now. Wouldn't fit in place of the IH original due to the 3/8" ballscrew shaft and the encoder mounting screw size. If the machine was apart, the ball screw would be an easy fix in the lathe and the plate re-drilling would be simple enough too (current holes are over sized). The flats I was worried about would have been a problem if the shaft could have stayed at 3/8", but if I do take them down to 1/4", it should work out.

I agree with you about that being the best location for them, but for now I got to get the gremlins out of the machine and work the jobs in the shop. I had read that the old cable would plug in and they did, but I couldn't snap the encoder pieces together like they're designed to. The new cables may be here tomorrow and if so I'll get the Z encoder working this weekend.

Whatever is providing the pulses is buried inside the encoder housing, so you can't look at it too closely. There is quite the printed circuit board there though. I had to get the magnifying glass out to see how to set the jumpers for ppr's. That is some dinky printing on it. All in all, they look pretty robust. I guess the concentricity of the rotating disk to the encoder body is critical, so you get a tool to set them relative to the shaft. According to the data on them, they work by generating output code through detecting changes in capacitance. The disk just has a copper ring mounted on it, so other than the exposed electronics to coolant, they should be fairly dependable. Dust and oil film shouldn't be a problem. That's two things I've fought in the past.

I plan to do some testing with the drive belt removed. I'm so tired of the head coming down on the table, so I'll block it up out of the way. No damage on anything to this point other than broken cutters and a couple scars on the old vise, but no point in taking un-needed chances either. I've got a large thick piece of mdf to cover up the table under the head too.

Not sure if I mentioned I got the news my spare drive was scrap. It had let go of the motor control around or just after Christmas time and faulted when the head ran into the vise. Head just fell from gravity about 6" above the vise after the power to the motor quit. That random stuff is hard for me to troubleshoot. That was in between running the same program over and over on multiple parts. No cutter change i the program, just running a 1/2" radiused end mill. Had to stop and order a new cutter after that, and change the drive out of course.

I've been using the mill down time to catch up on a bunch of lathe work. I at least got one machine I can count on and it's older than me. I enjoy running that old thing and it still holds nice tolerances.

Bob

[RustedOut]

Not much progress to report, but I have now smoked my second Gecko. Any way you try to spin it, that's not progress. I just seem to be spiraling into a black hole with the Z.

I've also decided I don't want to run the AMT encoders, preferring to stick with the US Digitals. I'm changing out the HEDS modules for the EM1 modules. The AMT's are probably fine, but I like the encoder on the screw and going this route is the easiest way to accomplish that.

My new 320X drive arrived today, but I've been working on another project so it's just sitting on the bench. Besides now I'm a bit gun shy of anything electrical. Not going to try running the mill again until the EM1 is installed. It's not due to be here till mid week.

Bob

[RustedOut]

Got the new EM1 encoder module mounted and the new 320X wired in yesterday morning. But I've been very reluctant to turn this thing on for quite a while now. I had the drive belt off and the power cables to the motor disconnected, so I did finally fire it up. The drive responded exactly like it should in both static and in hand turning the screw or commanding a move. Then wired the motor in and it took off at full speed when I applied power to the cabinet. Encouraging, but I was still nervous. But I put the belt back on and tensioned it.

With a tall timber under the head, I knew it wouldn't travel far. So I hit the switch and reset the drives and it just sat there. Reset Mach and it still just sat there. Jogged Z a bit and it moved down, then up, then down, then up, responding to my key strokes. Flipped the timber over to gain more travel room and commanded a rapid down and it went and stopped on command. And did it stop, not hunting for position, just stopped. And it did it smoothly and quietly.

Every command it followed and stopped on command. The motor now is pretty quiet, strong and responsive. And when the move is done, it just stops. It hasn't acted like that in a long time and it's good to finally see that again. This is all off of the default drive settings and no tuning. Pots all set at 11:00 positions like it says in the manual. I ran a test program for about 30 minutes then commanded the mill to verify. Numbers came back at .0001" Jogged the milled around for a while and ran the program again and verified it once more at .0001" I'm pretty encouraged with what I saw yesterday and today I have time all day to mess around with it. Heck, I may even try to cut a part.

I think a big part of the past problems has been the drive and I've thought often about just replacing it. But there wasn't a real compelling case for doing that, so I kept chasing my tail with it thinking I could tune it out of it's problem. Going with the new module is a low cost move to find stability and the death of the old drive may have been a blessing in disguise as it forced my hand there. Hopefully this is the end of this saga and I can move ahead with the planned belt drive conversion.

