Ken Nicholas Photography Astronomy & DIY CNC

Blog posts : "General"

Problems with stepper control

Here we are it's January already and my machine is stalled due to a issues in the upgrade to individual stepper drivers and a 48V system up from the 24V single board driver.

The checkers board that was intended to be made prior to Christmas was not done neither was anything else done. I've been pulling what little hair I have left out trying to fix this issue of at least one axis not driving and at times two not working, currently it's two, the Y and Z do not run at all.

I have swapped drivers, exchanged leads, replaced the BOB with direct wiring to the controllers, replaced printer port leads and even ran a different computer but still the issue will not be fixed.

It has always been either the Y or the Z axis that have failed, one at a time or both at the same time. It would seem to be a poor quality of signal as I have fed the X axis signal to the Y and Z independently and they both function quite well using the X axis signals but they fail repeatedly with their own signals from the parallel port.
As it stands I have measured the voltages and frequencies for each axis and found that the Y and Z axis show lower voltage readings for direction control but also the step frequency is unstable when everything is plugged in.
If the printer cable is unplugged and I measure the step frequency for each axis they are identical 3.7 kHz but the Y and Z direction voltages are lower than the X by 0.7 Volts.

So when connected,

X is stable at 3.7 kHz step frequency and 0 - 4V direction control.

Y and Z drop to 0 Hz or float unstably and the direction voltage is 0 - 3.3V.

NOTE: Direction control is 0V in one direction and 4V in the other.

It seems unusual that two completely different computers should suffer the same issues and this is why I’m still tearing my hair out.
I have just ordered a USB to parallel port adaptor so I can try and run it from my laptop in the hope that the adaptor does some signal conditioning and supplies equal signal voltages and frequencies to all three axis, I should have the adaptor next week to try out and will again try through the original break out board that came with the three controllers and if that fails then I will try the direct port wiring as well, wish me luck.

Go Back

What's happening ?????

Well it's been a few weeks since my last post as things have been a bit slow happenening and number one is, believe it or not, still waiting for the "T" nuts to go on the back of the bed for the clamping system. The incorrect nuts have been sent twice, so I now have 80 of these little 4 mm suckers and with luck the correct 6 mm ones will be here for the weekend.

In the meantime I have installed the limit switches to each axis and also a KILL switch because things do go wrong even when you think you have it all under control.
I was cutting a perspex sheet for use on the router dust hood, cut from 6 mm sheet there was a 65 mm hole and a 40 mm hole, the first cut was to be the 65 mm with a 3 mm pass followed by a second 3 mm pass with tabs.
The first pass cut started and by the time it had gone a half circle the 1/4" router bit had plunged through the 6 mm sheet and into the sacrificial board (fortunately a 16 mm thick board) to all but .5 mm from cutting through to the table.

The router bit had worked loose in the collet and slid downand almost out of the collet, still cutting due to the side pressure being applied by the steppers. I went for the Kill switch that I didn't have.
This meant powering down the router and trying to stop the Mach3 software at the same time. No Kill switch, so what eventuated was a stopped router fully plunged still being pushed along as the bit hadn't fully dropped out of the collet it was caught by the narrowest margin by the collet retaining nut and it bent the router bit, see the pictures in the gallery.

So thats why the limit switches and a Kill switch became the next on the list of things to do.

You will also see the vacuum tube trial installation held in place on the perspex sheet and I will most likely fit a brush shroud as well. Also in the pictures you will notice I was trialing a fence to catch the cuttings etc and help keep the place a little cleaner as the perspex cuttings are statically charged and stick to anything and everything.
The trial worked well as you can see and stopped 98% of the mess going over the side, made for easy clean up afterwards.

Stepper drivers are to be changed due to the poor peformance of the original single board driver causing very slow transit speeds, that's in the next installment.

Go Back

Job Request

Hi again,

While still waiting for the replacement (correct size) blind T nuts to arrive my daughter who has also seen the potential of this little machine has submitted an order for a Chinese Checkers board.

