I got a new HURCO KMB1

[bhurts]

You can also look at k-flop and kanalog. It will work with mach3 or they have a free g-code software that I prefer.

Ben

[tfuqua63]

You can also look at k-flop and kanalog. It will work with mach3 or they have a free g-code software that I prefer.

Ben

Thanks Ben I have contacted Dynomotion and now I will wait to see what there response will be. I liked there setup it did not seem to complicated and the price is more reasonable than some of the other routes I have been looking at. Hopefully this may be the answer I am anxious to get started putting the Mill back together again.

[tfuqua63]

Dynomotion was very prompt in answering my email they do offer 1 on 1 help for $500.00 for the first session up to 2 hrs and $250.00 an hr.for additional time
However they said to go to this forum and or there yahoo groups to find the help for free.
I dont know I like to have a salesperson or tech to talk with.
Any advice on using the Kflop and Kanalog would greatly be appreciated.

[bhurts]

I think you would find the help on the forums to be exceptional. I have not used kanalog but have used k-flop for several years and would not even consider a different system for my own projects. I already have a second k-flop and will be purchasing a kanalog early next year to do a hurco retrofit.(no time right now.) Not only are the prices reasonable the online support is 1st rate and it is such a versatile setup. My suggestion would be for you to join the dynomotion Yahoo group and search for setups similar to your own. Do the same in there forum here on the zone. It will help you decide if the support and products are right for you.

Ben

[mmoe]

For what it's worth, I have been using the CSMIO IP-A from CS Labs for over a year now, often doing large productions of parts that have the machine running 10-12 hours a day (I'm just a one man shop, so that's a long as I care to go). It has been absolutely flawless over the year. It could really go 24/7 if you need it to. It's easy to install, has a large number of inputs and outputs that run on 24 vdc (doesn't pick up interference like 5vdc can), and has worked great with my old DC servo motors and drives. There is very minimal setup to do once you've finished your wiring. For the most part, you just install their Mach 3 plug-in and do basic Mach 3 configuration. If you have more advanced level needs for running specific custom M codes, there is plenty of help since it's just the standard Mach 3 scripting.

I've worked with 3 machines using CS Labs controllers now, and I can't recommend them enough. In addition to the IP-A (analog version), I've now also used the IP-S and the IP-M. The IP-A is really the flagship of the bunch in my opinion. It may cost a bit more than the Kflop, but it's not that much of a difference in price that I'd make your decision based on that. Plus, the additional modules available for the CS Labs are plug and play of the highest quality. You can get I/O expansion capacity up to 288 inputs and 144 outputs, more than anyone could ever need. There is an excellent MPG module kit available as well. Rigid tapping is also possible, though I'm not sure if you need the threading module for that or if it's built in (I think the threading module is more for lathes?). All of the modules and the controller are DIN mounted to standard rails for fast installation, daisy chain together easily with the provided wires/dip switches and they have a very well constructed enclosure that I feel does an excellent job of keeping the electronics well protected. Plus, there are diagnostic LEDs built into the enclosure so that you can verify I/O and other signals at the controller itself, handy if you are looking in the cabinet and need that information (you don't have to go back to the computer to check).

I would guess that Kflop has a similar feature, but one of my favorites in the CSMIO IP-A is the fact that it will home on the index mark of the encoders. This feature provides very precise homing that makes using a fixture super easy to set up. I just note the work offset values in use for a fixture right on the fixture, home the machine, then set the work offsets to match what the fixture requires (assumes the fixture is indexed to the machine via indexing pins). The homing is so good that I can just run the program from there without having to fight with locating the work origin to the fixture (my old controller had a loose definition of "homed").

[tfuqua63]

I think you would find the help on the forums to be exceptional. I have not used kanalog but have used k-flop for several years and would not even consider a different system for my own projects. I already have a second k-flop and will be purchasing a kanalog early next year to do a hurco retrofit.(no time right now.) Not only are the prices reasonable the online support is 1st rate and it is such a versatile setup. My suggestion would be for you to join the dynomotion Yahoo group and search for setups similar to your own. Do the same in there forum here on the zone. It will help you decide if the support and products are right for you.

Ben

Thanks Ben for your help I will do some more research on the Kflop from Dynomtion I like what they have but was a little worried about support.

[tfuqua63]

Thanks mmoe I have looked at there products and did like that they seem to be plug and play. I also conacted them and they were very helpful and seemed to know what they were selling.The price tag was a little more than some of the other systems but then if you get what you pay for I do not have a problem with that.
You are the second to recomend the CS labs products so they are difinatly on top of the list.
Thanks again and I will check the CS labs out some more.
What mill did you retrofit ?