Bob

[RustedOut]

Two days after that last post, I was back to the same old thing. Couldn't get the motor to run quiet and running a test program the drive would fault on the second or third rep. Broke down and called Gecko for some advise and was told to look at the motor. Now I sure don't know how to test those things, but got some direction from them.

Did as suggested and the motor fell within the target amps. Went so far as to look at the bearings and measure the torque required to lift the head. It looked pretty good there too. I even installed my multimeter in the power line to the motor and took some current readings under load. Worst I measured was 3 amps at a change of direction on rapid feed and I certainly didn't think that was of any concern. But again, I'm electronics stupid, so I was stumped about that point again. It was becoming a familiar feeling, not a good feeling, but familiar.

Next morning I was back at troubleshooting it and now for the first time I was getting some feedback from the motor that all was not well. Finally a symptom I could definitely say was motor related. And it went downhill pretty fast after that. The motor's not totally dead, but it can't handle the load now.

I hate intermittent problems, especially when they're electronic in nature. New motor is on order and will be here middle of next week. I also learned how to test a drive to see if it's bad, so that was good too. Gecko's servo info says they are good for around 2000 hours and that's about what this motor had on it. At 226 oz in of constant torque, this motor should have had the torque to lift the head with a direct connection to the ball screw. Running a 4:1 reduction through the timing pulleys, I'd think it should have had a relatively easy life. I intend to take it apart and check it out to try to figure out what failed. If this has been the problem all along, it makes me wonder about the other two motors.

Bob

[Cruiser]

I had simillar issues with my Y axis motor. It had a very hard stop hit and stall and consequently burned out a winding. When I took the motor apart it could be smelled as burned. A new motor got things going again. The problem occured when I was trying to set-up my soft limits and botched a setting. These motors are very powerfull but also a bit delicate. Part of the issue is that I run the jumper on the gecko's fault terminal.

[RustedOut]

You know, at the first signs of trouble one of my thoughts was to just replace the drive and the motor and be done with it. But I figured I needed to learn more about the black art of electronics. Didn't think going in it would take this long. The mill hasn't been able to run for around 8 to 10 weeks for more than a few minutes. It always acted the same with the drive faulting or me being unable to tune the motor...again! Sometimes tuning multiple times and trying to test in between them.

I became so fixated on the drive being the problem it was only this week that I started to think about the other pieces involved, like the motor and bearings. I was headed for checking the bearings when on advise from Gecko I started looking at the motor. Was pure coincidence that it started acting up while I was focused on it. After taking the current measurements the day before, I wasn't sure it was the problem and was scratching my head again.

I've taken notes during this period and am adding them to the other documents I have on the mill. Hopefully this solves the problem and it can finally go back into service. File those notes away, chock this up to experience and move on. It ate up my belt drive money and put it in the future a bit more than it was. But I did get some electronics knowledge out of it. I think I'll use a bit of that and measure the new motor as a bench mark for future reference.

Bob

[RustedOut]

So now I have a new Gecko 320X and a new Keiling servo motor on the Z axis. Had to fiddle a bit with the drive to get it to reset as I had been adjusting it to try to get the old motor over it's issues. But let me back up a bit and provide some more history.

The last thing, after installing the new Gecko, was the drive would fault while running a test program that merely moved all three axis a bit. It has been my mill warmup and oiling program for the past couple of years and I had the feed rate at 80 ipm up until just recently. Once I installed the new gecko, I could run the head up and down with the jog keys at any feed rate I wanted. But if I tried to run it with that program, it would fault right after the first down move. I finally got the feed rate down to 25 ipm and could run all the repetitions I had set in the program. But bump it up faster than that and the drive would fault quickly.

It was after one of those many faults that I was trying to jog the head up and park it when the motor sounded like it was dying. It could hardly turn and displayed the same characteristics jogging down. Nothing I could do to the drive changed it. I never did get the head lifted and just blocked it where it was. The drive never faulted during this last bit of craziness. I cut the power to the drives and went and did something else for a while. Came back and hit the power switch and the reset button. The drive would not reset. If i held the switch very long it would start to vibrate. That's when I got on line and ordered a new servo motor.