At right you will see the 3D rendering from my CAD drawing rendered in Vectric Cut2D, the finished size is 13 inches measured across the flat sides. I will be running a test on a piece of MDF this weekend to see how it runs the tools. I will be using a 1/4" straight bit for the pockets and profile, a 1/2" 90 V bit for the bevel and star outline and a 1/2" round router bit for the marble holes done as a drill pattern.

The test run will only be on a narrower piece of scrap rather than cut the full board as all I want to see is how the pocket depth looks and also the bevels and marble holes etc so it will be about half the width of the final article.

More to follow.

Go Back

The fruits of ones labour plus 1/8th collet added

Still waiting for the blind "T" nuts for the bed clamping, the wrong size turned up so more being sent. While waiting I have been testing some more from Vectric's Cut2D samples, you will see in the gallery the locomotive wheel cut from MDF, a little fury from the spiral 1/8th cutter on the spokes but where the 1/4" straight cutter was used it was nice and clean.

The 1/8th cutter obviously needed a new collet to hold it and after looking around I found that a std die grinder collet was almost a perfect fit in the little Makita trim router. I did need to machine the upper shoulder down a bit by around 0.017" and then shorten it by 0.020" to stop it from bottoming out but it now fits perfectly and is a good quality collet, it cost me $30 and that was only after a bit of haggling.
The 1/8th end mill cutters from eBay arrived so they were tested out as well and seem quite good for the price.

The 25mm thick MDF bed is now screwed down and a level check with dial indicator showed a maximum bed variation from side to side of +/- about 0.007" or just under 0.2 mm and only about half that from end to end. So I'm pretty happy with that considering its made from MDF.

My daughter and wife are placing orders for things already, the green Perspex apple core is a stitch template for the wife’s patch work quilting and I just received a request from my daughter to make a Chinese Checkers board for the grandkids for Christmas, think I may be able to come up with something special for them.

I can also see I'm going to have to make a dust hood for the router, in particular when cutting the Perspex as it throws statically charged chips everywhere even if you chase it with the vacuum. Still have to tidy up the wiring and fit the limit switches just looking at different options there but I seem to have a pretty good handle on tool paths and am controlling it quite well.
Finally at this stage I am starting to enjoy the fruits of my labour but still a bit more to go yet, I'm also thinking about a perimiter fence to help keep the mess contained, it will be made from Perspex mounted in slide in rails to allow easy maintenance. Going to the Aluminium place tomorrow to see what they have that will work as a mounting for that.

More to come.




Go Back

First Cut

Hi all,

As the title states it was first cut day today. A test pattern of pockets was cut in a scrap board to check accuracy of the desired pattern and also the repeatablity of a second tool path over the first to cut a second series of pockets. The video shows that this was a success and after the three months of the build I'm very happy with results.
Feed speeds are still to be looked at a bit further but for the moment I'm happy to continue to tidy up the loose ends in wiring etc and finish the bed, the test pattern in the video is actually for the bed clamping I've devised, the pockets will be cut into the 25mm MDF slab that is the bed and drilled though to the top, blind nuts, T nuts commonly used in woodworking for furniture assembly will be fitted into the pockets and turned over only showing the through holes.

This will sit on a sub bed of 6mm MDF that will stop the T nuts from dropping out and will allow 6mm studs of any length to be srewed into the bed at forty locations, 100mm spaced so as to allow clampin of work pieces of any size virtually anywhere on the table.
This method I came up with was mainly to help reduce costs of using a T slot bed and I think it will work fine also allowing use of a sacrificial bed when cutting through.

Anyway have a look at the video, it's in HD but you can select lower res to help downloads and please comment in the visitors book, let me know what you think or if you have any questions.



Go Back

Motor Tuning

Well motor tuning has been done to the best that I can achieve with the driver board I'm using, its one of the ubiquitous Chinese manufactured Toshiba chipped single board for all three axis. I have played around with the settings on this board and within Mach3 for a week now, using the Mach3 instructions on how to tune the motors just didn't seem to work for scale and I was getting some strange results when telling X to move 1" and getting travel of approx 1.15" to 2.5" and anywhere inbetween.

