Printable View
I don't know if this has been covered and frankly at 2:30 in the morning I can be arsed to go back over the posts and look. BUT have you checked that the wiring to the stepper motors is solidly connected and that there are no internal breaks in the wiring. Is your machine new or 'pre loved'?
A quick bit of clarification, the "ON DP" marking on the DIP switches (the 4-inline blue switches) - the ON represents the position of the "On" state of the switch. So it's not a case that you have K3 on, but rather that you have K1, K2, K4 on. Looking at the markings on the board that represents 8 micro steps. Change one axis (and one only) to Off, Off, Off, Off (all toggles towards the numbers, away from the "ON DP" and retest.Quote:
Originally Posted by Palletlad
(I'm assuming in the chart on the board that '1' represents a switch in the ON position... not a safe assumption, but the existing settings support that)
Note, that axis will run at 8x the original speed, so you'll have to tune the motor for that axis.
Do this with the motors still hooked up to the machine.
I think removing the steppers from the machine to test... interesting if this does present the problem, but I doubt that it will - the sound is mechanical, and likely related to the stepper performance in situ with the machine. Doesn't mean the machine is toast - just that you have to track down the system, rather than component behaviour... altogether more fun.
This is probably shutting the stable door......... but, only bugger about with one thing at a time.
I know it's a temptation to change multiple settings on multiple axes in hardware and software at the same time but all that does is confuse you and make it impossible for anyone else to draw any conclusive conclusions.
K1 and K2 is the microstepping. K3 and K4 is the holding current. SW1 - SW3 are the motor current settings. As it is configured in the picture, it is set to 1/8 micro stepping, 50% holding current and 2.5A motor current. Micro stepping means that the total number of steps per rotation (most steppers 200) must be multiplied by the number set by K1 and K2, in your case 8. It means that the driver must receive 8 times as many step pulses as without micro stepping, so 1600 pulses needed for a full 360 degree rotation. It will simply take smaller steps.Quote:
Originally Posted by Palletlad
I guess you mean "jam"...Quote:
Originally Posted by Palletlad
Can it be that what you call jam is actually stalling caused by too high speed or acceleration for your machine? Or maybe something mechanical, misalignment for example. Disconnect the motors and try to turn all axes by hand. Is it easy all the way? If not then start with fixing all the mechanical issues first. All screws must rotate pretty easily and most of all, smoothly all the way from one end to the other. If the linear guides are misaligned or twisted then you will have problems, even if you may be able to force your way through using more current.
Going by the picture, K1 and K2 set the microsteps.
K3 and K4 set standby current.
And the SW set motor current.
If the new computer has made no difference, then that should rule that out.
You've got nothing to lose by trying with the motors disconnected from the machine.
What's making me think it's not mechanical, is the seemingly random position the thud occurs, and that it affects more than one axis. If it was something like a bit swarf in a ballnut, I'd expect it to stall completely until the direction is reversed. I'd also expect to be able to feel something moving it by hand. You could perhaps try running the screws with a cordless drill and see if you can feel anything, as something like swarf may only show itself at speed.
However, to me it sounds more like a glitch in the electronics side. My next step would be to try a different power supply, or hook up an oscilloscope to see what was going on.
To answer your previous post, don't get an all in one system. Get a separate BOB, and if funds are really tight, get some of the basic TB6560 stepper drives. They're far from high performance, but they'll match what you've got, and individual drives can be swapped later if needed.
I do plan on doing the checks in stages but I thought about seeing if I can prepare a check list and go through it step by step.
I have ordered another db25 cable so going to check the cable.
I am going to grab lithium WD40 spray and give the ball screws and rails a coat.
I am going to test the motors on their own.
If still no luck I will check the board settings as Doddy suggests.
If still no luck then I will purchase PSU I found this one https://www.ebay.co.uk/i/15261107092...9916224&crdt=0, are these safe to purchase or do i need to go to some specific website to avoid dodgy components?
then go from there by purchacing a bob and stepper drivers as recommended by M_C.
Thanks guys btw for the help thus far and don't panic I am not going to do all and everything in one go.
I tend to agree - but am not convinced yet that this isn't an interaction between the electrical side and the mechanical side. I fear a scope is literally beyond the scope of the investigation so far, and a PSU should show problems simultaneously on jogs on two concurrent axis (and likely exacerbated by such).Quote:
Originally Posted by m_c
OP: Are you familiar with microcontrollers?, I'm thinking in particular I could knock you up an arduino sketch to generate a pulse train to step an axis over a predetermined distance, reverse, etc. Even check out the effects of pulse-width and so on. This would give a health-check on the machine from controller through steppers (including PSU). But for this you'd have to be happy with how to program an arduino and how to connect it to the parallel port interface on the controller (3 wires - ground, step, dir).
VM (Max) for a TB6560 is 40V. I'd suggest a 36V PSU.Quote:
Originally Posted by Palletlad
Quote:
Originally Posted by Palletlad
Don't forget the cordless drill test on the ballscrew that m_c suggested - it's a good idea.
Just checked, and I should of said get TB6600 drivers if you do replace them.
TB6560 is the older chip, whereas the TB6600 is newer, far more resilient, and has better performance.
Which has also reminded me that the TB6560 has quite a large pulse width requirement. It needs a 10uS minimum pulse width, which limits it to 15KHz step rate. Some where in the Mach settings (think it's the motor tuning screen), there is a box for setting the pulse width. Make sure you have it set long enough.
The attached chart shows the key differences between the chips.