Today, I got the motor installed and finally got the new drive to reset. And I could jog it up and down at any feed rate I wanted, just like with the old motor. It sounded good and strong and it was now quiet again. Hurray I thought. The way I was loading the warmup program was by load recent and that's what I did this morning. It was still set at 25 ipm and it ran fine. I bumped the feed rate up to 50 ipm and it faulted right after the first down move just like before. I thought what the heck is going on. Back to the jog keys and it ran strong. A buddy calls and wants to go to lunch, so I was out of there.

Got to thinking on the way about the program and what if anything could be causing that problem. Got back and read through the program and found nothing. Ran it again and got the same results. Funny thing though, the motor kept trying to lift the head even after the drive had faulted. The pulley wouldn't turn far, but I could see it turning. I had a hunch then that the problem was in the program, or what Mach thought was in the program. I went to the folder it resides in and reloaded it. Same values as the one I just dumped and gave it a try. Mill ran perfect. I ran it a few more times to make sure and it repeated each and every time. Then I changed the feed rate from the 50 ipm it was at to 80 ipm and it ran perfect. I ran it all afternoon and it has been perfect every time.

I tore the old motor apart and inspected it and found nothing. Checked for shorts, nothing. Brushes look nearly new and the commutator looks good. Before I installed the new motor I measured the current feeding it 5 volts with no load and it was very close to the old motor. I'm talking an amperage difference between the two motors of .003 amps. That old motor is still good in my electronic idiot opinion.

Now I have no idea what was actually causing this last bit of strangeness, but it definitely was associated with the program. The one Mach was reading as a recent file. But not a problem in the program, just how Mach was seeing it I think. And I have no idea if that was what caused the motor to act like it was on it's death bed. Probably not as that was all off of the jog keys, no program involved at all. All I know for sure is the faulting problem went away immediately after loading from the folder. What sucks is I have chased this for the past couple of months and I suspect that if I had done this back then, the mill would have been running a long time ago. But why would I think the program was the cause after running it hundreds of times in the past with no problems. It certainly doesn't explain some other problems I went through though. Some I have no clue what the cause was, only know that today it was running fine again. But I've been to this cross roads before too.

Hopefully this is the end of this strange problem and I can get the mill back into service once I can establish some trust in it again.

Bob

[RustedOut]

More testing today and more failures. I knew it was too good to be true. One thing is good though and that is that the head doesn't drop on faulting of the Z drive. That allows more aggressive testing. I stated yesterday "Now I have no idea what was actually causing this last bit of strangeness, but it definitely was associated with the program" and I hold by that today. But what is different in running the program (or any of my other programs) is I'm asking the mill to move in all three directions all at once. That's where the problem starts.

Running a test program to just test Z shows no problem at any feed rate for as long as I care to run it. But bring the other two axis into play and the drive faults on the first down/up move. It has to be faster than 25 ipm, but will fault at 35.

I ran a peck drilling program yesterday and it ran fine, but bear in mind that X and Y were satisfied before Z was commanded to move. I ran it a number of times without fail.

So, pondering this for some time today, and knowing that the head would have dropped quickly from gravity with faulting of the old 320 drive, I'm thinking this problem has been around a long time and may have been the cause of the first drive failures last summer when the head dropped on a commanded Y only move.

Sorry if I'm a slow thinker, but having the head drop just on testing makes a guy a bit reluctant to do much with it. And I need the repetitions in order to see a pattern. One drop could cause much damage and it has been getting costly. But the new 320X was a good purchase if for just that one feature.

So, that's the problem and I'm thinking it's not due to drive tuning adjustments of the drive itself, but maybe something I've set in Mach. Funny thing is those settings haven't changed since I've gotten this thing running a number of years ago. Fairly certain the drive faults due to missed steps, nothing more. There are no hard stops as it's just running in the air.

The hunt continues, so here's a question for you Mach 3 experts. I have the X and Y settings the same in the Motor Tuning window at 126 Vel and 9 Accel, with Z being 84 Vel and 8 Accel. Would this cause this kind of problem and is there a compelling reason to have them all the same at the lowest common setting? The Z motor is the large 1125 oz-in Peak Keiling motor and the other two are the 850 oz-in Peak Keiling motors if that matters. All are 4:1 reduction belt driven to 5 tpi ball screws.

Bob

[fatal-exception]

Have you ruled out your computer being the problem? I saw the same type of problems when I had the pulse rate set too high and was doing a full 3d program. It would act as if it were a stepper motor machine that was losing steps, and it would happen on all the axis' randomly. This was never a problem with up to 2 axis' moving at the same time.