So after setting 1/2, 1/8th and 1/16th steps on the board with slow decay and fast decay with combinations of those settings including varying current and voltage setting to try and achieve smooth running while not over heating the Z axis when it was mostly just being held and to result in the correct scale of movement was a pain in the proverbial.

Settings finally arrived at are 1/8th steps and close to those seen in the image to the right but I am not happy with the speed achieved, not that it's a real problem but it would be nice to be able to have the Gantry do a full X traverse that is approximately 26" or 660mm in less than a minute and fifteen, 75 seconds and a full minute for the Y to do its 475mm side to side travel which is the final speed achieved per minute.
With these results I am thinking I may need to upgrade to Geko drives or similar to be able to gain any sort of improvements on those here as the maximum revs are only 124.6 rpm, so from what I've read 400 rpm would be a more realistic top speed. All I get when trying to up the jog rate or feed speed is motor growling followed quickly by missed steps and motor stall.

The computer I have driving this is an older Win XP machine with a 1 Gig processor so I followed Mach3 recommendation to set the Kernel speed to 35000Hz, I had also tried altering the step pulse width from 1 to 3 and even 5us but to avail.

At least I have the scale sorted and it moves an inch when told to or I can switch to Metric without issues so I'm happy with that, this will do for the time until I finish other things that need to be done and will try for more speed later.

Go Back

Road Runner Test

Well it works.

Three months to the day from the start of the build, this was the estimated build time so I'm very pleased that it has worked out this way as this has been a spare time project still working full time.

There is a video on my Youtube channel

I haven't fitted the limit switches as yet and the wiring is only roughed in for this test run to take place. All has worked well although motor tuning is still a bit of issue as feed speed is slow and when trying to force faster feed the motors miss steps quite badly so a bit of work to do to get my head around the setup of all this as the Mach3 manual is no small read.


Go Back

Wiring under way

Wiring has comenced as the main structure has been completed and more quickly than anticipated considering some of the delays I've had.
Anyway the end is in sight and very much looking foreward to a full test run under power, but lets not get ahead of oneself as I still have to fit the limit switches as the wiring is being done so a few more brackets to make yet.

The cable drags have gone in nicely and the cables being used are blue LAN cable for the limit switches, more cores than required but it's nice and flexible and I had plenty lying around. The black is a four core again very flexible but it also has a quite thick outer that should stand up to rubbing and ware and tear during use, I picked that up from Jaycar, the cores are heavier than the stepper leads but that will reduce any chance of voltage drop to the steppers.
I haven't chosen an enclosure for the power supply and stepper control as yet as I haven't decided on a table for the machine. I do have an old steel framed office desk with three drawers down one side that would carry it quite nicely and also give me plenty of room to mount the computer etc under the desk and the drawers used for storing clamps, hold downs and cutters etc so it may well be moved to that very soon to see how it looks.

Go Back

Stepper Motor Fitting

Fitting the motors to the screw drives has gone well after changing my mind about three times as to how I was going to do this.
I had purchased the flexible couplings you see in the pictures as an option to the ones that came with the ball screw set, those were a one piece cut aluminium type.

The type used here allows for the motor to be removed without undoing any of the screws on the coupling so assembly and alignment is so much easier as well, much better.
The motors were to be mounted on standoff posts but the final decision was to use the 80 x 40 rectangular aluminium, again this made the shaft alignment so much easier and the whole assembly nice and rigid.

All three motor mounts are virtually identical and fortunately the four jaw chuck in the lathe made producing these easy so this final assembly stage is going well. Mind you it is still a lot of work with more parts to manufacture test fit and then final fit.

Getting close to starting the electrical/electronics stage soon. Still to do before that begins is the final loctite of the Z axis screws, refit the X axis bell screw and stepper motor and mount the cable drag chain on the X and Y. I still haven't decided on the positioning of the micro switches for the limit switching so a few more brackets to make yet.


Go Back

Ball Screw shaft end float

Getting close to final assembly time so I started to set the ball screw mounts up and in the process I found they all had an end float problem.

The amount was at a guess around 0.020" and so I went looking to find out why, I removed both the bearings from the main bearing block to find they were not a standard single row ball but an angular contact type, this is given in the number of the bearing "7000AC".

