remember no touching ground with those wires, and try not to touch the copper bits.....lol
Printable View
remember no touching ground with those wires, and try not to touch the copper bits.....lol
Ive taken that board out, not taking a chance with it until confirmed. Sorting home / datum switches now, a bit confusing. Home is X all the way to the left, Y all the way to the person but the datum plate is on the other side.Quote:
Originally Posted by i2i
Is datum not home?
Ok. clearly not although it uses switches called datum, that is what is confusing. The Triac usermanual confirms the same - home is correct as I have it now.
Can anyone explain why, when the CS Labs unit has a fault condition, my machine basically drives both X and Y to the positions furtherest from home? How do I stop this from happening?
Also, spindle relay wired, connected to output 6 on the CS Labs digital output but spindle doesnt run. I try M3 and M4, I can see Mach 3 flashing that there is activation, but the actual relay (24V out on output 6) doesnt kick in.
Actually, I think I know why in the first instance. I need to get enable to work via the controller. Currently the DER (Drive Enable Relay - which cuts power to the drives), is hardwired as its 0V 'on'. That said, why is something controlling it (ie, giving it reference to move), when there is a fault? Some type of failsafe?
And we have life, although the motor output is being set manually at the moment and still need to sort my machine runs X Y positive when controller resets.
http://www.mycncuk.com/attachment.ph...id=15709&stc=1
Ok, spindle now starts on M3 code. Progress.
Not happy with the E Stop setup. Im currently using most of the original wiring but on 'reset' my X and Y still moves in the + directions. I then wired off the 24V switched to tell the CS Labs that the Emergency Stop was hit. This works but resetting is tricky.
I am going to remove the E stop setup in the current wiring mode and just wire it to the controller. Ill work another way to enable the drives.
I am basically deciding same stuff for my machine.
So i wired Estop to the servo drives not to the board. The board just enables the drives. So I hit the Estop and the drives stop instantly, they emit alarm signal that disables the board and puts all in reset condition. I reset from mach3 and board resets drives and enables them/ 2 different pins that go to relays/
What i am saying is that is ok for the board to have second circuit connected to the estop button, but its a must to stop the servos, so thats why servos usually have Estop and limit switch inputs, to react faster than boards.
The way it should be done is to have the Estop connected to an emergency relay that will cut off the AC going to the Servo drivers. The Servo drivers should have a DC supply that still powers the logic up, so when the estop is hit, even though the AC is dropped, the driver still knows where it is.
You can also have the Limits connected to the relay as well as the csmio.
I was going to say I never had any major E-Stop issues with my Denford Cyclone lathe (from what I've seen, Denford used a pretty standard E-stop setup across most of their machines), but then the KFlop allowed me a lot more flexibility in handling E-stops, and I added extra relays so the servo fault outputs triggered an E-stop. However to achieve clearing a drive fault situation, I had to use a custom program in the KFlop that clears the E-stop signal, then waits for all the drives to exit fault status before resuming normal business.
I used Kinco servos from Zapp, which allowed me to configure inputs/outputs how I wanted them to work, but I'm not sure what drives you're using, or what timing sequence they need.
.
I'm guessing your problem is the Drive Enable/Drive Fault timing once you're into an e-stop?
Your DER should be connected to a Drive Enable output from Mach3, so unless Mach is active, nothing should move. Simplest option may be to add an override push button to disconnect the drive fault input and/or trigger the drive reset so you can fool Mach into exiting an e-stop, and get the drives back online/out of fault.
Yep, spot on. Mine is about DER timing I feel. As the Denford uses inverse logic, its been a bit more painful as the CS Labs unit cannot provide 'switch on this 0V option' from what I can see. I also dont understand why, when Mach3 goes into 'reset mode' that the machine moves itself in X and Y positive direction. I think disabling the drives will help here, currently the DER relay is kept powered (well, zeroed, 24V is permanent) via hardwiring. Ill swap the logic and then use one of the enable outputs, like what I am doing with the new Kinco drive that replaced a damaged IRT drive.Quote:
Originally Posted by m_c
Thoughts on this to drive Mach3?
http://www.pcworld.co.uk/gbuk/comput...ml#srcid=11026
Pros:-
Touchscreen
Fairly Cheap
Windows copy is legal
Cons:-
Not massively powerful, but its not a Pentium 4 either
Not a lot of upgrade options.
