KMB1 Conversion

[Snipes44]

I built a custom control panel with a touch screen. Here is a picture. I made a swing arm that mounts to the top of the right hand side cabinet.

[Bob La Londe]

I have figured out how I am going to mount the monitor. Keyboard, controls, etc still to be determined.

I'm going to drill and tap some holes on the side of the head for an articulating monitor swing arm. This way I can easily swing the monitor to the front when I am doing setup, and working in the shop, and I can swing the monitor towards the door so I can see it instantly if I come into the shop or out of my office into the shop.

I am going to use a 15 pin gender bender in the side of the top cabinet as a bulkhead connector. The monitor will be able to plug directly into that for video signal, and right below it I will install a 120V outlet to plug in power to the monitor. I am thinking I'll install a single outlet rather than a duplex receptacle so there is never an open outlet to build up crud. If I can't find one I like I suppose a I can put a child cover on a duplex receptacle I suppose. (There are always options)

Speaking of 120VAC. I am going to need 120VAC in the top cabinet for sure. For the computer, for the monitor, for the cooling fans, and possible for low voltage power supplies for misc components, relays, etc. I am thinking the easiest way to do this would be to just install a heavy duty power strip in the cabinet. (Not one of those $5 ones from the drug store.) Then route cable out of the cabinet and plug it into the 120V outlet on the wall next to my 230V outlet. Can anybody see any issues with doing that? The only thing I can see as an issue is how to setup the safety interlocks on the doors to make sure when 230V is on the interlocks are on. All the 120 stuff would be hooked up satisfactorily for exposed usage so having 120V still on would not be an issue, and I could have the computer powered up for testing stuff with 230V off.

Well, off to the store again. I want some nice counter sunk head machine screws to mount the monitor swing arm.

[Bob La Londe]

Seems like the more I get done the more I need to do to get this machine going. I actually have not done much. As you can see I added a swing arm to put the computer monitor out in front of the machine for convenience when setting up a job. It will fold up flat out of the way against the side of the head when its not needed, and I can swing it over by the cabinet so I can see the display while working on the controls or troubleshooting.

You might be able to see that there is an electrical outlet for the AC power to the monitor mounted in the side of the cabinet. Panel mount connectors for video to the monitor, and for keyboard, mouse, e-stop etc will be added when they come in. The cabinet will be completely sealed except for intake and exhaust cooling fans when its finished.

I have the keyboard and mouse location figured out as well, but I need to fabricate a couple parts for that yet. They will swivel below the monitor on the same mount with a small box underneath for an emergency stop button. I'll probably make it wide enough to hang a pendant beside the e-stop.

I have not done a lot inside the cabinet, but there is a little work that has gotten done. I finally got around to installing the intake fan on the back of the cabinet. The intake and exhaust fans are towards the rear so there is no chance of liquid cutting coolant spraying into the electronics cabinet. I am hoping with the intake fan at the top and the exhaust fan at the bottom the amount of dust drawn in will be less. With the computer shelf added it nicely divides the cabinet and forces air to flow in around the computer and out around the servo controls.
An air flow test works as hoped. With just the intake fan blowing I can feel a substantial amount of air coming out through the exhaust fan. The cabinet isn't even sealed up yet. There are still half a dozen conduit connector holes and bolt holes from its previous setup that are still wide open. I am fairly satisfied that as long as either fan is still working I should not have anything overheat in the cabinet. I still have not decided if I will leave the cover on the control computer or not.

There are a few other changes. Obviously a shelf for the control computer has been installed. Most of the remaining garbage has been removed. There is now a power strip inside the cabinet for the 110V controlled and powered stuff like the PC and the monitor. Also it will be a handy place to plug in low voltage power supplies for relays and stuff like that. I moved the heat sink and servo controllers over directly in front of the exhaust fan so that it has to draw air over the cooling fins no matter what else is happening in the cabinet. The DC power supply for the servo controllers and servo motors is finished. My math says that should be adequate for normal usage, but I am considering the addition of a second capacitor to level out ripple under higher current draw when all three axis are moving.

Next steps.

Bulkhead/panel mount connectors for keyboard, mouse, monitor.
Long pin bolt for monitor mount to extend below the monitor and provide a swivel mounting point for the keyboard shelf.
Install the Smooth Stepper drivers and start configuring the Smooth Stepper.

THEN I can finally start connecting and configuring the machine.

I had a broken servo motor that I got a replacement for. I have not reinstalled that motor because I will use it for configuration and testing without risking crashing the machine into its mechanical limits while I do.