This type of single row AC bearing when used in this type of installation, to stop shaft end float, are fitted back to back so they can be adjusted to take up any free play.

This is where the problem was found as the bearings had been installed face to face and a second issue was also evident in that there was no room for adjustment as the inner race's were hard up against each other.

Other issues found, virtually no lubricant on the bearings and the surface finish of the sleeve's for the seals was very poor and coated with some sticky black substance.
See the gallery images for an explanation.

Corrective actions
1 Lubricate the bearings with a good quality grease.
2 Fit a shim spacer of 0.030" between the bearings to allow for end float adjustment, "must not interfere with inner race".
3 Install the bearings back to back "Part Numbers to Part Numbers".
4 Polish the seal sleeve's to stop them from tearing the seals.
5 Reassemble completely and adjust the keeper nut on the shaft end to remove all end float but not to over tighten.

Go Back

Custom Parts

The devil is in the detail.

Getting to this stage has been relatively straight forward but this has slowed down some what because the Z axis is proving to be very time consuming as assembly and disassembly test fitting everything is aligned with everything else is the culprit.

I have aligned the rails so it now runs smoothly top to bottom and then I had to make the drive plate to connect the ball screw to the router mount plate, the results you will see in the images at right. The easiest way was to bore the hole in a piece of scrap first and then cut the plate from that piece I started with. Finishing the block to the scribed line after rough cutting with the band saw was achieved with the disc sander quite well.

Both plates were then drilled and threads tapped to mount the plate but when mounted I found another issue, the ball screw shaft has not been manufactured perfectly and has a slight run out causing the whole assembly to bind slightly every half turn of the shaft and I have also found some end float in the Z axis drive screw bearing block as well so more work to remove the end  float but I don't think there is anything I can do about the out of round short of replacing screw shaft. Hopefully the Stepper will overcome the small amount of stiffness but time will tell with that so I'm not happy with it but will continue for now. Achieving the desired accuracy is becoming more difficult due to imperfections in the parts purchased and the small errors in hand finishing parts so a bit frustrating, oh well.

Go Back

Router Mounting

Well this week just gone didn't go quite as planned and not as much work was done as I expected.

Sunday was when I finally got to it for a bit of work and I've concentrated on the Z axis while it is assembled to this stage. I completed the mounting plate for the router and fitted it up to the Z axis rails, they are a little stiff as there is a small runout of one rail to the other, only a couple of thousands of an inch but enough to stop it running smoothly full travel but it won't take much to fix so no real problems there.

The eBay motor mount for the makita trim router fits quite well and gives the full range of hight adjustment, I think it is about twenty five mm or there abouts so this also good. Next I'll be pulling the gantry apart to fit the Y axis ball screw as I still have to drill the right side of the gantry for the shaft exit hole and stepper motor mounting. One thing that is surprising is the weight of the Z axis and never having made one of these before one hopes that the chosen Steppers have the required power to lift the whole shebang, but the motors I have selected are based on a similar machine that performs quite well so there should be no problems hopefully.

The third picture down showing the rule beside the router gives an approximation as to where maximum Z down is and also where the bed hight will be and that's the hight of the sacrificial bed. I've lost count as to how many times I've had this apart and back together again but a few more times to go yet.

Go Back

Z Axis

While I had the Ganrty assembly in place I decided to fit up the Z axis to the stage you see at right to allow for fit and alignment of the Y axis screw drive.

I have kept the Z rails as close together as possible to maximize the Y travel so its total width is 125 mm, this will give me a theoretical 470 mm Y axis travel but by the time I fit limit switches etc I'll be happy with 450 mm so I should get that without any issues. You will see in the gallery images I have had to cut into the rail mount base to allow the close fit up to the bearing blocks but this is not a problem.

Where the main Y axis plate mounts to the angles fitted to the bearing blocks the screw heads are hidden under the rail base quite nicely so these screws when finally fitted will be locktited as I think I will do most all the screws on final assembly anyway.

Things are still progressing nicely and I'm looking forward to putting in a fair bit of time this week on it so  I'll post more later in the week.

