Printable View
So here I can define Input Parameters. P23_15 was set to 13 (Overheating) but I set it to 127 (Set to Null). The problem is that in the list below, its only P23_10 to P23_14, Ill double check but I dont believe any of them were set to 13 (Overheating monitor).
Attachment 30615
Interesting that in the next section down it says 'add 128 to the number to make it NC type'.
The only other page I can see something related to motor temps is below.
My 'reading' of it is as follows ... not sure if that's correct or not.
P90_03 if its set to 0, it will Error and power off. if its set to 1, it will error but keep power on and 3 or higher it will 'no error, assuming it ignores' but offers limited torque %
Attachment 30616
So SZGH have confirmed that its a NC contact.
The mapping goes to P23_25 (port MapToDi15). Ive been told to change this to 127 and disable it, I dont think you would know what port it was mapped to from reading the docs. Ill try that later and see.
Parameter changed, it allows running but now comes up with Error 5, encoder failure. Its the same cable that is used for the Thermistor and a premade cable - cant help but think there is something wrong with this cable. Let's see what SZGH come back with.
If it was a simple DB connector, id put something together but this goes from a specific canon plug into a RJ45 (which in fairness looks good), so will test / check later. Might be a tuning issue - there's also some parameters that might need changing based on the manual.
So parameter 14_x has a number of options. One of them is to disable the encoder or suggest that there isnt one there.
There is also encoder feedback on the spindle itself back to the main controller (not wired in yet), so this might be enough for now, can sort things like rigid tapping etc later.
Ok, disabling the encoder stops the errors but the motor doesnt run well and doesnt speed up. It sounds like its single phasing but oddly when you press the spindle off button, it runs smoothly and then comes to a stop.
I've tried most of the encoder options, all of them give the same result, error after a few seconds.
There is a self test you can do which is is to set it into mode 2 which does a phase test (presume this is to check that the encoder and motor direction match). I had this working a few times whilst I was stuck with this motor temp warning but since then I cant get the test to repeat.
So will keep scratching my head, slow steady progress hopefully.
I was writing some more than we had a power cut.
There is a viewable parameter called 'H' which shows the absolute position of the shaft. This isnt reading what I would expect. It flickers between, 1, 3, then 4096 then back. It's like it is on the index but ultimately then it doesnt seem to move when the shaft rotates.
I'm not sure if any of this could be due to the way that I am powering this (via a VFD via a choke) however when it stops running (after the stop signal) its ultra smooth.
Will keep looking.
Spindle drives are essentially a Servo drive.
They need to know the orientation of the motor in relation to the encoder, so they know exactly how to power the motor, which is done orientation/phase sensing to learn the position of the rotor in relation to the encoder.
Typically the drive needs to know the encoder count, then go through the learning process to learn the motor.
Not doing this successfully results in what you're experiencing.
Unfortunately, that's as far as my knowledge on the subject goes, as I've never actually used an unmatched motor/drive.
OK, so major progress but sadly no go.
I checked the connector and rung out the wires from the cannon plug back to the RJ45 connector (but has 10 across, not 8). All looks OK.
I then found something that explains the issues, a loose wire inside the motor. Joy of joys, would explain why the encoder doesnt work properly and why the motor heat error is needing to be masked.
So soldered it back together, sadly no change.
I did figure out how to do the the Op = 2 run again, its all in the sequence. It trips out on an encoder error. I dont remember if it did this before or not as I was getting the over temp error but I think it did the same.
Its as if the encoder is simply not working. I can see no indication of a pulse or movement.
Attachment 30618
Attachment 30619
The manual refers to checking system parameters E04 but I dont know where / what that is. It does then talk about Parameter 14_1 onwards which are for the encoder, so must be related.
Ive tried to remove the encoder to see if all wires are there but struggling to get the rear case open.
So you get a bit of understanding of what the drive needs to achieve, have a read of the 3 Phase Brushless Motor Control on this page - https://www.dynomotion.com/wiki/index.php/SnapAmp_Info
It is Dynomotion specific, but it explains what the drive needs to achieve.
You just then need to figure out how to do it with your drive.
Having a quick read of the manual, E02 for motor poles, E04 to set the encoder lines.
The index offset might need a bit tweaking for smooth running, but I'd expect everything else to be automatically calculated after the phase detection has run.
Thanks. Ye the newer manual uses E codes, the older (which applies to mine) uses P codes.
Ive spotted that the unit might need external 24V. I dont think this is the case with mine as it has a connector going back to the SZGH control panel where it has 24V available. That said, Ive asked them to see. If there is no 24V that might explain it too but not sure where I apply it as there are no power connectors apart from 2 spare solder points (3 - one is not tinned) in the cannon jack. That wire (the untinned one) does not have a partner on the plug, so I'm wondering if that might be where 24V is inserted. Doesnt sound practical, who knows ....