[GT390]

Bob, I am following this post, and refering to the other KMB retrofiting by Jack and Dave, I have nothing to add yet, as I have just started to gut out one of my Hurco`s, I have two KMB-1x machines. Iam going to go with the geko / PMDX 126 combination that Jack used.
Really I just wanted to thank you, Jack and Dave for taking the time to post the information and thus helping others like me to pluck up the courage to take this on. I am sure there are many other guys out there that read and gain knowledge from this forum too, so please keep on posting and updating us all with your progress and some more pictures.

Thanks again

GT390

[Bob La Londe]

Bob, I am following this post, and refering to the other KMB retrofiting by Jack and Dave, I have nothing to add yet, as I have just started to gut out one of my Hurco`s, I have two KMB-1x machines. Iam going to go with the geko / PMDX 126 combination that Jack used.
Really I just wanted to thank you, Jack and Dave for taking the time to post the information and thus helping others like me to pluck up the courage to take this on. I am sure there are many other guys out there that read and gain knowledge from this forum too, so please keep on posting and updating us all with your progress and some more pictures.

Thanks again

GT390

Danno is another who I learned quite a bit from and helped me work up my nerve to tackle this project... I decided to put the PC in the cabinet as mentioned in a previous post, and I am now waiting on the rest of my bulkhead connectors, for monitor, keyboard, etc. The control cabinet will only be "open" for the intake and exhaust fans when finished. I am debating similar fans setup and sole openings on the high voltage cabinet.

I'm using a Smoothstepper (USB VERSION) and C23 breakout board, but I am also using the G320X step direction servo drivers.

I'm hoping by using the smooth stepper to take load off my PC and having closed loop control I'll be able to hold this machine to its mechanical accuracy. Atleast at slower feeds.

I'm still trying to figure out how to do a code switchable companion spindle. I may use the mist coolant command to switch a contactor for 230 and control wires from one spindle VFD to the other.

[GT390]

I have taken the first steps and have ordered my PMDX-126 B.O.B., PMDX-107 spindle board and 3 Geko 320`s.

I am curious as to how much over-voltage the geko`s may handle safely, when they say they are rated at 80v max, can a few extra volts be that bad, do they maybe has a over voltage protection built in ?

I am in Ireland and my electricity supply is 415v three phase, my machine powers the servo drives off one phase at 220v stepped up with one transformer to 415v into the second transformer, giving me 59vac out to the rectifier, that is giving me 85.6vdc at the servo supply tails with no load, what I never did was check the voltage here under load, I am wondering if it will be necessary to take turns off my transformer, or if the rating on the geko`s has a safety margin built in, a no load over-voltage of only 5.6v does not sound too severe to me, has anyone checked the voltage to the drives loaded and unloaded to see if it drops ?

[Jackal66]

Guys,

Mariss at Gecko told me that the 320x's were checked above 90volts. It seems he may have said they are tested at 100+volts before they leave the factory. ( safety measure only. DO not operate at this voltage above 80 volts)

This is because when an axis is coasting down it can generate more voltage that what the power source is supplying.

The PMDX-107 spindle board was really easy to connect.

The VFD has the directions marked out and I just connected 6 wires????????.

Steve at PMDX (morestuff here on cnc zone) is very helpful.

I don't get to check in here as often as last year.

Good thing that last summer was slow and it gave me time to do the retrofit.

I am covered up now 7 days a week. < Not complaining

You might also check in at the MACH3 forum.

Keep the pics coming,


Jack

[Bob La Londe]

I contacted Mariss (or Marcuss I forget) and they said that the G320X "could" handle upto about 100 volts for a short time so the 90 I had by default would probably be ok. I had a problem with a different Gecko driver on a different machine and they chose not to warranty it, so I chose not to over power these drivers. I unwound my power transformer, and I get 78 VDC off the terminals of my power leveling capacitor. When I looked up the servo motors on my mill it seems Electrocraft spec'd them at 45 volts, but Hurco was running them at 90. I figured 78 was a good compromise. We shall see.

[Bob La Londe]

Ok.. I'm a packrat. I admit it. The keyboard tray started out as a 6 foot long drop down work shelf in a service truck work cap. It was always in my way so I took out back in 2008. It was laying on my scrap heap out back. A borrowed benchtop bending brake, a few holes drilled, and a few pop rivets later I have a servicable keyboard shelf, and it matches the decor. Early American Scrap Heap.

I bought the arm for the monitor, but it was cheap. About $60 I think. I would have bought one with its own keyboard tray, but all the ones I saw were a couple hundred dollars or more. I just couldn't bring myself to pay that for a basic shelf so I made one.