I also had a computer that would randomly drop the pulse stream for a split second which would cause problems. It was a laptop with speed step processor.

There's no reason to have all your axis' set to the same speed and acceleration. None of my machines have the same settings for all the axis' and I have no problems.

Paul

[Runner4404spd]

I had this problem on my computer as well when using the on board parallel port. i bought a dedicated parallel port card and the problem went away. a few weeks later the computer died(kept locking up) and i had to replace it. i still used the new parallel card in the new computer and my machine works great. i have simultaneous motion in all three axes with no lost steps.

[RustedOut]

Paul,

I guess I haven't ruled out anything at this point except I believe the Gecko 320X is doing what it's supposed to, the new servo runs smooth, quiet and strong, so I'm thinking it's not the problem and I recently measured the torque required for the screw and don't believe it's an issue either. And thinking on this, if either were a problem, wouldn't they show up in single or two axis testing especially as I approached the higher feed rates. I've now ran two axis at 120 ipm, after changing the motor tuning vel and accel values of my Z motor to match the X and Y. Didn't matter which two as long as one was Z.

I recently swapped computers as I developed a parallel port problem on the original. That was my own fault. Scrapped that motherboard a few days ago. But ruled it out, no. It's set at 45,000 Hz, but tested better than that. No changes made other than the one I made today to the motor tuning of Z. But this may be something to look into as I put this one into service just before the new Gecko. It's a P4 running somewhere around 2.35 ghz. And I suppose I could have damaged the cable as well. I should be able to do a continuity test on that though.

I measured the amperage draw on Z last week under rapid feed rate and only got a bit more than 3 amps. I got 20 amps on the power supply, so if all three take that I'm still way under to my thinking. And at 50 ipm I got just a bit over 2 amps. But I'm still wondering if this may be an issue and I may pursue more info from the drives and motors so I can evaluate it. The drive seems to be faulting at the change of directions so I'm kind of leaning toward something other than missed steps right now, but what the heck do I know.

I did play around with the Gecko this afternoon, changing the dip switches settings for the following error. But I flat refused to change the trim pots as far as tuning the gain and damp. That could have had a bearing on the outcome, but at 256 following error the motor runs the best. All revealed the same tendency to fault on change of direction though. Maybe not the best test conditions, but I wasn't seeing results that made me want to go to the effort of re-tuning the drive to the individual switch settings.

And I changed the trim pot level switch (# 7) for the amps limit. Took it off of default and the drive faulted as soon as I tried to move the Z. Put it back where it was at the default position.

Runner,

That may be something to look into. I guess you just changed the address in Mach once you added the card? I'm thinking it's just something little, something simple (well not simple for me) that's causing this problem. And I'm really believing it's been with me for some time, but I have run the oiling-warmup program a bunch of times in the past and even after the first encounters last summer. So I'd be hard pressed to say it's been there much before last fall/winter time frame. That's when the head started dropping at an alarming rate for no apparent reason.

I think I'll try to find one local this weekend and give it a try. I've also been wondering about my breakout board too and I guess I don't know enough about it to test it.

Thanks for the advise guys, you've given me some more to look at and I'll check those things out and keep trying.

Bob

[Cruiser]

Now you said something !
First you say all three axis in an interpolation movement.
This would tax the power supply driving the axis. Also, I had a computer that was of 900hz and a power house but it could not keep up with the running math needed to keep movements fluid and it would loose occasional steps without faulting out. I also had an instance where it stalled and faulted the computer internally and it seemed that the drives didn't know what to do. I built a new hot rod puter and the problem was solved till my motherboard started popping caps. I lost count on the puters I had in the shop but I can say that the environment is harsh for a puter and I have since been puting home heater filter gauze in the intake air side of puter to filter out as much as possible and it does help a lot.

[RustedOut]

Really, it used to do it. But that was with an older Gecko 320. Now I've gone from bad to worse. I can't even get the 25 ipm 3 axis travel out of it today I was getting yesterday. I checked the computer and Mach says it should be good to over 75 KHz. I'm running it at 45 KHz and Mach diagnostics says its 43,859. That's what it tells me, but I'm not totally sure I understand all that added together. But doing the math, the feed rates should max out at 135 ipm and I've got them set now at 126. I can get 122.8 while running two axis, but I'll be honest and say I'm afraid to try it with all three axis. I've watched the head crash into the vise way to many times to want to go there again. Correct me if I'm wrong, but that velocity setting shouldn't matter to the slower feed rates I run in my programs, so I'm thinking it's not part of the problem.