Do you have an oscilloscope?
That would be a sure fire way to find out if the encoder is working or not.
Failing that, you could test using a multimeter connected across the encoder pairs (A+/-, then B+/-, then make sure there is at least power on the Z+/- as stopping on the Z pulse will be near impossible) and randomly positioning the motor to see if the voltages invert. That at least tells you the encoder is doing something.
And if your multimeter has frequency measuring capability, you could spin the motor and see if if you get a plausible frequency reading.
I dont sadly, but Ill measure with what I have. I know you can get cheap scopes for basic testing, maybe that's a plan.Quote:
Originally Posted by m_c
Nothing like spotting something, heading to the garage after 10pm, trying something, it fails, but then you see the cable that you fixed earlier still disconnected. You connect it up, you rerun the change / test and it all works.
https://youtu.be/NPxrjj5FobU
So here's a clue as to what the issue was (clearly the broken wire was also an issue).
Attachment 30621
Attachment 30622
I'm sorry, but I've got to say it.
Idiot!
But it just proves, always check the basics.
So the cable marked with X5 DOES NOT GO into the X5P.
It goes into X2A.
I spotted in the manual that X5P = encoder OUTPUT, so I investigated further. X2A it is. I'm not sure why there is an X5 number on the cable.
The other X5 cable (X51) is also marked X5 and goes there ....
Anyways.
In fairness, there was a wire off as well, so more than 1 problem it seems.
So, progress. Started to rewire and tidy up the cabinet.
First test cut (well, 2nd), - ignore the birds nesting, bad feeds / speeds.
I need to work out why the spindle override is being ignored too. Speeds are being set by G50 S2000 etc. Need to see the details behind this, I'm using the Fanuc Post Processor.
https://www.youtube.com/watch?v=p70y6Z6F8ok
And then a pic, I missed facing it fully but you get the idea, really happy with this.
Attachment 30638
Have I got my wires crossed.? I thought you bought the Spindle motor from SZGH same as I did. If so the drive is just a large servo drive so why are you using a VFD.?
I have only single phase. So I am making 3 Phase with a VFD, through a choke which then feeds the spindle drive.Quote:
Originally Posted by JAZZCNC
Ok, I get it now.Quote:
Originally Posted by Chaz
If it's any consolation I did the same with the X5 cable, the markings are clear as Mud.!
Ye, generally its decent in terms of cabling and quality. One broken wire and the assumption about the labelling.Quote:
Originally Posted by JAZZCNC
I adjusted the shape a bit and redid the CAM. I did trip the supplying VFD so had to slow down the ramp on the spindle which seems to have helped.
https://www.youtube.com/watch?v=ZgA6SV2jczc
Hi Chaz
I am really interested in understanding the kit you have used to upgrade these really good but slow and unreliable machines.
I have had about 5 CHNCs over the years, cannibalising the old AB controls to keep just one running.
I know have one in the graveyard and one that is a really good condition machine with a dead AB control that I really have given up on.
I have been looking to buy the Centroid upgrade kit but it is so difficult thinking about spending such a lot of money on what is a very old machine.
The problem I have is that I like the precision and the 5c collets and the special collects I have developed for various jobs.
I can't justify buying a new CNC (I have a 1988 Mazak QT15N for the bigger, faster work).
I don't run my machines on a 24/7 basis - they owe me nothing and I just use them when I need to make various components for my products we sell.
So seeing that someone so recently is doing exactly what I would love to do has prompted me to join the forum and ask if you can share your experience so that I might have another look at doing the same to save my old but loved CHNC.
Look forward to hearing from you,
Paul
Hi PaulQuote:
Originally Posted by Fairmount
Indeed, good machines once you can get rid of the old controllers (like many things). A lot of detail to consider, should I just PM you my number? Easier to talk through however in summary, I'd probably do the same again if I needed to. Once you understand these 'not pc controlled' controllers, they are fairly straight forward.
A key question however is then if you are replacing everything such as motors and how much of the original interlocking do you need (for example, I can have door alarms etc) however if I want to have a condition such as the spindle lock being detected to stop the spindle running, that needs some coding / considering.
Charl
Hi ChazQuote:
Originally Posted by Chaz
Yes please - I would like to chat this through.
01763263415
[email protected]
Just tried to call, was told you arent in. Ill email you now with my mobile number.
Interesting thread.
Yesterday I visited a customer, and expect to refit 1, later 2, larger (2.5 m long carriage) lathes with industrial quality retrofits.
The budget is likely to be around 15k, 18k with glass scales and or better screws.
Probably cslabs csmio-ips, and their sw, or maybe machx.
A 3 m long, 40 mm D screw, ground, c1-c3-c5 is not cheap.
Around 1 kW servos, x, y, maybe c (like mine), maybe B (tailstock).
The client is ready and has the will and the moolah, so ...
Itīs possible, perhaps likely, I can publish the refit.
