Printable View
Hi All
I'm in a bit of a tight spot currently, and really need some help.
I have a Mach3 converted Novamill, which I've only recently had running. With just a few parts left to make, the HDD decided to die on me!
It never occured to me to make a backup of the xml file, so I lost all my settings.
If anyone has a Novamill with Mach3 controlling it, and would be happy to share their xml file, or settings, I'd be incredibly grateful.
Brian
The configuration file will be specific to whatever controller/breakout board is fitted.
Can you upload some photos of the inside of the control cabinet?
Apologies. I've been posting all over the place and thought I'd already put the images up
Attachment 32140
That's a pretty generic Chinese BOB, however I'm not entirely sure of the pinouts.
I do however know the original Denford wire numbers, but they don't seem to match the hand written marks on the terminals -
57 - Home 2 (Z)
58 (assuming the 5 has fallen off) - EStop monitor signal
66
67
68 - Estop relay coil
78 - SGR (Spindle Go Relay)
66 & 67 would normally be the wiring numbers for part of the E-Stop loop (depending on the spec, the EStop loop can go through EStop button(s), limit switches, Guard interlock, and spindle motor overheat, with the wire number incrementing between each switch/sensor capable of breaking the loop).
This is the BOB, which will give you the pin mappings
Attachment 32145
I'm going to open the box up and see if I can trace some of those wires/numbers
From the BOB pinouts and looking at the wiring I have set the following:
66 - X Home
67 - Y Home
68 - Z Home ( however this one doesn't show in the diagnostics panel as being made at either end of the Z axis travel)
57 - E Stop
58 - Ground
Z Home should be near the top of the travel.
I can't remember how far off the top (I sold my Novamill a good few years ago), but it's definitely near the top.
I'm guessing homing was working OK before the PC died?
Homing was non existent before the HDD died.
I could drive right up to, and past the switches on all axes.
The only way I was able to work was with soft limits, but if there were any problems with references I would lose my absolute position and have to start again.
Right now I can see LED's for X and Y light up at M1 Home and M2 Home but not for the Z axis
I can probably guess why homing and limits don't work then.
Denford for a while used inductive NAMUR sensors for homing, which are a sensor type designed for volatile atmospheres (basically a two wire sensor that changes resistance/current flow, so any wiring faults don't result in sparks). These are not easy to interface, but the original controllers used them, so I'm guessing whoever done the retrofit probably never changed them.
Instead they've probably modified the limit switch wiring to use them as home switches, as the limit switches would have originally broken the E-Stop circuit, but then they've never really implemented them. Or possibly disabled them because the Z has gone faulty.
Before you dismantle anything major, if the machine still has the big Harting multi-connector for connecting the machine to the control box, unplug that and very carefully look to make sure all the pins and sockets are in position and the same height. I've seen them become unseated, so they don't make contact.
The homing situation makes a lot of sense now, from what you've said. Can the physical switches be used for homing or is that impractical?
There's no reason they can't, if they've been re-wired.
For the size/power of a Novamill, you can probably get away with using the limit switches just for homing, then relying on soft limits.
It's not like a Novamill axes have enough power to do any damage if you run into the end of the physical travel.
I noticed they aren't the strongest when they just keep going if I hit the hard end stop.
Looks like I have some fixing to do. In the meantime I've managed to get all the axes to jog and found the pins/port combination to get the spindle to run, but I have no control over it.
Keep getting the PWM bumped to minimum warning even when I enter big numbers.
Any clues on how I get control? The screen overrides won't work but I can start it from there or the F5 key.
It's been a long time since I've used Mach3, so can't really help you there, but it'll depend on exactly how the speed control is being done. You'll probably have to trace the wiring to establish how it's wired in, as going by your photos, it doesn't look like it's using the BOB's 0-10V output.
Thanks for your help. Seems like I've hit a wall at this point
Think I'll just have to leave it now as I've pretty much run out of time. Back to knife and fork to make my final components
B
Are you using the original Sprint speed controller? If so do NOT try connecting the BoB 0-10v output to it as the controller is isolated from ground and its 0v floats at something round half mains voltage. An isolation circuit is needed. I've got a Novamill with a slightly different Bob but happy to share xml when I'm back home. Also can possibly help on the isolation.
Thanks John
I took some more photos earlier which might show what's what. Spindle was running fine before the dead HDD, so it my just be a case of the right pins/outputs.
Like I said before, the spindle will start with the pins I've picked but there's no control like before. I ned to take a closer look after m_c pointed out the lack of control connections to the 0-10v outputs on the BOB
If you think your XML might work I'd be very happy to give it a try
Thank you
The machine has been running, so I'm guessing there must already be some form of isolation.Quote:
Originally Posted by JohnHaine
Going by the photos, I'm guessing P17 is being used to control the SGR (Spindle Go Relay aka switches on power to the Sprint drive), and then the little red wire next to it from the P1-PWM terminal is being used to control the speed.
However to use that, I'm guessing there is another PCB mounted somewhere, because as far as I can remember, the Sprint drive only accepts a 0-10V supply, which must be isolated. Unless it happens to have an isolated Sprint drive, but AFAIK Denford didn't use isolated drives.
Just to be clear what we mean, due to the design, SCR motor controllers (which the Sprint is), the speed control voltage sits at roughly half mains voltage, which is why it needs to be treated with caution. You do risk giving yourself a tingle if you touch those pins/wiring with it powered, but the equipment risk is by connecting them to a non-isolated circuit, magic smoke will be released.