Just swap the relay wiring. Instead of having the relay having a permanent 24V with switched 0V, swap it so it's got a permanent 0V with the 24V coming via the CS.Quote:
Originally Posted by Chaz
That's the way analogue drives work. When the controller is active and running closed loop, it's continually adjusting the supplied voltage to hold the servo stationary, remove the closed loop or disconnect the control wire, and most servos will move, due to the internal circuit design and/or even interference/noise generate in the wiring. That's why the drive should be disabled when the controller isn't active. Most modern servo drives do let you program in a 'deadband' zone, which allows you to create a voltage range where the drive will not move, but it's not something that's needed when a CNC controller is involved. It's a feature more aimed at where servos are being manually controlled via something like a speed pot, where you need a stationery position from a basic voltage control.Quote:
also dont understand why, when Mach3 goes into 'reset mode' that the machine moves itself in X and Y positive direction. I think disabling the drives will help here, currently the DER relay is kept powered (well, zeroed, 24V is permanent) via hardwiring.
One thing to note with the Kinco drives (certainly the CD4x2 ones I used), is you can invert the input/output logic via the software. IIRC by default, things like fault outputs are only activated on a fault (bad choice from a failsafe point of view, as it means any fault in the fault circuit could cause the fault signal to go unnoticed), but can be swapped by clicking the suitable button on the input/output config screen (that's not the correct name, but I've not got the software installed to check just now!)Quote:
Ill swap the logic and then use one of the enable outputs, like what I am doing with the new Kinco drive that replaced a damaged IRT drive.
Capacitive or Resistive touchscreen?Quote:
Originally Posted by Chaz
Capacitive is useless in a workshop environment, as it doesn't like gloves, or oily/wet fingers (just try a touchscreen phone with wet/cold/oily fingers to get a good example of just how poor they are!)
Resistive isn't quite as accurate, but doesn't care what you use to touch it.
Thanks, agreed on swapping polarity, done that with the Spindle Relay already.Quote:
Originally Posted by m_c
On the Kinco, I cant get my Serial / USB cable to work, so its all been done by hand. Sometimes the drive doesnt like being reset and I need to fiddle a few times.
Oh, I dont know actually. I dont think the PC World lads will know either. Any way to know / check based on anything visible? In the same breath, if you said '99.9% of all screens are capacitive', then that answers it. Suppose Ill take my biking gloves and see if it works on a demo unit. If it does, it must be resistive.Quote:
Originally Posted by m_c
Slightly off topic.
Suggested vendors for:-
Alu, Brass - Clickmetal?
Decent Quality endmills / engraving bits / etc ?
T Nuts (I got some with the machine, but not enough) - 10mm.
Parallel bars etc (the items you put under material in a vice to ensure its flat)
General holding kit - ebay has loads, what is the quality like?
Thanks
The CS Labs unit actually has something called 'HV Enable'. Its for enabling drives, just wired it, works well. It will stop within 1ms if there is a fault or anything that it is not happy with.
What USB/Serial lead have you got?
There was an issue recently with FTDI clones, and FTDI releasing a windows driver that bricked them.
.
I'd guess the screen will be capacitive, but the quick check is to try working it with something like the blunt end/top of a pen.
.
Alu/brass, figure out what you need and do some searching. If you'll be buying full lengths/sheets, it's often worth trying your local metal suppliers. Personally, I use a mix of local for full size bits, metalsandplastics for cut alu/brass bars, and ebay for odds and ends for one off projects.
.
Endmills I mostly use APT (shop-apt.co.uk IIRC), as I know I'm getting decent cutters. Ebay is cheaper, but I've found quality can vary. Before you buy anything, run some figures through FS Wizard / G-Wizard (I personally prefer FS Wizard/HSM Advisor as HSMA I think is easier to use/configure) and try out different speeds/sizes/flutes and compare to what power/feedrates you have available. On a Triac, I'd guess you'll be wanting 2 flute to keep a decent chipload for the given RPM/feedrates, but run the figures and see.
.
Holding Kits, you can't really go wrong with the generic T-slot kits from ebay as a starting point. Once you get going and have some experience, you'll probably find you'll want to invest in other mounting methods depending on what you're doing. I'm just in the process (was, I'm currently waiting for a helicoil kit after I picked up a 4.8mm drill instead of the 4.2 to make a couple M5 threads) of making up a work plate that'll use Mitee bite clamps to hold flat bar in place, as I'm fed up with clamps being in the way and needing extra material to clamp.
.
For other bits, ArcEuroTrade is good. Not the cheapest, but they're stuff is always good quality, unlike some of the ebay offerings. They also dispatch quickly. Personally, just buy what you need when you need it. I've got stuff I bought years ago that I thought would be handy, but has never been used. On the other hand, it can be annoying not being able to get a quick job done because you've not got that couple pound widget that would make it so much easier!
http://plugable.com/drivers/prolificQuote:
Originally Posted by m_c
Supposed to be decent, many reviews. The lead is ultra short tho, not ideal.
So the coolant pump was seized. Managed to free it although cant get the 'bowl' thing at the end open to clean it. Ive tried both directions, no fluid coming out. Does anyone know if I need to prime the pump?
On a similar topic, what fluids do I use? The stuff in there looks old, ideally would like to replace it.
