Losing steps/position, again

If these are wired in a unipolar manner - which is sounds as though they are... you will need to isolate the centre taps. I appreciate the buggeration that comes with this - and recognise that at this time you're probably trying to keep the machine in a recoverable format (normally I'd just cut/isolate the cable at the stepper), but I understand you may want to be a bit more circumspect. the centre tap will be either to ground or to +V and it'll be causing problems for the driver - certainly your described behaviour I can imagine with the centre taps connected.

Doddy - I wasn't too worried about leaving the machine recoverable I was simply worried about cutting the wrong wires and disabling Y and Z.

This appears to be getting complicated.
Out of the stepper motor I have Black - White - Orange - Red - Blue and Yellow
Blue and yellow are not connected to the cable
White is connected to the yellow in the cable

I can't be sure if these are the same in the cabinet but in the cabinet I have the following pairs
Read and Yelow
Black and orange
These are all labelled X

The Stepper motor is Sanyodenki 103H7124

Attachment 27782
Attachment 27783
Attachment 27784

So by my reckoning I I am connecting to the taps on the stepper?

It's simple to test just check the resistance between the three wires of each coil. One pair will be double the resistance of the others. These two are the ones you want for each coil.
Do the same for the other coil and you have found your 4 wires needed. Just leave the others disconnected but terminated for safety.

If I understand Carl properly, what he's saying is (I think - please confirm) from the first paragraph: that one the centre-tap to one end of the coil is wired into the cabinet - a half-coil configuration. This presents a reduced torque for the machine (obviously by Denford design) - and those motors aren't particularly torquey to begin with. You can work with this for now at least - it was always a choice on this conversion.

But you're still just getting a knocking noise from the stepper under drive?, can you take a photo of the driver, in particular the connectors/wiring AND the DIP switches? I'm curious about your current setting and the microsteps.

If I was in your boat I'd be setting the driver to 2.5A and 8 microsteps to begin with... maybe even just 1 step to keep things simple. I'd be measuring the volts on the supply to the driver, and even swapping out the driver for another just to check that.

If we're ever allowed out I'll offer you a second pair of eyes if that would help - but I can't see that happening any time soon - and I think you'll resolve this a long time before then.

EDIT: One of those DIV268s has a label, something like "Modified" on it - try that one, it's geared to handle fast signals.

Original pairs has resistance of 2.3 and I have now wired up pairs with a resistance of 4.0
Still no movement just a knocking noise though the knocking isn't as harsh as it was (if that makes any sense).

Quote:

---

Originally Posted by Doddy

EDIT: One of those DIV268s has a label, something like "Modified" on it - try that one, it's geared to handle fast signals.

---

That's it. Changed over to the "modded" stepper and I have movement - but slow.

Set steps to 2 and X now moves at same speed as Y (by eye).
The X axis stepper is silent now when running compared to the whining of the Y stepper.

Quote:

---

Originally Posted by ngwagwa

Set steps to 2 and X now moves at same speed as Y (by eye).
The X axis stepper is silent now when running compared to the whining of the Y stepper.

---

Do you understand the relationship between Micro steps and Steps Per setting in Motor tuning that need setting so you get correct move distance? Also, the affects higher and lower micro-steps have on the motors.?

The first is something you need to know if you want correct movement and the second is worth knowing get a better idea of the effects of micro-stepping.

If not just say and will explain.

Quote:

---

Originally Posted by JAZZCNC

Do you understand the relationship between Micro steps and Steps Per setting in Motor tuning that need setting so you get correct move distance? Also, the affects higher and lower micro-steps have on the motors.?

---

Not a clue to be honest I would appreciate an explanation. After homing the machine appears to be going to the correct point when I zero all.

One thing I did notice that wasn't shown well in the cabinet photos is a "Control Transformer" with outputs marked 28, 110, 120 and 240 if that's something we can reuse. All my parts should be coming next week.

Interesting that "Modded" worked - that suggests your pulse width (somewhere in the motor tuning pages of Mach3 by memory) is less than 3uS. Try changing that to 5us for reliable operation.

EDIT: Silent?, Nonsense - you'll appreciate silent when you replace with a digital DSP controller.

Quote:

---

Originally Posted by Doddy

Interesting that "Modded" worked - that suggests your pulse width (somewhere in the motor tuning pages of Mach3 by memory) is less than 3uS. Try changing that to 5us for reliable operation.

---

They are all set at zero! Can I leave them for now and look at them when I go over to UUCNC?

Quote:

---

Originally Posted by ngwagwa

Not a clue to be honest I would appreciate an explanation. After homing the machine appears to be going to the correct point when I zero all.