[mmoe]

I have a large industrial CNC router (4'x4' work area and weighs about 8000lbs) which I put the CSMIO IP-A into. I've also put their IP-M into a small lightweight CNC router and their IP-S into a heavier laser machine. The laser machine works OK, but Mach 3 is not the greatest for lasers and there are minor issues still to work out that aren't really related to CS Labs.

As for the IP-A, it really is about as good as I can imagine a controller can get. The ethernet connection has been rock solid with a distance of about 25 feet of cable between my control panel (computer) and the controller module, which was my main concern about the system when I was installing it (seemed like a lot of opportunity for interference on a critical signal).You have to make a custom cross over cable, but it's not that hard to do and there is a diagram provided to show how it should be made. I believe you can also order them pre-made from most computer supply stores online as it's a standard that is used when connecting two computers directly without a router between them, but I just bought the crimping tool from Radio Shack and made my own so I could make the length exactly what was needed and no longer. The inputs/outputs are simple to use and the instructions are very well written despite the fact that English is not the primary language of CS Labs who are from Poland. If you have any specific questions, feel free to ask and I'll do what I can to answer.

Another user here has installed several CS Labs products on his machines with great success as well. He goes by "Hood" and has converted his Chiron mill with the IP-A. Here's a link to quite a few short videos showing his conversion:

https://www.youtube.com/user/HoodScotland/videos

Here's some videos I've put together:

https://www.youtube.com/user/mmoecnc/videos


One thing I think you should also keep in mind is that the controller will almost certainly outperform the servos/encoders you have, so while they will work fine with older DC servos, the servos won't likely be able to go at the kind of speeds that the controller is capable of. I've been noticing that my servos feel outdated compared to the controller, and plan to upgrade to AC servos and drives soon, which will have triple the feedrate/rapid capabilities of the current servos, while most likely also having better acceleration while maintaining control of the machine. I'll also be updating the encoders to something very high count, hopefully something close to 20,000 ppr. Currently, I only have 2000 ppr (original encoders from about 1989) and I think that the trajectory correction suffers a bit when the controller doesn't have much data about the motion. I believe you can go to around 3mhz worth of feedback per axis, which would mean that a 3500 rpm motor could have up to a 50,000 ppr encoder count. My CNC router has very high quality, but course ball screws at 20mm/rev, so if my math is right that would be insanely overkill at .0004mm (.000016") resolution which would far exceed any real mechanical capabilities. Even 20,000 ppr would be running at a resolution of .001mm (.00004"), again far exceeding the actual machine resolution. If you have finer lead screws/ball screws, you would be dealing with even higher resolutions than that, though slower feed rates due to the finer pitch, so the controller is really able to handle a lot of data for trajectory correction (closed loop).

[tfuqua63]

I have also looked at Pico Systems PPMC does anyone know anything about them ?

I do like the Vital systems but not sure the added expense is worth it

Narrowing down the field now and hope to have an educated guess as to which way I plan to go.
I want to thank all for your input. I will be posting pics and the build as it progress's.
I am going Mach3/4 to start out it seems to be the way to go for the beginner.
My brother inlaw is a machinest and he suggested that would probably be the best route.
He would help me out if I decided to go with some of the other programs and help me program codes.

Options available;

Dynomotion KFLOP with Kanalog about $500.00
CSLAB CSMIO/IP-A $763.00
Vital Systems DSPMC/IP with 7535, 2-7711 about $1334.00
Pico Systems PPMC with other components $780.00

[tfuqua63]

OK I am ready to bite the bullet but still am undecided which way to go ?
Plan 1 keep my drives and servos encoders and tach and purchase the Vital Systems DSPMC/IP
OR
Replace my drives with Viper 200F's and go with the PMDX motion control and breakout board

At this point I will be brave and say cost is not a factor but ease of completing the retrofit and customer service as well as quality of the end result comes first.

Any input would be greatly appreciated.

My decision was based on location of products and who seem to offer the better customer service and both of these companies seem to handle that very well

Larken made several attempt to contact me by phone before we actually were able to have a 1 on 1 discussion and he was very helpful as well as the folks at PMDX

Vital systems also had very good technical help.

I am new to all of this and all 3 were very thorough in explaining things to me and it seemed to me they were only too happy to help educate a completely stupid new to the cnc world person.

My thanks to all who I talked with.

[mmoe]

I would go with Vital between those choices. If I understand the other products from a 2 minute glance, you're comparing a true closed loop ethernet based system (Vital) to a step/dir based open loop system (closed only between the drive and the motor, not the controller) with a USB connection. No offense to the PMDX or Larken, but I don't think there is much of a comparison here. USB is far more likely to pick up interference than Ethernet and is limited really to a 10 ft maximum before it really gets sketchy. Installing a controller into a larger machine usually places the controller and the computer further apart and that could be an issue. Ethernet is quite well hardened against interference and is more stable connection in general while being pretty much unlimited in how far you can run it between the controller and the computer. True closed loop is also a very big difference in my book. Find a machine over $50k that is open loop (closed loop between the motor and the servo drive is not "closed loop" because there is no trajectory correction, only error counting). Most people go with open loop because of cost savings, but since you already have the servos and drives to make closed loop happen, it makes more sense in my mind to go with the controller to match.

