Machsupport Forum
Third party software and hardware support forums. => SmoothStepper USB => Topic started by: rally.sport on June 09, 2009, 09:29:47 AM
-
hi all
read a few long posts on here and i have similar issues to others
installed a smoothstepper to have greater control over a pacer 2500hd i own that we converted to mach3 about 4 years ago
now the ss is in a small abs box with a 8 relay board next to it
this box is in the main control unit metal box that contains the 3 msd stepper drives, 2 power units and a phase converter for the spindle speed control.
all cables are shielded except for the 2 to 4 iches exposed inside the control box. no issues prior to fitting ss using all the same parts.
i have the same issue of mach freezing every 3 to 5 mins.
if i close mach the ss actually stays locked on and holds the relays switched on, only way to switch it off is to pull the usb cable
i have tried the following:
if i only connect 2 stepper cards in any combonation it works for 15+mins (as long as i left it in my test) connect the 3rd and it trips out in 3 to 5mins
if i connect any 2 cards and switch on any 2 relays it trips out, so i thought maybe a load issue so i wired out the relay board (relay 1 switches on machine relay 2 spindle)
but it still does the same if i issue the same commands ie turn on 2 relays but they are not wired so feed is to pins in plug only doesnt mater if the spindle is on or not, if i turn the machine off i loose the stepper driverr feed so icant do that.
disconnecting the wires to the stepper cards at card end elimates the problem of the freeze in any state that i have tried (the steppers are msd units and have there own opto isolation at the card end.
is this another noise problem, if so is this a coman problem to the ss system.
i can only be getting noise from the opto isolated stepper drivers which do not have a cable run near any other wiring, or the ss acting as an aeriel or the good quality 3m usb gold plated cable.
or i could hhave a port/pin problem but i cant see that as its the same settings
any ideas would be good.
-
i also see that one guywith same probs put his ss in side the pc and ran out 2 parralel cable for control
did this work for him?
and if so why for him but others do not have the same issue
thanks
-
Basically you trapped your SmoothStepper in a box with EVERY noise generating component in your system. The SS circuit board and USB cable can pick up any electrical noise that all those other parts are generating.
You need to protect your SS from electrical noise. The first step in that is to make sure that all your equipment is properly (earth)grounded. The next step is to make sure that all I/O on the SS is optically isolated. Then make sure that your signal wires (limit switches, etc) are not running parallel to your noisy stepper wires. These first steps ally to any system and are not unique to a SS.
Now for the SS: make sure your USB cable is as short as possible and not running along with your stepper wires. Mount your SS outside of the cabinet that all the noisy equipment is in or move it to the PC as has been suggested.
-
ok thanks for those suggestions
i have removed the ss and refitted a breakout board on the end of a parralell cable and all issues have gone away
as for putting the ss in the computer the reason for buying the ss was to avoid 2 par cables and to increase what we could do onboard. (and for laptop control in the future)
i have to say it seems a common fault reading other posts and i am disapointed in the product if it has this flaw, surly is should be either protected during the manufacturing process and have io opto's as standard or come with at least a warning or instructions to say "fit it in its own box to prevent noise and use i/o opto's" then i think iwould have not bought one, its sold as a complete solution and an upgrade from 1 or 2 p/ports
personally, as a manufacture myself, i consider any product cable of hanging a computer and producing potentially a dangerous uncontrolled machine as not fit for purpose, no matter how badly it may be considered it is installed, the machine we own is just 6 years old and all i have done is to remove the oem (with onboard relays)control from the oem box and replace it with a ss and relay board.
i will try to earth systems seperatly but as the fault occurs on the stepper lines (with all else diconnected)which are opto isolated at the driver end anyway and use a good quality industrial machine cable back to the ss i think we will bin it as i dont want to have a system that may fault at some random time in the future.
its not just us thats experianced this problem perhaps warp9 might like to comment and look to develop this issues out, seems from the post ive read that both art and warp 9 have said umm dont know.... to it. maybe they have seen more problems and can comment further?
thansk
-
Could the SS perhaps have some additional HW noise filtering that would help? Probably yes. Would it solve all these problems we here about? No.
The first lines of defense when building a machine are proper (earth)grounding and proper isolation (optical or otherwise). Most machine control I/O boards will NOT have optical isolation built in as some of the external boards that are driven (like Geckos) also have optical isolation on their inputs. It is best not to 'stack' them. For example, I refitted an old Bridgeport with a Galil motion control card, the card, the extended I/O board and the breakout board from Galil cost about $2500 and they were not optically isolated. They did offer an isolated version of the breakout board though. I used solid state relays instead, as two Opto22 boards fit in the same place as the old main I/O board did. The solid state relays not only gave me the optical isolation, they also allowed me to switch different voltages directly, both AC and DC.
The problem comes about when folks, who don't know any better, build a machine and don't quite get the wiring neat and tidy and also don't get the grounding done properly. Through in the lack of any form of isolation on most if not all of the I/O and you have a problem. You thorw a SS into the mix (or any USB device) and the USB cable will pick up all that electrical noise causing the main hub in the PC to stop trying to communicate as it says, heck there is a lot of noise here, I can't even hear myself think!". The problem in the noise.
