[robhrzic]

txcowdog, I am using (4) Gecko G320's. Got them last fall. They are identified with an "H7" marking (August 2007?).

[cam1]

Hi:
What's the game plan for the Z axis, are you going to use a ballscrew on the knee?
We have a First CV 250 CNC mill at work with the Z axis on the Knee, and it's a really nice feature.


regards

[CuttersCov]

Quick note:

Putting CPU fans and heat sinks on your Geckos is ingenious. I'm stealing that idea.

[txcowdog]

I just bought four of the HSF from http://www.surpluscomputers.com/stor...&item=ACC10812 for only $5 ea. They are a good large size that seem perfect for the Geckos.

[jdelaney44]

Dude! Looks excellent. Very nice.

John

[robhrzic]

cam1, I plan on driving the knee for the Z-axis when I need a precise milling depth and/or lots of travel. I am also planning on driving the quill for higher speed operations like peck drilling and boring, where Z depth is not necessarily that important. This would provide less wear and tear on the (knee) Z-axis servo motor, trying to move all that mass at a rapid rate.

I do not have any plans, yet, to change to a ball screw on the knee. I do not think that a ball screw may be needed, as the weigth of the table, and gravity, should be able to eliminate any backlash in the lead screw.

I will be using a double air cylinder counterbalance to lighten the load on the Z-axis servo motor for the knee.

[cam1]

Thanks robhriz:
Using the knee for boring operations could be more desirable than the quill, as in order for the quill to move, it must be unlocked, the clearance (.0003-.0005) in this interface can sometimes be a hassle for boring.

regards

[protman16]

cam1, I plan on driving the knee for the Z-axis when I need a precise milling depth and/or lots of travel. I am also planning on driving the quill for higher speed operations like peck drilling and boring, where Z depth is not necessarily that important. This would provide less wear and tear on the (knee) Z-axis servo motor, trying to move all that mass at a rapid rate.

I do not have any plans, yet, to change to a ball screw on the knee. I do not think that a ball screw may be needed, as the weigth of the table, and gravity, should be able to eliminate any backlash in the lead screw.

I will be using a double air cylinder counterbalance to lighten the load on the Z-axis servo motor for the knee.

You might want to consider a ballscrew for the driven knee. It will help with travel speeds and allow you to use a servo with less power. I woouldn't worry about wear too much as long as you use a proper oiler.

The convesion I accomplished some time ago can drive the knee up to a software limited 135ipm. I find this plenty fast for drilling operations. Soon I will be making upgrades to this mill that will allow travel speeds of 500ipm.

[robhrzic]

Can anyone out there tell me if i could possibly use a PC Based Oscilliscope to tune a servo motor for use with a G320, such as this on that I found on eBay:

http://cgi.ebay.com/5-Channel-USB-Os...QQcmdZViewItem

Are there any (explicit) instructions anywhere as to how to go about doing this (tuning with a scope). Where do you connect the leads / probes? How do you setup the scope for the measurement?

I know that one can "tune by hand", but I would really like to use a scope to tune the servos. That way I may be able to have a bit more confidence that the servos are critically damped as they should be.

Since the servo tuning / electronics portion of this project is, by far, the most enigmatic and misunderstood for me (I'm have a mechanical backgound), I'd like to be able to confirm my tuning by hand or "guesstimate" with something that is a little more concrete (in ny eyes).

If I can't use this type of PC Based Oscilliscope, what type of scope would I need to do the job?

[bemfarmer]

I'm interested in this question too. Maybe post in electronic forum or gecko forum? I see an oscilloscope trace over at homeshopcnc, with current probe.

[robhrzic]

bemfarmer, I've posted this question in the Gecko Drives forum, also. Hope I don't get flamed for double posting.

[bemfarmer]

robhrzic, apparently tuning is done in mach3. There are instructions on the artsoft forum. Also there is a yahoo Geckodrive group, but I cannot understand the postings. lol

[robhrzic]

bemfarmer, yes, I am aware that the Mach 3 user's guide has a a "Tuning motors" (Section 5.5) chapter, where it shows you how you calculate steps per unit, steps per revolution, acceleration and deceleration curves, etc.... That's not what I am talking about.

I want to know how to go about critical damping / tuning the servo motors to the Gecko G320 drives, using an oscilloscope, as shown on pages 2-3 in the G320 manual:

http://www.geckodrive.com/product.cfm?pid=13

I have never used an oscilloscope before, and would like some sort of guidance / direction as to how to procede in doing this correctly, and most of all safely, without making any expensive "smoke".

I know exactly what you mean about not understanding!
:withstupi

[robhrzic]

I finished installing the sub-panel into the enclosure. I found a "hammer finish" green Rustoleum spray paint that seems to come close to the green on the mill, so I painted the enclosure. Now I am finally ready for Phase II of the wiring, and hooking up all of the connectors for the inputs and outputs.

I also got a chance to finish up the drive design for the Y-axis. I should have all of the stock already, so all I need to do is start making some chips.

[robhrzic]

I finished wiring up the machine control enclosure today. I took the covers off of the G320's and soldered a lead to the test point and ground point, which I'll connect to a common terminal strip for easy access for tuning the drives / motors with an oscilliscope. I wish that the G320's would come with an externally accessible test and ground point, as I REALLY, REALLY, REALLY, did not like soldering anything to the PCB inside the G320 enclosure. I hope I didn't fry anything internally doing this. The "test" access hole is totally useless in my opinion, and I do not see how you are supposed to access the test and ground point through this little opening, that doesn't seem to be anywhere near where it should be.