Can we get a full photo of the Sprint drive (lower right corner of your photo)?
And where does that little red wire go to?
The little red wire goes here:Attachment 32146
Close up of the Sprint drive Attachment 32147
Attachment 32148
Attachment 32149
Looks like they've patched the PWM signal into the original interface board, in which case you just need to work out the PWM settings to control the spindle speed.
You'd need to confirm, but if you wanted, you might be able to connect the Sprint direct to the BOB, as I think the 0-10V circuit on the BOB is isolated from the 5V control circuitry.
You could confirm by removing the 12/24 wires to the BOB and checking everything else still works, and also checking that the Gnd for the 12/24VDC supply isn't connected to the other BOB Gnds, and only connects to the 0-10V Gnd. If that's the case, then you'd just need to add an isolating DCDC converter to get an isolated 12VDC supply to provide the required isolation.
Given that it was working well before the drive died, I'm reluctant to make any big changes right now.
If all I need are the right PWM settings I stand a chance of being able to finish my work in time.
Just a case of figuring out what they are
The PWM is connected to pin 1 on the parallel port, so it shouldn't be too hard to figure out.
Having just installed one of those BoBs to replace the old one on my lathe, I can say with confidence that the 12v and 5v grounds are commoned! Therefore the 0-10v output is not ground isolated. Some years back I published on here a simple circuit to take the PWM output, isolate it with an opto-isolator, and smooth it to generate the 0-10V needed by Sprint type drives. A bit of searching should bring up the post.
I did actually see that post and concluded that it was way beyond my paygrade currently.
Whatever the actual ins and outs of wiring/connections could or should be, the thing was working fine somehow.
I just need to get back to that somehow, then plan to update or upgrade the internals.
Another point about these BoBs (for which documentation is very scarce) is that the logic inputs have an opto isolator with a 1k resistor connected to the +ve supply, the 12V one not 5V, and expect to see an "open collector" drive such as a microswitch pulling to Gnd or an open-collector output ditto. Caught me out and I had to convert a couple of sensors that generated a 5v logic signal.
Can you suggest, based on what you've seen, which pins etc I might try to get the spindle to work?
Attachment 32150
I attach a Sprint manual which may not be for your precise type but they seem fairly generic. For a quick fix, you could either
(a) connect the speed control input on the drive direct to the +10v output they provide - that's pin 1 to pin 3. Make sure ping 2 isn't connected to anything. This will simply tell the drive to run at maximum speed when switched on.
or -
(b) wire in a 10k manual speed control pot, with the clockwise end to +10v and the other end to COM which is pin 5. Again pin 2 not connected. Make sure you have a pot with a plastic shaft, mounted so you can't touch the metal casing or screwed bush.
I can't vouch that the exact same pin numbers will be found on your drive.
Nothing internally has changed with the wiring, so some combination of pins/outputs and values should surely do the trick?
I'm not in a position, time or mind wise to tackle rearranging the spindle functions and it's controller right now.
I'm going to give up the fight and get someone else to make my parts before I go completely crazy with frustration and chuck the damned thing out of the workshop
Thank you for your help. Both you and m_c have been brilliant in your efforts to educate me.
B
So a long awaited update on my Novamill situation.
The day after my last post I took a call from Brian at CNC4U, where he made some suggestions to get me running again.
The one thing that helped most was to use a higher quality cable from the UC100 controller-pc. Seems there is a lot of stray EMF and cheap cables don't have enough shielding.
Combined with the help I've had here I was able to finish the project I was working on.
Once again my thanks to everyone on this forum.
That leaves me with just one issue; the limit/home switches don't work unless virtual limits are set. If I lose my initial datums for some reason (light smashing into the part) I have to start again because I have no references.
Activation lights in Mach3 show the X/Y switches working (no Z though) but when I reference the table just runs over the switches until it hits a hard stop and spins the leadscrews.
It was mentioned in a previous post that Novamill switches are inductive, so the question is; do I try to replace them with mechanical switches, or can they be made to work as they are?
My workaround idea is to add a pin or hole on the bed somewhere are a permanent reference. That way I can get back to datum if things go wrong.
Any advice would be welcome. Thank you.
Novamill switches, on my mill anyway, are standard microswitches. Maybe there's a wire dropped off the z switch? If you are seeing the lights operate on the diag screen then the X and Y switches are working.
There's a configuration setting for the homing speed as a percentage of rapid, what do you have that set to? When the home switch is triggered the mill goes into a controlled deceleration to stop then reverse until the home switch deactivates so there is a small overtravel. If going too fast this may allow the axis to hit the end stop. IIRC I have the homing percentage set to 5% which is probably too slow but I never hit the end stops.
On another point, I almost never home the machine anyway unless I'm doing a job that is very close to the limits of travel. I always reference the axes to the workpiece using an electronic edge finder and tool height setter. If I do have a crash I only need to re-reference to the work again.
Update: the homing speed setting is in the config\homing/limits tab
Attachment 32292 Far right column.
Good advice as expected. Many thanksQuote:
Originally Posted by JohnHaine
I'll check the outputs and adjust the speed down to a lower value. It can't hurt to give it a try.
Reassuring to be reminded that every setup has it's own reference points, and if they are noted they can be attained after any "issues"
For anyone coming across this thread, I'd like to say this group has been invaluable with it's help and suggestions.
I can't imagine how many total years of experience is here amongst the members, but being able to tap into it is just fantastic!