[whateg01]

I recently bought a used SpectraLIGHT benchtop milling machine. I didn't get the computer, but I did get the control box, all steppers, and the spindle motor. From what I've read the control box is worthless without the computer, so I may reuse the case for what I have.

I was contemplating scrapping the steppers for something else, as I haven't found much info on them. I have a few similar motors that I thought of using, but then decided to get out the ohmmeter and check a few things. Turns out there really isn't any magic to these steppers at all. It's just that nobody seems to have posted much about them. I don't know if all of these machines are the same or not, but here's what I found about these.

Sorry. No pics yet. I'll try to get some before I really get into it.

There is an 8-pin connector on each motor. Of these, it first looks like one broke when pulling the connector. I've read other stories where the owner thought they broke the pin off, too, so it wasn't just me. Pulling the shroud back from the connector on the wire harness reveals that the 8th pin is a key. So, I didn't break anything after all. The motors have 7 pins remaining, while the 8th was broken off to make room for the key. On the motors I have there are no markings, but the female connector seems to have pin-1 opposite of the key, so we'll go with that. I haven't figured out the polarity yet, but pins 3, 4, and 5 are one set of windings, with 4 being the CT. Pins 1, 2, and 7 are the other set of windings, with 2 being the CT. The windings seem to be about 0.8 ohms end to end. Haven't figured out much more about them yet.

My other motors are 100 on-in, like these are supposed to be, and are almost exactly the same size. But my other motors are the old "HP printer" steppers so they have a much higher DC resistance, 8 ohms across the full winding.

I'm not sure if it's possible to get an accurate estimate of the current needed for these secret steppers or not. I haven't found much data on them, but would prefer to use these, as the couplers won't fit my other steppers.

[whateg01]

I should also mention that I decided to use LinuxCNC. I had a computer given to me that I wanted to use, and downloaded the latest LiveCD, which had Ubuntu 10.4 and LinuxCNC 2.5, I think. However, it wouldn't boot. So, I downloaded 8.X and 2.3 and it ran. So, I installed it all and everything seems to be working great. I did have to download a program to test the parallel port and then mess with the BIOS to activate it. But all-in-all, it wasn't too bad. A computer novice may have had a bit of hair pulling, but it does look like there is plenty of support for LinuxCNC, so maybe not.

I bought my driver board from HobbyCNC. I've heard that dealing with them can be a bit difficult, but hopefully, I won't have any troubles. I've read quite a bit over on the Yahoo group, and it does look like the responses tend to be less than friendly most of the time.

Anyway, I want to test out the software and hardware on my plasma table, but I want to get the mill running to make a couple of the parts for the plasma table. A bit of a chicken and egg scenario. I guess I'll just have to see which is ready first.

[cameraguy]

I have seen the inside of a couple of those in the past, they were really rather rudimentary due to their age; old round-can steppers and some pretty inefficient and weak old-style unipolar low voltage l/r drives IIRC.

I would just dump the lot and reuse the enclosure. Put a Gecko 540 or individual Keling drives or something something in it, probably be cheaper in the long run and certainly less headache than trying to get the original drive hardware moving again. Besides, even if you did get it moving again, they didn't run even as well as a crunky old Xylotex board even when they were new due to thier design limitations (i.e. verrrry sloooooooooooowly). Pretty much the model T of desktop CNC driver.

[whateg01]

Well, it's been awhile, but I have the machine running. Made my first chips - er, sawdust - tonight. For the sake of making sure everything was okay, I went ahead and used what was there. Got some info from Home [MeshCMS] that helped me get started. Still had to get the scope out and do some tracing of signals to get everything working right. I think there may have been some differences with different vintages. Glad I understand logic. Anyway, I'm using a PC running LinuxCNC and the existing control box. I found that I am limited in the RPM that the motors will turn, and I imagine, from past experiments with steppers, that is largely due to the way they are being driven. I am going to try to get another HobbyCNC board and run them from that. I think I'll have better response. Otherwise, all seems pretty good.