The connectors and sockets were a lot easier to (solder) use than I thought they would be, and they give the appearance of a really nice looking professional install. The only (male) socket that I've done so far is for the E-stop mushroom. I still need to install the sockets on all of the other input/output wiring.

I am a little concerned, as when I powered it up (not connected to a PC) for the first time, only two of the green LED's lit up on the status board connected to the combo break out board. I also noticed that as I turn the power on and off, I would get some of the other LED'd to light up, but could never get to see all four come on. Why this is, I do not know. On my previous power on test (with only the power and step, dir, and common connected to the G320's) all 4 green LED's lit up.

I hooked up a motor and respective encoder to a one of the connections to see what, if anything happens. To my bewilderment, the motor seems to turn and accelerate, then stall (fault light comes on the G320), repeatedly over and over. I guess the good thing is that it does do something, but i need to figure out why it's operating as it is, and how to adlust whatever needs to be adjusted.

There is a lot of wire and connections here!! I triple and quadruple checked all the connections before I powered it up. I hope that all of the connections are right, and all I need to do now is tweaking and tuning. I hope that I did not damage anything when wiring it up. At least I haven't seen any expensive smoke, yet.

I am really confused about the ERR/RES connection. It seems that if it is not connected to the ENC+ terminal, the G320 remains in a fault state. That's the way that I left it for now(ERR/RES jumpered to ENC+), though I am not sure if this is the right way to connect it, or if I need to do anything else.

I am getting 77 VDC at the power terminals on the G320's, which I believe is right where it should be. Though, I am getting about 7.4 VDC across the ENC-/ENC+ terminals. I thought that this was sUpposed to read 5 VDC. I am not sure why I am getting a higher voltage.

I have yet to receive my proximity sensors, which are backordered. Once I get them, I'll try to see if I can hook up the X axis, and see if I can make anything move under power.

I haven'y connected any of the output relays yet, as I am not sure if I'll be using any, yet. I also made a couple of extra access holes that I installed a couple of large grommets and plugs in, for and future additional connections that may be needed.

I did have to leave one fan out, as I did not leave enough room for it. It interfered with the AC input wiring from the transormer to the power supply prep module. I hope that the 7 fans that are in there will be able to keep everything cool!

[robhrzic]

I hooked up a motor and respective encoder to a one of the connections to see what, if anything happens. To my bewilderment, the motor seems to turn and accelerate, then stall (fault light comes on the G320), repeatedly over and over.

I switched the "A" phase and "B" phase wires on the G320's, and that solved the accelerating / stalling problem.

I set the gain and damp settings to the default 10 o'clock position, and the limit all the way CW. The servo seemed to be "singing" like it should. So, I figured that it was finally time to install the X-axis drive. I took the left side bearing retainer apart, and placed a .014" shim between the outer races of the angular contact bearings to get rid of the "sponginess" or "springy" feel that I have had in the X-axis since I put the table back together. To my surprise, I noticed that I had installed the AC bearings backwards.

Once I had everything put back together (with the .014" shim), I was pleased that I had no more "springy" feel in the handle. I have yet to actually physically measure the backlash, but I am pretty confident that, at least according to feel, that there is very little, if any.

Once I had everything tightened up, I fired up the controller, PC, and Mach 3 to see if I could make some motion. I tried to jog the X-axis, and I had a brief very slow and jerky movement, and then the G320 would fault. I was a bit disappointed, as I really wanted to see some smooth servo CNC motion. Oh well................

I e-mailed some questions to the support group at Geckodrive, and Marcus very quickly replied, and suggested that I separate the step and direction wires going from the combo board to the G320's, and to use shielded cabling to make this connection. Thanks again Marcus for the quick reply. I see why people say that Geckodrive provides top notch support

I have placed an order for some shielded cabling from McMaster Carr. Once it gets here, I'll re-wire the step / direction wiring and hopefully make some motion.

[robhrzic]

:banana::banana::banana:

Finally!!!!!!!!!!!

I re-wired the step and direction wires using (2) separate shielded 22 gauge cables. I am currently able to move the X-axis (using G0 manual input) at about 71-72 ipm. I still have to play with the pulse frequency and accel / velocity settings to get it up to 150-200 ipm rapids, which is my goal.

The only thing that concerns me is that the motor is still a bit "noisy" as it moves the table in the -X direction. When moving in the +X direction it is relatively quiet and smooth. It makes some noise though, and the handles on the table vibrate a little when going in the -X direction. It also seems to get a lot worse at lower feed rates.

When turning the handles manually (with everything turned off, of course) it is very smooth in both directions, no matter how fast or slowly that I try to turn them.

Anybody out there have any ideas as to what may be causing this?

[txcowdog]

So the shielded cable did the trick. Nice to see it move, but, for the life of me I don't know why you only get problems in one direction. I wouldn't think it is mechanical since you can not reproduce the problem manually so I have to think electronics and something in the resonance of the driver is causing your problem... but this is really just a guess.

[FPV_GTp]

hi

robhrzic excellent work and nice thread write up

cheers

[jarhead1017]

I know this is a old thread, but i just joined to get access to the pictures and information.

How has it worked out assuming you still have it?
Thanks, Joe