I still haven't got my panel mount USB connectors in so the cables are just poked through a hole in the cabinet for now. Tomorrow if I have time I'll install the Smoothstepper drivers and start trying to figure out how to hook up a servo motor.

[Bob La Londe]

Being the second Smooth Stepper I setup I figured it would be a piece of cake to install the driver and plugin. My plan was to be well on my way to configuring the breakoutboard and the servo drivers by now. No such luck. The SS driver installed just fine, and the plugin seemed to do what it wanted, but Mach refused to bring up the little selection dialogue.

After fighting with it for a while I decided to do a fresh root install of everything. I started with Mach (latest version), and BOOM! Everything worked like it was supposed to. I'ld read where they say download the latest verison in the SS manual, but I figured I was current enough. NOPE!

Oh, well. Atleast now the PC is talking to the SS and some glitches in operation have smoothed out.

Next was some more time reading manuals. Geckdrive's manual says something about getting the noload RPM of the motor at the voltage level provided directly from the power supply and then planning all the rest of your calculations at about 80% of that. Ok.... at 78VDC off my unregulated (with a giant cap) power supply the servo I used for testing turns almost EXACTLY 2000 rpm no-load. It fluctuated from 1996 to 2002 overall, but when I took care to keep my optical tach pointed directy at the reflective sticker on the shaft it only fluctutated from 1999 to 2001, and mostly it read exactly 2000. I wonder if there is any significance to that?

So based on that and Gecko's suggested 20% derating I should plan for a max of 1600 RPM peak out of the servos. 1600/5=320 IPM. (I seem to recall reading that the ballscrews are 1" - 5 TPI, but I could not confirm that at the time of this posting from my manuals.) That's over twice my initial target rapid speed. I figured I would start at 150 IPM, although the Hurco manual does claim they set the machine originally to rapid at 250 IPM. Then again Hurco was running them at a little higher voltage too. 90VDC +/- as oppsed to the 78VDC I am providing to the servos. I guess the acid test will be the same as on my little mills. Load the table up with a bunch of weight, and see when it starts to fault. Its going to be a while before I get to that though.

Next is to install the new 1000 line encoders, and do intitial tests between the motors and the geckodrives.

Then to start figuring out my breakout board.

[Bob La Londe]

Correction. The ballscrews are 1.25" x 5 TPI. I just found it in the manual in section:

4.3.3 Lead Screws And Nuts
...where it says 1.25 inch diameter and 0.200 inch lead per revolution.

P.S. I don't trust the manual 100%. The spindle motor is labeled as 5hp 3phase by Leland Motors, but the manul says its 4HP. Its got a 3.7kw VFD in the cabinet. That is apx 5HP. Still 4HP seems like a good safety margin. That would be a 20% derating. Pretty typical.

[Bob La Londe]

Limit / Proximity Sensor KMB1

Ok... I need some help. I'll post this as a separate thread also in case somebody else is looking.

The Z-Axis clearly has some for of proximity sensor for limits. There are covers ont he X & Y that I assume protect similar sensors. Another zone member also said that is what is under those covers.

I could use some information on how they work, how they are powered, and what kind of output they provide. If somebody will even just point me to a product info page I'll be glad to research it myself and report back what I find.

I seem to recall somebody here had said the sensors was a standard part available form guys like MCM or Digi-Key.

I spent all afternoon and evening reading manuals and planning my next steps. Having working limit devices before I move the table and quill under controller power is something I NEED to be considering now.

[Bob La Londe]

Thanks to some help from bloke I found this:

Found this so far:
http://stevenengineering.com/tech_su...DFs/31PROX.PDF
Page 32

I finally got out my little step ladder so I could look in the top of the head. The number was just barely visible on the proximity sensor.

Its 101XK5-1. According to the document above it looks like it is 5VDC Normally Open. Now to decipher exactly what that means with a 3 wire device. I use 4 wire and 2 wire sensing devices all the time in my day job.

[Bob La Londe]

As is often the case I got a bit side tracked with the time I had available to play with the mill today. I may play a bit more after dinner.

When I discovered the proximity / limit sensors were 5VDC powered devices I decided to see if I could power them from and hook them directly to my breakout board which has 5VDC all over it with pull up or pull down capability on most inputs.

That meant it was time to power up the breakout board. I have felt the wiring diagram and the documentation was a little confusing ont he C23 and power was no different.