I ended up getting a CNC4PC BOB to get signals in and out of the PC. The biggest reason for that was that my PC wasn't able to drive the inputs of the controller. So, with the BOB putting out a full 5V, it works.

Here's the first thing I ran!



Dave

[whateg01]

Here's the video...

Spectra LIGHT mill making the first chips - er, sawdust! - YouTube

Dave

[ZipSnipe]

So yours never came with the card that plugs into the computer? I had one that I regrettably sold awhile back. But I had the card, I think its a lsa slot that only old computers have in them and thats what I had running it was an old computer. The accuracy was not good, was always around .005 off. But I did kinda like the controller software , was fairly easy to use.

I agree with the other member about getting rid of its current steppers and controller box and getting the gecko and new steppers for it. It was a fun lil cnc machine while I had it.

[Bowman]

I have one I got off ebay for less than $300 and without any control box etc. So I junked the Uni-steppers and got some Bi-Polar steppers and a Xylotex board with a power supply and retro fitted the machine. I haven't really used it much since doing this as life had gotten in the way but I can say that it works fairly well for the amount of work I put into it. I would go with Gecko Drives if I did it now, Xylotex is ok but not up to the same quality as Geckos.

[whateg01]

ZipSnipe,
No, I bought just the controller box and the machine itself. The computer is long gone. The guy I bought it from never got it and never figured out how to drive the controller.

Bowman,
This was a similar deal. About the same money, though I did spend a bit on gas driving to get it.

What I have found is that the steppers on it are bipolar, wired unipolar, so all I really need is the driver board. I am leaning toward the HobbyCNC board, as that is what I am using on my plasma table and I have had no issues with it. As I stated, the downside to using the existing steppers is that I don't know much about them, so I would be kind of guessing on the current settings. I guess I can measure the static voltages at the pins of the steppers and the winding resistance and just call on Mr. Ohm's law to get the rest.

When I was trying to figure out the CNC thing way back when, I was using a set of FETs and a bank of resistors to experiment with. I found that I had a severe lack of acceleration. The way this machine works kind of reminds me of that. That's why I'm thinking that the steppers are probably okay for what the machine is intended to do. They just need some better drive.

There are a couple of things I don't like about the machine. First, in the enclosure, it is difficult to get to the leadscrews for lubrication. Not sure much can be done about that aside from a new enclosure. Second, I don't know what is inside the enclosure, but whatever it is sings pretty loud when it's just sitting there. I'm guessing it has something to do with the spindle drive, but I'm not sure. Right now, I'm just controlling the spindle manually. I have a couple of toolchanges in the program and that's my queue to turn on the spindle.

I am running LinuxCNC with it and although it took some work to get it all figured out, I am pretty happy with the results so far.

I'd like to add an ER-16 spindle, but that isn't on the immediate list. Maybe after I get my big lathe back together, I'll turn one.

Dave

[ZipSnipe]

Well I too had that problem life getting in the way and I didn't use mine for a long time, basically had writers block, so I sold it.

Sure enough after I sold it, I find my new calling ....Pellet guns and all the cool mods that I could have used it for :-(

Don't sell it !!!

[NSR_CIVIC]

I have the same machine. Old round can steppers with no model numbers.. I'm currently building a drive system with a tb6560 board and some odds and ends. But I have yet to get movement. Mainly because I don't want to damage the drive board. How are u currently running yours? My original drive box was cooked so I traded it for a pair of 3 and 4 axis tb6560 boards but kept the spindle controller. I just got my 24v PSU this past week so maybe ill finally get to cut something.. what voltage are you running? I an not sure if these motors are rated to 24v or not.

Kevin.
Nsr customs.

[whateg01]

Kevin,

I am using the original control box. I don't know off-hand what voltage it runs at. I can see if I can probe it on the drive side of the resistor and see what it is there. Depending on what I end up spending getting the front of my car back together (post-wreck), I may order another HobbyCNC board today. I've got power supplies of different voltages available to me so I'll just try to match what the existing box does. I had been trying to figure out how best to get the current rating (the important parameter) for these steppers. I'll try to check that this week sometime. Need to get the car back together first, though.