Go Back

Y Axis setup

Setting up the Y axis has been a bit more time consuming as I've chosen to make this from stock size materials so its been a bit of make it up as you go along. I have changed the way I'm approaching the assembly of the Y axis as the main plate that will carry the Z axis will mount to the two 60 x 60 mm angles you see here attached to the bearing blocks on the gantry and that same plate will also have the Z axis linear rails mounted to it so the bolt positions and assembly sequence will have to be thought out before I continue.
You will also see in the gallery photos the Z axis layout and beside it the motor mount to carry a Makita trim router.


Go Back

Gantry test fitting

Moving right along.
Well it seems to be going along well with the trial fit up of the gantry main frame assembly with only some minor adjusting to make on the Y axis channel, one cut wasn't quite square causing the side plate to not fit flush up against the linear bearing block. Still to that as yet and following that I'll mount the lower cross beam to the gantry plates and mark out for the fitting of the ball screw nut assembly so no major hick ups at this stage, fingers crossed that the remaining fit out goes as well.

I have made a small addition into the design and that was mainly to assist in assemble and making sure the gantry sides were kept in alignment with each other. You will see in the gallery images a shot of the linear bearings attachment point to the side plates, I have added a small piece of angle to both side plates ensuring when they were placed on the bearing blocks they would be sitting vertical and aligned the same on both sides, doing this made it a piece of cake to fit the cap screws as it held in place perfectly aligned over the mounting holes.

Time to pull it all apart again and do some finishing work, more holes to drill and threads to tap. If anyone wants any info regarding the build just drop me an email, my info is in the contact menu on the home page.


Go Back

Base frame assembly check

Things are moving along nicely after a day in the shed. Maybe I shouldn't say that as it's probably inviting some sort of failure next time I go out.

Anyway I couldn't resist wacking a few screws in to see how it's progressing, but then again it needed to be done to check alignment and fit as can be seen in the image at right.
This is just a quick check of the bearing block clearance to the bed when fitted as the bed will overhang virtually right up to the gantry sides.
You will also see at a later stage that I intend to overhang the bed at both ends of the table to allow ease of fixing beds or even work pieces.

I have also included in the gallery images as you will see an hammer and square purely for size reference to allow a better visualization of how big or small this machine will be, that just depends on how you look at it I guess.
I was originally only going to fit two bed rails but decided to add the centre rail for more rigidity of the bed when fitted as it will primarily be of MDF or just plain old chip board as a sacrificial bed for the whole base.
I had been recommended to use chip board to reduce the fine dust that MDF tends to make when machined. I may well use both, MDF for the permanent base and a thinner layer of sacrificial chip board.
I'm pleased to say that having assembled the base that all parts have so far lined up perfectly square and parallel, I have not allowed for any adjustment of the frame parts at this stage and that will be the test of my drilling and cutting in the end. There will no doubt be a need for it at some point for fine tuning.

Go Back

Linear Rails mounting

Much drilling and tapping of threads later.

The Linear rails will be mounted on the 6 mm wall thickness channels for the X and Y axis, I have chosen to drill and tap rather than bolt through with nuts and washers etc for rigidity and location accuracy, when final assembly is made all cap screws will be fastened with Locktite. Only about another fifty threads to tap.
The drilling jig (wooden block) held the rails quite nicely while drilling and hopefully maintain the accuracy I'm trying to achieve.


Go Back


So over to see Cameron, he's the one with real milling machine, not that mine isn't going to be real that is.Hopefully mine will be as real and functional when all is said and done with a little help from a friend, well for this bit anyway.
After rough out Cameron applied some of his milling expertise to refine the edges, saved me a lot of hand finishing that's for sure.


Go Back

Side Plate cut

Well I made the roughing cut on the side plates for the gantry today, I wasn't sure how the bandsaw would handle it but there was nothing to worry about as it went through like a hot knife through butter as they say. Off to my mates place tomorrow for the profile cut.



Go Back

Prepping Gantry sides

One part I dont have the gear to finish is the Gantry side plates so I'll be doing a roughing cut to size on the bandsaw and then a trip to a mates place to do the finishing profile cuts in his Bridgeport Mill.

Go Back

20 blog posts