Wow three posts all in one thread. With attitude Realy?
Can you spell T...R...O...L...L?
My momma taught me if I cannot say something nice about some one not to say anything so I am going to be quite now.
Good evening everyone
Ray, any clue on a potential price point yet for the new ATC?
Nate
Fine Line Automation
Home - Fine Line Automation
Fine Line Automation
www.finelineautomation.com
Boy, was this one of those days.... Went out to make the rest of the new coupler parts. Halfway through the first one, the Y axis faulted, with an encoder error. I removed and re-seated the connectors at both ends of the encoder cable, but no change. So, I swapped all the cables between the X and Y servo drives, to isolate whether the problem was in the motor/encoder or drive. But, after the swap, it's worked perfectly for hours, so I have no clue what happened this AM. Weird...
Got the rest of the couplers done, and installed, and they work really nicely. Only problem is after just a few minutes of testing the little servo motor died - one of the windings opened up. Sadly, that is the ONLY motor for which I don't have a spare on-hand, so I may be down for few days until I can get another one in, unless I can manage to repair it. But, the couplers work really nicely - they center the tool to the coupler motor, which is exactly what I wanted - not only providing accurate, repeatable location, but also locking it in place
Two steps forward, one back....
Regards,
Ray L.
Hi Ray,
I just wanted to let you know that I'm sure I'm not alone that I look forward to your posts in this thread and the lack of responses occasionally certainly doesn't mean we aren't reading and enjoying!
Eric
Eric,
Thanks! I figure most people just read and go back to more important things! :-)
Unfortunately, my son gave me a cold, so I was moving kinda slow today. I did get the new transfer arm fork made, and it came out nice. Just need to find the right spring for it.
I hope to have a new motor for the lock in the next couple of days as well.
Regards,
Ray L.
Ray....I'm right there with Air Gunner. I really look forward to updates in this thread!!
Are you familiar with Coil Springs Manufacturer - Stock & Custom - MRO & OEM | Century Spring Corp. centuryspring.com ? Tons of great info and options. My local hardware stores carry some of their stuff, so I've been able to find what I want online, then pick them up locally.
Get well soon,
Mike
Don't have to be too bright to be me
Now that the forum finally seems to be back almost to normal....
The last week has not been as productive as I'd hoped. I was sick for several days, then had out-of-town company for several more. I do now have the new transfer arm fork made, and today I re-made the lower end of the transfer arm drive, which includes a radial support bearing for the transfer arm shaft, re-located gear-motor to provide proper tension on the belt drive, and the encoder has been moved from the motor down to the belt drive, inside the (sealed) belt-drive housing. Backlash on the transfer arm is now very greatly reduced. In fact, it is considerably less than I expected.
Unfortunately, all the little changes end up rippling through other parts, so there are a few more parts I have to re-make, including the transfer arm shaft. I also have some changes to make to the lift drive, to incorporate the encoder into the thrust bearing housing.
I got the new 10X+ faster MCU boards last week, and it only took a few hours to port the firmware over to the new processor. Unfortunately, I now have to rip up and re-do part of the electronics, as the new MCU is 3.3V, while the old one was 5V. So, I have to remove all the level-shifters I had to add so the 5V MCU could talk to the 3.3V FPGA! But, the new processor will give me a lot more capability, so well worth the bother.
Here are a few quick pictures:
New Transfer Arm Fork:
Attachment 240970
Transfer Arm Encoder tucked away inside the belt drive housing:
Attachment 240972
New Transfer Arm Drive "Bottom End":
Attachment 240974
Also had a small problem with my machine today. I think I mentioned that about a week ago my Y axis servo threw an encoder fault. I swapped the X and Y axis drives, to narrow down whether the problem is really the encoder or the drive, and it's been working fine since. But today, the Y axis again threw an encoder fault. Since it's now happened on two different servo drives, clearly the encoder has a problem. Luckily, I have a Pulsar as well, so I stole the X axis motor and encoder from the Pulsar, and swapped it into the Y axis of the Pro, and had no more problems the rest of the day. Swapping motors only took about 20 minutes, so not a big deal. I'm not sure if it's possible/practical to replace just the encoder, as it's integrated into the motor. So, I may have to get a complete replacement motor.
Regards,
Ray L.
I have taken apart several servo motors in the past. While some use an custom PCB based encoder. The last couple I have opened all had an off the shelf optical encoder mounted under the rear cover. It should be nothing for you to do. Start by carefully taking the back cover off and see what is inside.
.....so can you take apart the bad one when it gets replaced so we can see what's in it? I know you're busy, but I'm sure we're all curious.
I look at those beautiful parts you're making and get excited about what I can make with my Pro! Thanks!!