And I called trying to find a parallel card and just got laughed at. Looks like I'll have to mail order that. I checked the cpu and it's a P4 2.4 GHz desktop running windows XP and not much else. It's pretty stripped and it's only goal in life is to run that mill. There is some idle programs that I could remove if it would help. 1 meg of ram is all it has though and I just upgraded it from a half meg a last weekend.

I tried talking to Gecko tech yesterday, but they were all busy. I'll try again on Monday. What is really puzzling me is the intermittent state of it. Two days ago it seemed fine running the old program up to 80 ipm, then the next day it can only get to 25 ipm and today I can't even get there. Maybe I'll try the old computer tech advise and reboot the computer and try it again.

I went back and looked at a program I recalled the head just fell out of nowhere and it was doing some 3 axis interpolation in the approaches and retracts. I can't recall exactly where in the program it happened, but I'm thinking it was just after a rewind while I was changing parts on the fixture.

What kind of hotrod computer are you running Don? I run filters on the intake of the cnc electronics cabinet, but not on the pc. This one's pretty clean right now as I just put it into service.

Bob

[RustedOut]

More testing, more faulting last night. Never did get the mill to run my test program at the low 25 ipm feed rate. And it didn't seem to fault on the same place in the program. Once it faulted after going halfway up the total travel on the last rep. Then back to the first up move following the first down, which is the most common. Then one time on the third up move. But I found myself sitting there watching it fail time after time wondering if the computer is the problem, would a Smooth Stepper solve the problem. Or is it really missed steps that's causing the fault. How would I go about proving one or the other?

Changing from the default error limit value of 256 to 512 on the drive didn't help. There are two other settings, but in testing it seems I'd have to re-tune the drive to get them to work as the motor struggled to run smoothly when I chose them. At 2048 the motor was so completely out of tune I didn't try running it too long. Not too sure why that would change, but it seemed to.

Forgot to mention I have the 320X switch for the trim pot set to emulate a 320 drive and have the trim pot turned all the way to 20 amps. The power supply is giving me a tad over 78 volts and has a tag reading 20 amps max. It's the standard IH 65 volt power supply. The Z drive only used 3 amps more or less in the area the head is traveling in at a rapid feed rate and between 1 and 2 amps at 50 ipm feed rate so I don't think I'm close to the 20 amps supposedly being provided by the power supply. I'm guessing the other two drives are similar to the Z, but I suppose I could measure them to be sure. This is where I thought the problem was coming from last week, but haven't seen anything so far to support it.

I'm going to mess with the Accel value in Mach today to see if I can affect the drive faulting.

Bob

[RustedOut]

I think I'm back in business.:cheers: I lowered the Accel value for all three motors to 7 and tested first at the low 25 ipm. Success, then up to 80 ipm. Again it ran without fail. Ran it a few times there just to make sure of what I was seeing. I just finished a couple tests at 100 ipm and it ran great. Showed 97.47 ipm actual. Pretty sure it would go faster, but I have no need of that so I'll quit there. The difference in commanded and actual feed rate matches the difference in my set pulse rate and the actual of 45000 and 43859, so I'm thinking all is good with the pc, parallel cable, ps and drives. I left the following error switch set for 256 with the motor running pretty smooth and quiet. I'm thinking there is room to give on the Accel value now. 8 looks like it was border line, 7 seems to work fine, but I doubt I'd have much issue with a 6 or even a 5 setting there.

I got a bunch of lathe work to do today, but I'll continue to test the mill to be sure it doesn't fault. If I can make it through 30 or more reps of the whole program, I think it'll go back into service tomorrow. That will be the first chips in many months.

Thanks to everyone for the tips and advise. Sometimes a guy gets so focused or frustrated on one thing he doesn't look at the broad picture. That's what happened to me on the drive, but once I could get the head to stop falling it got easier to stop watching and reacting to it and check someplace else.

Thanks again guys, hopefully it's fixed.
Bob

[Cruiser]

My Hot Rod puter is just an assembly of parts from New Egg.
Now I am getting the notion that your ails were induced by your desire to make it go as fast as you could tweak the settings.
I set my accel to were I can just hear it smoothly accel to speed and go. I found that the top vellocity is seasonal and temperature limited, so I just leave it at 100 and the accel at 7 on X & Y and 5 on Z.
I started out being greedy too and quickly found out that it was foolish all the way down the line, starting from setting the Geckos to setting up the motor tuning !