Provided the pump head is submerged, it should pump. Can you see if the rotor itself is turning?
.
Fluid depends on what you'll be machining. There's not a coolant that is ideal for everything, but there are some that are generally good on most things.
Ok. It's working now. Had to check direction a few times. Lots of dirt in the pipes etc. Pump not healthy but manages some pressure now.
Can anyone link a fluid I can use mainly for ALU?
So, not a trick question .... but would like some confirmation of what my X and Y steps should be. I have it very close but Id like to know how the number is calculated. I know the formula, but on the current info I have, the 725 value does not look correct.
The issue is that I dont know the config of the drives. I do not have access to them and therefore there could be a host of different 'steps per rev' measurement. could be a number of values.
So If I take some of these, I cannot match the value that I am currently using which is 725. This seems very accurate, I am getting measurements within 1% but I need to do more testing. I may have a bit of backlash which *might* account for some inaccuracy, certainly on my Y.
Please see the table below. I take a number of options and nothing matches 725. I then swap the gear ratio around and the closest I get is 819.2.
The third table (in green) I simply reverse engineer from 725 and get some random steps / rev, could either of these be correct?
The claimed maximum output from the driver manual is 1024, but am aware that this could be multiplied by 2 or 4?
Thoughts?
Thanks
http://www.mycncuk.com/attachment.ph...id=15714&stc=1http://www.mycncuk.com/attachment.ph...id=15715&stc=1
for a stepper fitted machine, that has 400 steps per rev, and the gear reduction is 2.5:1, with the 5mm pitch you get 200 steps per unit in metric. For a 2:1 gear reduction it should be 160 steps per unit. All you have to do now is equate that to the steps per rev your motor has.
steps per rev / ( ballscrew pitch/gear reduction ) = steps per unit
If you have a bit of backlash check the angular contact bearings at the pulley end of each ballscrew, they may need adjusting.
Ps. don't forget to fold the tab washer flat to adjust the nut.
Thanks, so based on my pulleys being 12 / 24 (motor end / ballscrew end), 5mm screws, using 725 steps for damn near 100% accuracy, what do you believe the steps per rev is on the current IRT 1306 drives via the simulated encoder outputs?Quote:
Originally Posted by i2i
I have a favour to ask, not sure if anyone can help or suggest someone that can.
I bought a replacement drive from a vendor. Before I bought the drive, I asked if it can work with my current CS Labs and Mach 3 setup. I was told that it would.
Once the drive was installed, it was apparent that it would not work, not unless additional electronics was added as it does not output a suitable differential encoder output.
I attempted to get the drive refunded. The seller rejected this. Eventually he agreed that the drive doesnt work as I need and offered a partial refund. This was then upped later but not 100% as I requested.
Distance selling rules allow me to return the device however as this was a seller from EU (Finland), there is a clause that states that I can return only if the device / item has not been used.
I ended up raising a dispute with Paypal and forwarded them all the email correspondence that highlights that the vendor has accepted that the drive cannot output differential encoder outputs as needed (without building some custom circuitry).
Paypal have now come back and asked that I make contact with 'an expert' who can verify this. This is their request.
"Please obtain a document from an unbiased third-party, such as a dealer, repair shop, appraiser, or another individual or organisation that's qualified in the area of the item in question (other than yourself), which details the extent of the damage or clearly explains how the item received significantly differs from the item that was advertised.
This document must be on a letterhead that includes the name, address, and phone number of the individual, business, or organisation so that we may contact them if necessary."
I spoke to them at length this morning as I dont think this is reasonable. Why should someone, who has no role in this, assist me (at their cost and time).
I told them that the emails confirm what the seller had indicated. I was then told that emails could be altered. They offered a new option which was even more complex.
So, my question is, is there anyone that is either on this site or falls under the category of 'unbiased third party etc' that is willing to look at the info and put something in a letter than I can send to Paypal for the refund?
If not, Ill reluctantly accept the part refund offer that I had and cut my losses on this.
Thanks
Give me a link to the driver that you ordered.
Also did you get an email form them to say that it will work with the CS labs controller? I find it strange that a servo driver that accepts +/-10V and works with a resolver would not have emulated encoder output.
Thanks.Quote:
Originally Posted by Gary
http://granitedevices.com/digital-servo-drive-argon
http://granitedevices.com/w/images/f...gonFlyerA4.pdf
http://granitedevices.com/wiki/ARGON_Servo_Drive
It does do some level of encoder output, but you have to add additional circuitry.
This is explained here:-
http://granitedevices.com/wiki/Argon_EXT_connector
Due to my lack of knowledge (at the time, not that I am now wise ;p), the drive cannot do differential encoder output. The seller of the drive (nice kit otherwise) indicated it can be done but not only do I need to solder into the drive to connect this EXT connector, I will need additional electronics. I believe (if I am not mistaken) that his method to create 'differential' was simply to invert the signals and present them for use. My understanding of differentials is that one of the reasons we use these is for error checking. If one of the feeds looks 'bad', it will compare it to the opposite one. Duplicating it is not exactly helpful as this wont work in this case.