I was about to dismiss this - the controllers I've seen (and modded) have had a 74-series monostable oscillator to pulse-extend the clock signal - however the image of the board (and various clones thereof) don't appear to have this luxury, unless it's rear-mounted on the board? If not, this board is seriously junk. The 10us minimum pulse width is actually dependent on a capacitor value, and can be significantly longer than 10us dependent on value. Mach3 advertises a limit of 1-5us for step/dir pulses, but appears to let you enter a larger value (I can't test as my motion controller is hard-limited to 2.5us). I'd agree - experiment with larger pulse widths (and reducing the micro step to 1 will also help to limit the PRF to a sensible value)Quote:
Originally Posted by m_c
EDIT: My bad - just checked, and the controllers I did modify are TB6600s, not 6560s.
Just a quick update, I have confirmed that there is no issue with the frame alignment or ball screw.
I tested the motors when not connected to anything and even on low settings (320,1000,150,0,0)=motor settings.
And they still jam/jar.
so definitely something to do with the electrics/settings.
Ooh just remembered to ask I saw the TB6600 stepper motor drivers via amazon and ticked them for a purchase later idea, however when looking for a standard BoB (I know so little on them) I found https://www.ebay.co.uk/i/25208589694...4960578&crdt=0 and I kept having this one pop up https://www.newfrog.com/product/elec...waAsFpEALw_wcB everywhere I looked but not many others? I would have thought there would be all sorts of brands, low, medium, high spec etc
I would not go with the drives all on one board for the same price you could have :- https://www.ebay.co.uk/itm/5pcs-CNC-....c100677.m4598
or
3 of these:- https://www.ebay.co.uk/itm/20KHZ-CNC....c100677.m4598
The bob you linked to is OK if you want a cheap one. I have used them a few times
What's the
part?, have you not tried larger pulse widths on step/direction?Quote:
Originally Posted by Palletlad
Good. At least we know that the mechanics are OK. Now change the pulse width to 10. If that doesn't help KEEP it there and change speed to something really low, like 500. If that is still an issue then you probably have a faulty driver card, or a faulty PSU, or both.Quote:
Originally Posted by Palletlad
I started off with a 3 axis 6560 (blue card). Fixed all the electronic design issues but still managed to fry two before I gave up. In the end I bought a $5 BoB and a handful of DQ542MA drivers and never looked back. My CNC immediately became more quiet, more powerful, faster and with much better acceleration, even though at that time I still had a crappy 24V switched PSU. Changing the driver was the key to success in my case because it solved a lot of issues. That's the route I'd advice you to walk through even if you manage to sort out the current issues. I am still using the same drivers but upgraded the PSU and the CNC in many other ways...
Doddy I did attempt with pulses as well but no difference it was the same result what ever setting I attempted.
Is this a good BoB? https://www.amazon.co.uk/Motion-Cont...breakout+board as I am unsure what I am looking for other than how many axis, connector type (USB/Parallel) would this be a good card to have https://www.amazon.co.uk/gp/product/...9I2AQIQH&psc=1 also https://www.amazon.co.uk/Switching-P...ywords=psu+48v.
I wont be buying straight away if need be as I still want to check the PSU 1st before any buying but I understand my currant card cant handle a 48v only 36v but is the one I linked good for the above linked items?
Would the setup I have linked be a good idea or are they not a good combination?
Cheers guys for your time.
Did you not read post #54 as the drive above is very similar and nearly double the price
I think I have 5 of the ones that you linked to sitting in a corner collecting dust... will check tonight and if they are, and my memory serves me correctly, will post a photo or two. There might be justifiable caution to buying from our Chinese friends.Quote:
Originally Posted by Clive S
the only reason I queried that one is because its via amazon so if there was any issues I can replace them via amazon for free under their warranty. I haven't gone through all of them so far as it is its still just a query.Quote:
Originally Posted by Clive S
Ah yes, I have a number of these, advertised as TB6600s but which are, in fact, TB67S109AFTG drivers.Quote:
Originally Posted by Doddy
Attachment 23772
Taking one apart...
Attachment 23773
No TB6600, but on the rear....
Attachment 23774
A tiny QFN package, pasted to a bit of a heatsink that is then coupled to a secondary heatsink.
After you've been caught once with these, spotting them is pretty easy. For one thing - they advertise a 32 micro step, which isn't supported by the TB6600 (only 16 micro steps).
There's some discussion online that supports these as a reasonable alternative to the TB6600s, but you need to read the data sheets and understand each on their own merits. Personally, I think the thermal solution on these is poor - you could end up with the chip shutting down under load. The R(on) is 20% higher than a 6600 (that's a bad thing that exacerbates the thermal solution). The max clock is 100kHz - doesn't sound bad unless your using 32 micro steps (then it's 100,000 / 32 microsteps / 200 steps/revolution stepper = 15.625 revs/sec, or 937.5RPM max out of the stepper (the 6600 is rated at 200kHz, and only supports 16 micro steps). Lower working voltage and lower drive current (4A vs 5A). Although rated upto 50V worryingly the 67S109 data sheet gives typical characteristics at 24V.
I bought them, and they didn't work well for me (missing steps), but my replacements (TB6600s) also had similar problems that needed the optoisolators updating. I'm not saying they're good, or bad, but different to that advertised.
That's NOT a BoB, it is a complete motion controller. Don't buy it. In my opinion buying drivers based on the Toshiba chip is a mistake. Get the DQ542MA drivers instead, they are miles better. I would also avoid switching PSU. Also, don't be afraid of eBay... there are also plenty sellers with UK (or other EU countries) warehouses with quick shipping. If you have any issues PayPal will give you a refund.Quote:
Originally Posted by Palletlad