---

If that case you have been lucky and set MS to same as old setup (you can pick my lottery numbers)
Edit: No your probably not lucky, obviously it will go to the same point because the Home switch stops it and sets the zero. If you tell it to move say 20mm does it move 20mm.?

Ok well, it's not hard but it is important. I'll try to keep short n simple but it will still be long.!.

Your steppers will have a base count number of steps per revolution, often this number is 200 or 1.8deg per step. But this can be different depending on the motors but I'll work on that for this example because it's most common.

Now the drives offer a function that allows each step to be broken down into sub-steps or Micro steps (MS). The drive will offer a range of MS dividers ie: 2, 4, 8, 16, etc or sometimes you'll see them wrote as 400, 800, 1600, 3200 etc. They all mean the same thing and tell you how many times they are going to divide each step up or like in later example how many steps make up one Revolution of the stepper.

Now the controller needs to be told how many Steps(pulses) it must send out to move the machine 1 unit of measurement, So if set to metric how many Steps per MM.
This is determined by a calculation for which we need some information about the machine.

We need to know how far it moves for one turn of the motor shaft. Now linear movement can be done in many ways ie: ball screws, rack n pinion, belt-drive, etc but this doesn't matter all we need to know is how far it moves for 1 turn of the motor shaft.

The easiest way is to just turn the motor one Rev and measure how far it moved but this need to be measured very accurately. But we can also calculate it if we know some details about the machine. In this case, I'm going to work on using ball screws.

So using a 5mm Pitch ball screw and set the drives to 8 which is like saying 8 x 200 or 1600 MS we will get the following Step Per figure.
1600/5=320 Steps per MM

Now if the ball screws and motor are connected via a belt with a ratio then we need to factor that in as well. So lets say we have a 2:1 ratio then we get.
1600/5 =320/2=160 Steps per MM

This is the number you enter into Mach3 Steps Per setting so that the machine moves the correct distance.
The Steps Per setting is very important and not something that can be guessed or random numbers entered.

The speeds you'll get from the machine will be determined by the Steps Per value and the Frequency you have available. The pitch of the screws, motors and voltage etc all play a part here and are why a balanced setup works best. I'll leave that for another time!



Now the affects Micro stepping as on the Motors and the system as a whole gets a little deep so won't go too far into it. I'll just give the key advantages and downsides.

The main advantage to micro-stepping on the motors is they make them run smoother, you'll often see it mentioned it offers high resolution, which it does to some degree but it shouldn't really be relied upon. The largest effect is on the smoothness of the motors.

The down side is it lowers the motor torques and puts a higher strain on the device which is creating the pulses.

Now to you both of these are important because in the first place these motors don't have massive Torque and your also using the parallel port which is limited in frequency and on how fast it can pump out pulses.

So in your case, you'll be better using lower MS because it will allow a higher number of pulses which translates to potential higher feed rates, but you still want some motor smoothing, so don't drop right down to the lowest setting as it can make the motors run very rough.

4 x (800) is a good number but some times you may need 2 x (400) to give you the speed because of the in-balance as mentioned above and lower frequency etc.

Some times a certain MS setting won't suit the motor because resonance in the machine when working can affect how the motor runs, so in cases like these, esp with a parallel port and cheap drives, you have to move the MS range up or down to find a sweet spot the motor likes. This happens a lot on analog and shitty drives.
Newer Digital drives are much much better at handling resonance and along with faster motion controllers like the AXBB completely transform a machine, even with old lower torque motors.


I hope this hasn't blown your mind.
Simply put if you don't set the Steps Per correctly it will never be accurate and if you set the MS too high you'll stress the parallel port and run slow.!!

Quote:

---

Originally Posted by ngwagwa

They are all set at zero! Can I leave them for now and look at them when I go over to UUCNC?

---

Your call.

I'll tell you the behaviour if the pulse width is too low... You lose steps.

Quote:

---

Originally Posted by Doddy

Your call.

I'll tell you the behaviour if the pulse width is too low... You lose steps.

---

Right I will change them when I go in tomorrow.

Quote:

---

Originally Posted by JAZZCNC

If that case you have been lucky and set MS to same as old setup (you can pick my lottery numbers)
Edit: No your probably not lucky, obviously it will go to the same point because the Home switch stops it and sets the zero. If you tell it to move say 20mm does it move 20mm.?

---

Sorry I didn't explain that too well, after homing the machine it went to the work datum ie G55 0,0,0 (machine position X-355, Y-130 and Z-45}. I was assuing that as none of the Mach3 settings had changed all would be good.

Thanks for the explnation and I get the general idea - the "steps per" on the current setup is <Steps0>31.50619418</Steps0> guessing maybe imperial pitch lead screws.