The client is urgently looking for a big heavy large lathe we can do this on.
First xyc(b), and soonish linear guides (35 mm+) on all axes.
Very Large Lathes tend to be cheap, and are good moneymakers.
When no-one else nearby can do 50-60 cm diameter pieces, or shafts 3.5 m long.
Personally (and this is just my view), I'd not touch another CS Labs setup. I still have it on my one mill, its fussy, Ive had issues seeing the controller, firmware bugs etc.
Hello and sorry to revive an old thread, but I have a question about the original decision to replace the spindle motor.
Did you decide to replace because it was 3ph and you wanted to be able to operate from 1ph, or was there a problem being able to operated and control the original spindle drive from the new control?
I have a somewhat newer 1985 CHNC II with a GN10 control I would like to retrofit, but would rather leave the original spindle motor and drive if possible.
Thanks Gary Breiling
Hi Gary,
In my case, probably 2 factors.
The first was I had no DC controller for the motor, at least, not what I could see in the 2 machines I bought. There is meant to be a controller / drive in the 2nd machine's cabinet however the way the machine is placed means I cant open the cabinet and I cannot move / access the machine (long story).
The second was due to single phase only, when I ordered I had assumed that it would be a single phase drive, unfortunately some misunderstanding and it turned out to be 3 phase but I do run it from another VFD and it works, albeit that Ive needed to slow the ramp down time else I trip the feeding drive.
Thanks
Thanks Chaz, I need to do some more study but I have two complete machines and the spindle runs and operates on both now, I assume without knowing for certain that the control probably just provides a run signal to the original spindle drive probably like a 0-10 commonly used on a VFD. Not sure I understand how the encoder plays into that for the lathe to know the spindle location, but it’s good to know that the original spindle motor and controls might be an option. That seems like it could be one of the bigger costs. Thanks Gary
Just one more question, if you were doing this again would you stay with the same control and components from SZGH or would you use something different?
Thanks again Gary
Yep, would be good to validate, I would assume this but would need to be checked.Quote:
Originally Posted by glbreil
Interesting questions. The SZGH controller eliminates the PC stability / compatibility issue. The downside is that it feels like you are missing some flexibility and configuration options. It's all still there but not as easy / pretty. I think from a stability point of view, its great.Quote:
Originally Posted by glbreil
The only other controller Id consider at this stage is using Centroid as I have used on other machines (Mill and Lathes before). Each has its own benefits but Id never run Mach and/or CS Labs again personally.
Thanks, I have what has become an older CNC mill 1997 that was built new with a Centroid M400 control. It was my first CNC anything and I found it easy to learn and operate and the few times I have had trouble I have been able to figure out the issues and repair myself, so I definitely have been looking at that as an option.
Thanks Again Gary
Yep, Centroid is nice and mature and the conversational options are good too.
So after doing some research it looks like I would have to replace the axis motors to go with Centroid. They have a pretty nice system that only works with AC servos, but the encoders are on the motor.
Currently the CHNC 11 has the encoders on the ball screws and there is a belt between the motor and the ball screw.
What is your feeling about the effect on accuracy if the encoders were on the motor with a belt driving the screws?
Thanks Gary
My Tree CNC lathe had encoders on the ballscrews rather than the motor, with a toothed belt between them. I replaced the brushed DC motors with LiChaun AC servos (ie encoders within the motor) and Centroid Acorn. Works fine.
I don't see any big distinction between the 2 arrangements. You still have some backlash between the ballscrew and the table / toolpost and within the slides themselves.
The next step beyond encoders on the ballscrews or motors) would be closed loop control using linear encoders (usually glass or magnetic scales). Centroid have just released the "Hickory" and "Acornsix" controllers which build on the Acorn system by catering for encoder position feedback. That might be an option if your finances permit.
https://shopcentroidcnc.com/hickory-cnc-controller/
https://shopcentroidcnc.com/acornsix-cnc-controller/
The main difference seems to be that the Hickory uses ethercat - Acornsix is probably more appropriate for the likes of us.
Thanks my only real concern is that the CHNC II is billed as a super precision lathe, not that I am doing anything super precision, but I hate to introduce any additional error if it isn't necessary. When emailing with Centroid, the recommended the Hickory system and the Leadshine EL7 AC servos. If having the encoders on the motor is an ok option, it appears to be a pretty simple retrofit solution. The only other issue to overcome is that I want to use the Fanuc AC Spindle Drive and Motor, but looking at the Centroid retrofit manual, it looks like that would work as well.
I did ask them if they have a motor that will work with the Hickory controller that has a separate encoder so it could be mounted on the screw, but I haven't heard back from them yet. There are only certain motors that will work with Hickory at this time. I am not sure what is required to get the Leadshine motors mounted in place of the Siemens DC servos yet. Still looking for a drawing of the motors.
Thanks for the thoughts, Gary