Ok... unless you read everything on every document it can seem like you can power the board from the Smooth Stepper. YOU CAN NOT. Even if you read everything it can be a bit confusing. You can only power a very limited part of the board from the Smoothstepper. It still requires an external power supply. Totally separated from the SS power. I'm not sure if I would classify this as poor documentation, because it is there if you look. Its just that the various required circuits to be powered is somewhat ambiguous.

Where you hook up the main power to the board has some additional wording that can also mislead you to think these are just some extra terminals for powering addititonal 5VDC devices. Its not as clear as it could be.

Anyway... I have the board responding to and communicating back with the computer throogh the Smoothstepper now, and have tested several signal inputs and outputs, and have turned the charge pump back on.

I still don't know for sure if I can hook the proximity sensors up directly to the board inputs and 5VDC power, but I suspect I can once I figure it out.

I also installed one of the new encoders today on a free standing motor.

Either after dinner or sometime tomorrow I'll hook that motor up to a servo controller and some power to do some basic testing.

[Bob La Londe]

P.S. I am really starting to consider a window in the cabinet door again. There are a ton of diagnostic LEDs for every little thing on that C23 board. I'm good working with the door open for now, but when I start powering up motors I'll have enough DC in that cabinet to weld with.

[Bob La Londe]

I would sure appreciate if somebody would double check me. Feeling a little punchy today.

Here is my latest quandry:

C23 with 5VDC Proximity Sensors

NPN NO 5VDC
Sensor Labeled as:

EDIT - DO NOT HOOK UP A SENSOR THIS WAY - SEE MY NEXT POST

Black -
Red +
White [Load] +

Here is how I guess the connection to the C23

Board 5V to Red and C23 input.
White to C23 input.
Black to board ground.

Is that correct?

The sensor is found here 101XK5-1
Steven Engineering - Page Not Found
Page 32


The C23 Documentation does show hooking up an NPN NO proximity sensor here:
http://cnc4pc.com/Tech_Docs/C23R3_1_USER_MANUAL.pdf
On page 9

I think I have it right, but theirs shows an external power supply, and a resistor to drop voltage down to board level. Since I have proximity sesnors that operate at the same voltage level as the board it should be simpler as described above.

[Bob La Londe]

Okay........ I was way over thinking the limit sensor connections. I did some experiments yesterday and risked frying my 5V power supply. (Its a little low current for the application, and I need to replace it soon anyway.)

If I hooked it up the way I said it would short the board 5V to ground. I thought that's what it would do, but the connections in the C23 manual (admittedly for a different setup) would lead one to think that way.

Since the only source I found for an exact replacement is $75 each (I have since found an alternative for about $13) I decided to simulate the sensor with jumper wires. Yep. The instant it activates power drops out across the board. That is what I expected. Amazingly it didn't kill my power supply.

Then I just looked at it as if I hadn't seen any confusing wiring diagrams. Again simulating with jumper wires. Swapped pull up/down setting jumper for the input (so a short to ground of the input instead of a short to 5V (stock setting) changes the input condition on the board. Works perfectly. I had to change the high low setting in Mach and then Mach worked as predicted as well.

This morning I received this picture in my email. That is the way I "tested" it (except wire colors are different). I'll connect an actual sensor this evening after work and see if it works as predicted.

[Al_The_Man]

Pretty much what was described here?
http://www.cnczone.com/forums/hurco/...nsor_kmb1.html
Al.

[Bob La Londe]

Pretty much what was described here?
http://www.cnczone.com/forums/hurco/...nsor_kmb1.html
Al.

Pretty much.

[Bob La Londe]

I can see the light at the end of the tunnel, and I am convinced its not an oncoming train.

I moved a free motor with the whole conglomeration today. I've finally got just about everything figured out.

I have a problem with my pulse count though. I'm running a US Digital E6-1000 single ended encoder. According to my math that should equate to 5000 pulses per inch. Not even close or 1000 pulse per revolution. Not even close. It takes a setting of 20000 pulses per inch to be right.

The G320 has a pulse multiplier, but its set to 1. Besides its a pulse mutlipler not a pulse divider. LOL. Is it counting on and off on both A & B as pulses and adding them together?

What obvious thing am I missing? Maybe something with the C23 breakout board?

I cycled the motor at speeds equivelant of upto 250 IPM (1250 RPM) with no load and it worked fine. The in position light only wavered at the very end of travel, and the fault light never came on. I also had accelleration set at 20 so the reversal forces were enough to make the motor jump slightly. Now I need to hook up some of the motors on the machine.

Overall I am feeling entirely to pleased with myself today. LOL