I am hoping to have all the parts for the retro-fit by the end of next week (the toroidal transformer may be later as they say This was only ordered today! It takes 2/3 days for us to wind the order.) Anyway I can hopefully forget any serial port imitations when the new motion controller is installed. Changing the wiring on the Y and Z steppers is going to be a bitch if they are wired the same as the X axis.

Meanwhile I will set a job running tomorrow and start working out the layout of the new parts and how to get them in the control box.

Quote:

---

Originally Posted by ngwagwa

Thanks for the explnation and I get the general idea - the "steps per" on the current setup is <Steps0>31.50619418</Steps0> guessing maybe imperial pitch lead screws.

---

No that will be metric. If it was imperial it would be much higher number.

Apparently I havn't got as much room in the cabinet as I thought.
Attachment 27787

Quote:

---

Originally Posted by ngwagwa

Apparently I havn't got as much room in the cabinet as I thought.

---

QFT: That single statement is what we've all felt like at some time. Today I've been stacking BoBs on top of BoBs, and Relay boards on top of PSUs... and the cabinet always looked too big on paper.

It's not big or clever, but there's always the third dimension.

Quote:

---

Originally Posted by Doddy

It's not big or clever, but there's always the third dimension.

---

That's why I modelled it solids.
Attachment 27788

Sorry, I think I'm somewhat less than clear. My post, for what it was worth, was lamenting that I've ran out of space in my cabinet and and busy stacking boards and components - into the third dimension. Intended a little tongue-in-cheek.

One quick suggestion - having stared at some real-estate in my cabinet where a PSU burnt itself out - a lesson learnt to not seal enclosures without cooling - if you've not already planned this, make sure that you have a cooling fan or two in the cabinet for extracting hot air.

Advice much appreciated - my layout was posted to see if anyone commented on it.
I am using the original Denford control box which has an air inlet bottom right and (I think) a 4 inch fan top left. It gets a good flow through as in winter I had to put an external deflector on ot to stop the cold air freezing my legs.

One for Jazz.
Got my PM25 clone running today and works brilliant but...
I used overkill Nema34 (the grief of my X2 knocked me a bit) seem good motors so that's all fine.

Anyway. If I hit my soft limits at 2000mm/min the motors slightly over travel and put my position out by 0.2mm.
I would say this is down to the rotor inertia of the overkill motors. Would you agree?

Simple solution... Don't hit the limits :whistle:
I do have some decent nema23 size for it. I'll make some mounts so I can try them out.

Quote:

---

Originally Posted by dazp1976

One for Jazz.
Got my PM25 clone running today and works brilliant but...
I used overkill Nema34 (the grief of my X2 knocked me a bit) seem good motors so that's all fine.

Anyway. If I hit my soft limits at 2000mm/min the motors slightly over travel and put my position out by 0.2mm.
I would say this is down to the rotor inertia of the overkill motors. Would you agree?

Simple solution... Don't hit the limits :whistle:
I do have some decent nema23 size for it. I'll make some mounts so I can try them out.

---

Soft limits shouldn't be out by any amount. The point of them is that the controller controls the velocity/acceleration and position so it doesn't let you go past the soft limit value and brings the axis to a controlled stop, unlike when you hit a physical limit switch which just turns off the motor output. In this case, you always get over travel from inertia if moving quickly and even at lower feeds you should consider position lost if you hit the E-stop or limit sw.

OK, Step pulses set to 2 - what about the Dir pulses (currently zero)

Regadng my rewire here is the bottom rail that wouldn't have been too clear on the previous pictures. I suppose I can use some of these for splitting the rectified output from the toroidal transformer?

Attachment 27789

I'll bow out of the conversation on re-using DIN rail terminals - I'm not familiar with them enough to advise (Only recently realised there's such things as shorting links). If you have the rail installed (which you have) it feels like a good opportunity, but others here with more experience will offer a more accurate view.

About the DIR - probably best to be set similar - if you miss a DIR change then you might end up with a single step in the wrong direction - but as this would only be on change of direction the error is unlikely to accumulate.

I take your point though that you'll be migrating to UCCNC next week anyway... I wouldn't sweat it too much.

Quote:

---

Originally Posted by JAZZCNC

If you inventory what you have and find the voltages etc and tell me what controller you are going to use I will happily help you work out a wiring schematic.

---

Here is the list:

Existing items

Power supplies:
24V 3A
24V 14.6A
12V 2A
RFI Filter
Omron relay MK2P-S
Omron relay G2R-2-SN
Control Transformer" with outputs marked 28, 110, 120 and 240
DIN Rail
Fan (voltage unknown will check tomorrow)
Main on/Off switch


On order:

Airlink 500VA 230v to 2x45v toroidal transformer

Power supply 50W 5V DC 10A
bridge rectifier 400V 35A
4OFF Electrolytic Capacitor, 4700 µF, 100 VDC

ABXX-E Motion Controller

3-OFF Stepper Motor Driver 6.0A, 80V CNC Microstepping CW8060

Terminal Blocks and wires


I am pretty sure I can wire the ABXX to the drivers and steppers and the VFD is seperate so really all I need help with is getting the mains to the PSUs and on to the devices and the E-Stop.
I think I can work out the limit switches.
I thought it would be tidy to use the existing connectors on the rail to get my rectified output from the toroidal transformer the the drivers, I could have the wiring run behind my mounting board.