At least, that is my understanding.
Any help appreciated.
Thanks
I am not sure if my lack of knowledge has been the issue here. In both cases I forwarded the seller both the IRT 1306 drive manual as well as the CS Labs manual to show exactly what I needed. I therefore assumed that the drive would work as I required.
My first email to them:-
"Hi,
I have had damage to my IRT 1306 drives. I have managed to repair 2 of them but the 3rd remains faulty.
The motor runs with revolvers.
http://shop.granitedevices.com/produ...vo-drive-argon
I believe that your driver should work. Do you have a single unit in stock? How quickly could you get this to me in the UK?
There are 2 resellers selling them but a 25-50% markup is not cool.
Please help.
Thanks"
Their response (removed their contact details):-
"Hi Charl,
Thank you for your inquiry!
We have few Argons kept in UK warehouse stock for urgent cases. Please let me know if the need is urgent and I can arrange you one.
Best Regards
Granite Devices Oy Finland, www.granitedevices.com"
My response to this:-
"Yes, please !!!
Can you just quickly confirm that this should work with my current system? It uses IRT 1306 drives with resolver input. That drive then outputs as encoder output which is then taken by the CS Labs controller. Ive attached the drive specsheet for info.
What do you need from me to process this and arrange postage for tomorrow or Saturday at worst?
Thanks
"
And their response again:-
"Hi Charl,
We can ship today with express courier if ordered ASAP and it would probably deliver tomorrow (but there is always uincertainty if it makes there in one day, or if theyre is saturday delivery option).
I quickly checked the pdf and the drive looks quite close to Argon. Do you require also encoder output from the drive? This feature in Argon has few limitations, see: http://granitedevices.com/wiki/Argon_EXT_connector
I hope to hear from you soon!"
Obviously I have missed something from the thread. What CS labs board are you using and why you close the loop in the board and not in the drives? Why would you need at all the drive to output encoder signal to the board?
Im using the CS Labs IP-A.Quote:
Originally Posted by Boyan Silyavski
My original drives are IRT 1306. They take motor resolver feedback to work and then output an emulated differential encoder output which the CS Labs unit needs to know where the motor is in terms of position.
Looking at his email, it looks like he left it up to you to read through the specification sheet to see if it will work or not.
Looking at the details on the encoder output, this is very poor and not even a connector accessible form the outside of the box, so will require going into the driver and connecting to the ABZ, but this is only 3.3V and also very unacceptable to esd and electrical noise.
If is was just a single ended encoder output with a good 5V signal and a connector on the outside of the box, then this would not be a problem, by connecting a standard board to convert from single ended to differential.
If you want me to put something on paper i can do.
It makes absolutely no sense to have a good driver like this being limited by no differential encoder output.
Thanks. Agreed. I like the drive, its nice kit. I love the open source idea but its not what I needed.Quote:
Originally Posted by Gary
It feels like I bought a car with an Aircon and then when it doesn't work, I get told to have some pipes and gas installed to make it work.
Appreciate if you can draft something up for me and sent it across.
Many thanks.
emailed you back.
Thanks, appreciated.Quote:
Originally Posted by Gary
Quick update time.
Machine is running but I need to do more tweaking. Need to recheck the PID control but generally fairly happy. Taken most of the backlash out both mechanically and with Mach 3.
Needed to move the mill but no method available. The engine hoist I bought isnt wide enough. Got an idea to buy some heavy duty castor wheels. Tried to fit them but the metal I wanted to put them on was too flimsy. Came up with a good plan, had some 20mm Alu plate for something else but milled myself 4 plates to basically mate the wheels to the load bearing beams, using the original shafts.
Some pics below, the CNC has now been moved to the location I wanted it in the garage. Last pic is of a new X plate but I ended up having a problem with it, so it wont be used.
http://www.mycncuk.com/attachment.ph...id=15860&stc=1http://www.mycncuk.com/attachment.ph...id=15861&stc=1http://www.mycncuk.com/attachment.ph...id=15862&stc=1http://www.mycncuk.com/attachment.ph...id=15863&stc=1
I wanted to thank a few people that have made this conversion possible. Its not 100% complete but mostly there. Still need to get ATC working and build a new control panel.
JazzCNC
i2i
Gary from Zapp (please support him)
There have been others too but I would not have completed this conversion without their help.
Many thanks.
Can you post some pics of the ATC parts. I have a (I think) similiar machine with Pickup ATC and know what you need for this part. IIRC aou want to buy 110V solenoids for the pneumatic harness. I would better use only 24V, you can them
connect directly to the PLC output (if the amperage of the Output is enough)