Don't take that the wrong way. I'm not picking on the OP or anyone else. We all start out at the same place, and learn as we go. But, we need to learn what the source of the problem is and improve out own machine building practices too.
-
so do i understand it correct the usb hub/line is actually more prone to noise? and not so much the ss picking it up or developing a fault?
the machine we have is a £50k machine its now 7 years old lives in its own cage for health and safty that has its own pressuiresed air filtration system as well as dust collection/recurculation, (the materials we mill are considered unhealthy!)we run a pc with 3cctv camera's on it to monitor the machine without entering the room.
the control pc sits in an operators booth, there is a 2 inch hole (rubber grommet) through which the usb and also a manual overide and estop control panel cable runs it is only since going to ss and usb weve had any issues and by simply taking out the ss and replacing the usb with a parellel cable and breakout (no isloation no additional earthing) all is well.
as a test i will try the ss in our room with a long parellel cable and see if that changes things.
as for the opto isolation i agree you shouldnt stack them the relay board is shmitt triggered, the drives are opto'd at there end
we have no inputs connected so no noise there, one point i did pick up on which we might have is our stepper drives have the option of pulsing plus 5 and common neg or fixed five and pulse neg, we have them wired as pulse 5, our negs are connected together and back to a common (pin 20 i think) we also have a ground to pin 18 from the relay board if i read another post correctly this can cause problems? although the post did not say how to wire this problem out. i could reverse this and take plus5 from pin 1 on enable and then switch the neg line, but will that make any difference?(takes out the pin 18/20 common and earth possible cross over)
so for good earthing,
we have an earth from the ss to the metal outer cab, we also have one from the relay board, all other components are as wired by the manufacture and earthed. the machine is 240v and earthed to the mains correctly. there is no link between pc and control except the usb cable.
do we need to run and extra earth along side the usb to link these 2?
thanks for the reply and dont worry its not taken as any offence help is good from any angle but we are fairly experiancnd here and the install is to a high standard just think of the room the machines in to see hte level we had to go to to get iso approvall
-
oh and i forgot my manners!
hi geoff
thanks
james
-
jeff even
i work with a geoff engage brain before fingers!
-
No worries I answer to either spelling. Probably just moving the SS out of the drive box, or shielding it would work. I agree that a bit more of an 'installation' guide would go a long way to preventing headaches.
I've been designing a new 4-axis drive box that uses a SS. It is an aluminum enclosure with the PS and other high voltage stuff mounted on one level, then there is an aluminum 'attic' floor that the SS is mounted to. This shields it quite nicely.
-
so would putting it in a metal box instaed of a abs one make a difference? or is it a case of it might help?
thanks
-
Yes, if the box is (earth)grounded it will prevent the SS from picking up the RFI from the drivers and such in the same box. If the SS is picking up noise from the inputs you'll need to optically isolate those as well.
-
A couple of observations based on my SS. The USB cable I originally tried had no continuity between the metal shells of the connectors - which suggests there was no shielding. Also the onboard USB 1 ports in the PC didn't seem to work reliably at all.
The only way I could get stable results was to use a shielded 5m USB extension cable which has a hub integrated into the end. I now have that coiled up inside a metal enclosure, with a "normal" usb cable coiled up in the same box connected to the SS.
Aside from my x or y home inputs getting stuck on randomly - it seems to be performing as expected i.e I actually cut some parts with it and the machine is definitely running smoother.
Having said all that - the main source of interference I have is from the VFD/spindle, which, without a complete disassemble and rewire, is going to be tricky to resolve.
-
we tried a few more things today, but i have to say its not a reliable product suitable for comercial use.
ive had a reply from greg today and in his own words"I think that a ground loop exists from the PC through the USB
cable," no mate you design and sell this product you should know not think, a cnc machine with a spindle turning with a cutting tool in it can injure or kill a person
ive put mine in the bin its a liability i employ people i cant have a situation of "the only way i could get stability apart from!" my insurance company would have me in prison if it injured someone especially with the knowledge i have of around 50 percent of users having problems, and all seem to be with commercial machines which are better built to start with but are expected to return better results.
-
ive put mine in the bin its a liability
ill buy it from you if the price is right :)
Hood
-
rally.sport,
Putting the SS in a solid metal box would be shield and that shield would be considered 100%. If it were shielded
in say a fine metal mesh screen ( there are rough equations or quides on how open the mesh should be depending on what you want to do ) then that also may be appropriate ( but not as good ). You could also just put Al tape, overlap the seams, ( like the tape used to seal duct seams ) around the whole thing and that would shield it.
Now if the noise is being carried to it via wiring inside the the shield then your shield won't be worth much in that sense.
Define the source of noise / rf, contain or minimise it, shield as best you can, move the SS to it's owne or appropriate location ( like in the computer case ), Use the same ground for the 5v as the gound of the supplied 5v.
Sometimes there is logic to it all ,sometimes i just call it black magic ( because it worked and nobody can realy say why!
Just some thought's, no one stop solution for you,
RICH