Obviously if I have missed anythng please let me know.

Do you have momentary Button on the machine that can be used for safety Reset. You really need this as part of the E-stop safety system.

Quote:

---

Originally Posted by JAZZCNC

Do you have momentary Button on the machine that can be used for safety Reset. You really need this as part of the E-stop safety system.

---

I asume you mean the big red one on the front? If so then yes.

EDIT.
No that will be the E-Stop button won't it?

At the moment if I bend over and catch the E-stop with my backside I click on the Mach3 reset button to continue.

Quote:

---

Originally Posted by ngwagwa

I asume you mean the big red one on the front? If so then yes.

EDIT.
No that will be the E-Stop button won't it?

At the moment if I bend over and catch the E-stop with my backside I click on the Mach3 reset button to continue.

---

Ermm probably not usually Big Red button is E-stop, show me a pic.

Quote:

---

Originally Posted by JAZZCNC

Ermm probably not usually Big Red button is E-stop, show me a pic.

---

You may have missed my edit before you replied but yes big red is the E-stop and Mach3 reset turns everything back on.

Attachment 27794

JAZZ I have just remembered the E-Stop switch stays in when pressed, I have to turn it (I think) to get it back out.

Quote:

---

Originally Posted by JAZZCNC[/QUOTE

I hope this hasn't blown your mind.
Simply put if you don't set the Steps Per correctly it will never be accurate and if you set the MS too high you'll stress the parallel port and run slow.!!

---

So, if the machine is running right if I change the steps on the driver from 1/4 to 1/8 I need to double the steps per in UUCNC?

I believe you use cncdrive motion controllers? If so do you know if the utility will see the controller on the network when UUCNC is connected? I can connect with UUCNC but it keeps dropping the connection and I'm concerned that the utility has never found the controller.

Quote:

---

Originally Posted by ngwagwa

So, if the machine is running right if I change the steps on the driver from 1/4 to 1/8 I need to double the steps per in UUCNC?

---

Yes

Quote:

---

Originally Posted by ngwagwa

I believe you use cncdrive motion controllers? If so do you know if the utility will see the controller on the network when UUCNC is connected? I can connect with UUCNC but it keeps dropping the connection and I'm concerned that the utility has never found the controller.

---

I've never had any issues like that, they either connect and stay connected or they don't connect at all. Never drop out.?

You say "on the Network" are you trying to run this over a network ie: thru a switch or directly from PC to controller.?
Also are you using a patch cable or just an ordinary network cable because some network cables can be wired differently to others.!

Quote:

---

Originally Posted by JAZZCNC

You say "on the Network" are you trying to run this over a network ie: thru a switch or directly from PC to controller.?
Also are you using a patch cable or just an ordinary network cable because some network cables can be wired differently to others.!

---

PC connected to a hub then hub connected to controller, both standard cables nothing else connected and even tried 2 different PCs (I have a crossover ordered to try)

Quote:

---

Originally Posted by ngwagwa

PC connected to a hub then hub connected to controller, both standard cables nothing else connected and even tried 2 different PCs (I have a crossover ordered to try)

---

Try it direct to controller first to elimnate the hub

Quote:

---

Originally Posted by JAZZCNC

Try it direct to controller first to elimnate the hub

---

Direct connection and no joy, am trying to resolve with CNCdrive support

Attachment 27907

Yhey are telling me I need V1.8 or newer utility tool but I can't find it on their website.

This might not help at all, but for each of my 3 ethernet based motion controllers I installed a second ethernet port and created private LAN between control PC and the motion card without any switch (or hub?, really?, you really don't throw anything away :) ), subnetted accordingly. Now, I'll also say I've never used a cross-over cable but think that most PCs nowadays are auto MDI-X sensing and will sort that out for you.

Quote:

---

Originally Posted by Doddy

This might not help at all, but for each of my 3 ethernet based motion controllers I installed a second ethernet port and created private LAN between control PC and the motion card without any switch (or hub?, really?, you really don't throw anything away :) ), subnetted accordingly. Now, I'll also say I've never used a cross-over cable but think that most PCs nowadays are auto MDI-X sensing and will sort that out for you.

---

Thanks Doddy. PCI ethernet card on order! Have realised my laptop can use a standard patch cable and have tried that today but still disconnects.