Dave

[whateg01]

When you say you kept the spindle controller, which part is that? Do you mean the part that is in the base of the enclosure? How do you plan to control it?

[NSR_CIVIC]

I didn't get an enclosure with mine. It had the spindle controller that I believe was mounted in the enclosure. It's a minarik dc motor control with an e-stop and a pot for manual speed. I am working on setting it up to receive signal from mach3 but have not tried it yet. In theory it should work but I haven't fired it up yet.. lol
If it works I have all the documentation I used to set it up and ill post it here once I confirm its viable. I don't want to be responsible for cooking someone else's parts..

Kevin.
Nsr customs.

[whateg01]

Can you take a pic of what you have? I have thought about controlling mine through LinuxCNC, which I think I can do. According to the info I've found on the spaceopera page, I think it just receives an analog voltage, but I'm not sure.

If you want to PM me whatever info you have, or think you have, I'd be happy to look it over, too. I'm not worried about cooking mine. I can do some reverse engineering myself.

Dave

[NSR_CIVIC]

my spindle controller looks like this.

i think it was mounted in the enclosure.
i have the datasheets for this controller its a minarik pcm21000a and yes it does need an analog 0-10vdc signal. i have built a circuit that will convert the pwm output from my tb6560 board to the 0-10vdc signal needed but i have not tested it yet.
here is a shot of my mills.

spectralight on the left, sherline on the right. both are CNC "converted" but neither are running as so yet. the spectralight was more complete so i decided to start with it. the sherline mill itself was brand new in box. but it was missing all the cables for the "backtrack MFG." controller and the steppers seem to be welded directly to the lead screws which i don't like at all..
my driver boards and drive enclosure attached.
im using the 3axis for this mill, the enclosure pic is outdated. i have replaced the computer PSU's with a 24v 15a unit and rearranged it a bit. ill get an updated pic tomorrow morning.

[whateg01]

Thanks for the pics! I'm not really familiar with those drivers, but what are you figuring the current rating for the steppers to be? I just put a DVM on the leads going to the motors and was seeing about 2 VDC and 1 VAC with it just sitting there. I guess I was expecting it to be a static voltage, DC only.

That looks like the PSU setup I have on my plasma table. Couple of stacked PC PSUs. Seems to work! I got a few xfmrs from some equipment at work that was being scrapped. As I recall, we get about 18 V of rectified DC out of them, so I am thinking of stacking a couple of those and then rectifying it. That gets me closer to the "ideal" 36 V that HobbyCNC recommends. I'll probably run it a couple of volts lower, just to be safe, but it should be better than the 24 V for accel.

[NSR_CIVIC]

I honestly haven't messed with current settings. I received the boards from a gentleman who is familiar with these machines and he set the dip switches in advance. I'll look at the settings and see what its at.

Kevin.
Nsr customs.

[NSR_CIVIC]

My board is set up for 100% current, 100% decay, & 1/16 Steps.
Apparently the person I got it from is familiar with the spectralight mills and he says with these settings and his xml file I should be able to get 700 ipm.. not sure but we'll see.

@bowman: are you using a counter weight on your z axis? I see something like it in your pic. Is that something you found necessary?

[whateg01]

I don't see any counterweight on either machine. The Spectralight doesn't seem to need it, regardless. I think I can get a lot better speed out of my machine than I am getting now. I haven't "clocked" mine, but even at the recommended 30 ipm, the leadscrews seem to be turning awfully slow.

[NSR_CIVIC]

On his pic it looks like there is an eye loop on top of the spindle motor and a cable/ rope going up and over then down behind the desk. Weird.. lol
Are your cuts coming out at the right dimensions? I believe these are metric lead screws and if your program is set to sae screw pitch it may think its going faster then it actually is.

Kevin.
Nsr customs.