Don't have to be too bright to be me
This has been a couple of weeks filled with the "thrill of victory", as well as the "agony of defeat". I'm now running on the MUCH faster MCU, which has enabled me to greatly speed up the servo loop. I"m now running at 2 mSec (it was 10 mSec), and I'm confident I can easily run at 1 mSec. I've been able to simplify and speed up the servo code, and also made changes which further improve the positioning, and better deal with the "stiction" and backlash. The carousel is now working very nicely, and reliably, and it can be run as fast as I see any point in running it. The tool "lock" is also working nicely, and does a great job of pulling the tool receiver to exact final position, and holding it still as the tool is inserted and removed. I re-made several parts of the transfer arm drive, mounted the new high-resolution encoder on the output shaft, and got proper tension on the belt drive. I also made a second pass through the FPGA and made some minor improvement there, and the electronics as a whole have been working perfectly for some time now. I haven't even blown up a single motor H-bridge!
I was able to go through the entire toolchange sequence quite a few times over the last few days. though I didn't have a chance to capture a video. The *only* potential problem I see getting close enough alignment between the tool in the transfer arm and the tool receiver and spindle. Unlike "real" ATC toolholders, like ISO, CAT and BT, TTS is very unforgiving on alignment. the tool MUST be pretty well centered before it can be inserted. A little off-center, and it WILL cause jam. The tolerance is only about +/-0.05", which is not much. I REALLY wish I could do this with CAT30 instead of TTS....
The one fly in the ointment is the motor and gearbox I've been using for the transfer arm are simply not up to the task. There is too much backlash, and nothing I can do in software mitigates it adequately. This makes getting adequate alignment of the tool to the spindle extremely difficult. Today I swapped out the gearbox for the same one I use in the PDB, which only required a few very minor modifications to a couple of parts, and a simple mounting adaptor. This one has lower backlash, and also has the advantage of a very wide range of selectable gear ratios, and a very wide range of motor choices. However, even with this, I think I will also have to give up on my dream of being able to position the transfer arm based solely on encoder readings, and will instead have to provide a "hard stop" to set the final alignment to the tool receiver and the spindle. This is quite trivial on the carousel, a bit trickier on the spindle, but it still do-able. Unfortunately, I ruined my last spare timing pulley, trying to bore it out for the new gearbox, so I'm dead in the water for a few days until I can get a replacement in.
All in all, I'm pretty far behind where I'd hoped to be by now, but I see few potential obstacles beyond this point. I have a good feel for what remains to be done, and none of it appears terribly daunting, the worst of it being the problems with the transfer arm. So far, everything - hardware, electronics, software - is proving to be quite reliable, and behaving at least consistently. That at least gives me confidence there should be no major surprises down the road.
BTW - I don't know how many of you guys are signed up as "Novakon Demo Centers", but I signed up a while back, and I've been surprised at the number of contacts, and visitors, I've had. I've had, so far, four people come visit over the last few months. Most just want to look at the machines, and talk about the "Novakon experience", a few had specific operations they wanted to see demonstrated, ,so I setup a demo for them. Average visit has been on the order of 2 hours. Its actually been kinda fun, and the people I've met have all been very nice. I'm confident at least a couple of them will be buying machines. Seems like most of the interest is in the Pro, though one guy was looking at the Pulsar, mostly because he didn't have room for the Pro. Seems to me like a nice way to meet other potential Novakon owners, or just CNC geeks, in your area.
Regards,
Ray L.
I did actually open it up, but didn't take any pictures - really not much to see. The cover comes off with four screws, and inside is - SURPRISE! - an optical encoder. Coolant had gotten inside, and the encoder was wet with condensation. I dried everything off and left it open for a day to dry out. I then put it on the Pulsar, and it seems to be working fine again. Oddly, the connectors are all IP67 rated (i.e. - waterproof), yet the housing itself was not even gasketed. At some point, I intend to take all of my covers off, and either add gaskets, or sealant, to ensure this does not happen again.
Regards,
Ray L.
Ray;
With the importance of massive coolant flow, I would think Novakon would step up and offer a set of gaskets for the servos. Unless protected, the coolant will go everywhere and most often, where it shouldn't be!!!
Thoughts?
Bill
billyjack
Helicopter def. = Bunch of spare parts flying in close formation! USAF 1974 ;>)
Well, I was able to graft an encoder onto the motor of the new gearbox/motor assembly, and do some testing. Backlash is well within the acceptable range, and with a 100:1 gear ratio, it's more than fast enough, with more than enough torque. It is a bit noisier than I'd like, but there are ways that can be mitigated Positioning accuracy is very good, despite the relatively high friction that comes with the high gear ratio. Fortunately, it seems to be "dry" friction, rather than stiction, which is FAR more difficult to deal with. On the plus side,the friction acts as a built-in damper, which appears to make the servo tuning relatively non-critical. My one concern at this point is how much backlash is added by the belt drive. But, if need be, I can, with some difficulty, eliminate the belt drive. I also know from experience on the PDBs, that these gearboxes are VERY rugged, and do not show any measurable signs of wear, even after long, hard service.
Hopefully I'll have the new pulleys quickly, and be able to test the whole assembly, with belt drive.
Regards,
Ray L.