[tfuqua63]

Thanks mmoe thats what I wanted to hear so the decision will be Vital I have been leaning that way all along but was not for sure. The Vital system is a little more expensive but as you said a true closed loop ethernet system is the best way to go.
I will hopefully get parts ordered first of the week and start putting things back together again.
I will do my best to get pics and show progress as I go along.
I know you strongly recomended CS LABS there controller was a little less expensive but they have no distributor in the US, that for me put them out of first choice.
How would you compare between CS Labs and Vital Systems ?

[mmoe]

I was on the fence between the two and I doubt there is a whole lot to tell between them from a practical perspective. I think Vital may have already finished up their Mach 4 plugin, so that would have made it a tougher decision if I were to do it again now, but I'd probably still have chosen CS Labs (they'll have a Mach 4 plugin eventually, and I'm not sure Mach 4 is really ready for daily use yet anyways). The main things I liked about CS Labs (and still do) is that their system is modular and laid out very well for installation. It's a small thing, but I really like the way the CS Labs controller mounts and the way the CAN bus system works with the add-ons. It's really almost idiot-proof. Here's a shot of my controller right when I put it in the machine, before running the wiring. You can see that it lays out very nicely and is easy to run wire to/from.



When I compared with the Vital, you have to mount the Vital to the cabinet like a servo drive which is nice in that it's a smaller footprint, but then you have to make some sort of mounting plates to support the breakout boards which are sold separately at $55ea. (included when you buy the CS Labs IP-A and/or I/O modules). I considered that a bit of a hidden cost when I was comparing and when you need 6 of them it adds up quickly. By the time you buy the breakout boards and the controller itself, you could have pretty much bought two CS Labs IP-A controllers or three of the Kflops if the I/O built in is sufficient for your needs. In my case, I figured the extra was better spent expanding my I/O since I have a lot of control panel functions that I wanted to maintain. The link shows how their breakout board goes:

http://vitalsystem.com/portal/motion/7761b.jpg

I've ordered several controllers from CS Labs now, and in every case it arrived within about 4 days (they charge about $80 for the shipping though), so it's really not any different in that regard from if it were shipped in the US. If anyone reading this purchases from them, I would highly recommend that you buy everything you need the first time (I/O modules or MPG, etc.) and get the most out of the shipping expense! If I had an issue, I'm sure that I could get a replacement in a few days as well (I would probably buy a second unit and send the first back for warranty service if that were to happen), so I didn't really feel that it's any more or less of an issue for down time. I think that no matter what you go with, a complete controller failure would be several days of downtime. In the case of CS Labs, the nice thing is that you would just pop the old one out and pop the new one in, then flash the current firmware just to be sure it's the same. It would literally take 5 minutes to change it out if there were a problem and you wouldn't even need any tools. All that said, I have not had any issues with their controller anyways.

Even with the cost differential, I did give very serious consideration to the Vital controller, and I really don't think you'd be disappointed as I've heard nothing but good about it. Without having used both, my best guess is when it comes to installing the system, the CS Labs is somewhat more refined and easier to deal with. Even the LEDs on the face of the IP-A provide a visual check that helps with trouble shooting the I/O (LED indicator for each I/O). If you turn on a signal at the control panel (computer), then walk back to the electronics cabinet area, you can still see if specific outputs/inputs are active without going to the diagnostics window in the Mach 3 plug-in, which can often be hard if you're a ways away from the computer (my machine electronics are not within eyeball range of the control panel). Otherwise, I don't suspect there is any significant difference in performance either way when it comes to machining parts.

It's funny because I've always thought that CS Labs would be much more popular if they had a U.S. distributor as well, to the point where I have tried to talk the guys at the "cncrouterparts.com" store into carrying them (they are local and I met one of their guys while teaching CAM locally). I may have to look into it myself one of these days. I think highly enough of their products that I'd be willing to distribute them. I just don't know that much about importing as I'm really just a small job shop and CNC/CAM consultant for shops just getting into CNC and needing to learn how they can put machines to use instead of putting off using them. I've retrofitted and helped retrofit machines for other businesses, and it may eventually become a more integral part of what I do, but for now it's maybe more of a distraction than I really need.

[tfuqua63]

mmoe,

Thanks again for your input. I appreciate your opinion and why you made the decision you did to go with the CS Labs products. I do see how simple there system is to put together and there expansion system is really great you do have me giving them another look, but I since I started my research I have always been leaning towards the Vital systems product I will look into the CS Labs, maybe give them a call before I make my purchase for of the hardware.
You need to go ahead and be a distributor you have me sold on there product but them not being US is the main reason I am having a harder time considering there products.
I have a couple of other machines to work on and the CS Labs would be a lot less money for three retrofits over the Vital systems.

[tfuqua63]

After much deliberation on which way to go for my Mill retrofit I ordered the Vital DSPMC/IP-S 8 axis motion cotroller along with 2 of there 7711 breakout boards and an 7535 P breakout board.
Will update and post some pics on how things are going.

I still have not got the VFD functioning properly but have not really worked on it since the first time I got it hooked up. I can run the spindle in forward and reverse but don't have the variable part working yet I can run it at a slower rpm and a high rpm but not been able to get it to change when running Will hopefully get that resolved this weekend.

[tfuqua63]

The Hardware was delivered yesterday and already ran into a snag. I have single ended quadrature encoders and I need differential quadrature encoders I have to convert the ones I have so I contacted US Digital and have the parts on the way. I think I have this part figured out as to how the encoders are wired to the break out board, should be an easy project.

Now I have been trying to figure out how to wire the servo drives, the ones on my mill are FET servo drives there are seven wires to connect for each axis plus the 2 wires that power each axis.

I need help to figure out where these wires need to go on the break out board there are two pairs with a black and clear and one with a black,clear and white. I believe one pair will connect to the tach but not for sure any help here would be greatly appreciated I have attached some images hopefully this will help

[/ATTACH].

[Bob La Londe]

I didn't know any of the KMB1's had 5 hp spindle motors. Is it an NMTB -40 taper spindle?

My KMB1 has a Leland 5HP motor, but Hurco calls it a 4HP machine. Mine has the same Kwik 200 spindle nose as many others.

[tfuqua63]

Thanks Bob for the info.on your machine do you know anything about the servo drives I have posted pics of ?

[Bob La Londe]

I do not know much about the servo drives you selected. At the time I retrofit my machine I only studied the Gecko G320X drives, the Dugong drives, and the Viper Drives. I dismissed the Dungongs due to availability at the time and some negative comments by users. I chose the Gecko Drives primarily due to price, but also because they worked with single ended encoders so I could run the encoders that were already on my servo motors or run the AMT-102-V encoder which was so cheap if I had a problem with mine. As it turns out the 102V is physically too small to fit on the Electrocraft servo motors, so I wound up using USD-E6 1000 line encoders. AS it turns out I could have used the E5s just as well. The G320X encoders worked ok, and I can run my machine at 150ipm under nearly any loads. As it turns out I may be able to run them faster. I'm in the middle of an upgrade, but as soon as I finish it I'll test again. The current limit pot raises the current limit by turning the pot COUNTER CLOCKWISE. LOL. All this time I've had my servo drivers set fully clockwise. LOL. I do have to mention that I tested the machine at speeds upto 250IPM, but turned it down to 150 for reliability.

Another thing to think about with regard to line count depending on your PC interface is acceleration. I think some motion control combinations struggle with higher line count encoders at high speeds. I am running 1000 line encoders. With 5 IPM leads that's 20,000 lines per inch. That's a theoretical resolution of 0.00005". The mechanical parts are nowhere near that good. Running all three axis in a rapid generates a lot of pulses for your controller to handle and it won't be any more accurate. My machine originally had 300 line encoders. That's 6000 lines or a theoretical resolution of about 0.00017. That's still better than the mechanical bits. If I knew then what I know now I might have gone with a something like 500 lines instead of 1000, and I think my machine might have been able to manage much faster rapids. (Well, that and turning up the current limit. LOL) Getting back tot he beginning of this paragraph. I think reducing the pulse stream might also help with higher acceleration rates without faulting. If you are doing a lot of detailed complex 3D work like I am acceleration makes more difference in job time than high rapid speeds.

[mmoe]

He's got plenty of headroom with the Vital to run 14bit and most likely 16bit (just over 65000 ppr) encoders without any issues up to perhaps 3000rpms (probably more). That's why you spend the big bucks on a controller like that (my CS Labs will apparently take 6mhz of data, per channel, or 16 bit encoder resolution at around 5000 rpms, so I expect the Vital is not far off from that) . If you are getting faults, you have to go back and see what the following error numbers are and adjust them to something reasonable. You can't have a following error max of 10 with 16 bit encoders. If you have that kind of encoder resolution, you can set the following error much higher and it's still equivalent to a lower number with less resolution. The advantage is that the controller will have more data for trajectory correction and will perform those corrections better. Acceleration should not be affected by more data, since the controller is capable of taking that kind of data without problems.