Yes twin screws so 4 steppers.Quote:
Is that PSU of yours running 4 off 3nm stepper motors from that or 3.
The toroidal does not give out 68 volts . It is about 50V AC out but when rectified it is around 1.4 higher and becomes DC
Printable View
Yes twin screws so 4 steppers.Quote:
Is that PSU of yours running 4 off 3nm stepper motors from that or 3.
The toroidal does not give out 68 volts . It is about 50V AC out but when rectified it is around 1.4 higher and becomes DC
Ok, I understand the 4 wire ref now I just had to ask to eliminate any chance of any silly misunderstanding because series or parallel you are using 4 wires.Quote:
Originally Posted by PaulP
The noise you are hearing is normal and if you haven't done the auto-tuning they will even scream/squeal sometimes. The reason why the motors go quiet when ENAble is set is that the motor output in the drive is disabled.
Your motors are Nema 34 so running them in parallel becomes even more important for a router. Large motors like these have a higher inductance than smaller NEMA 23 motors which means you require more volts to get the same rpm. Torque drops away much sooner up the curve with NEMA 34's so if you have them wired in series this compounds this even more which is why higher volts are needed to get high speeds out of nema34s.
What does bother me and makes me wonder if you have an issue is that while they won't spin fast when wired in series they shouldn't easily stall at lower RPM and no way should a few bristles from a dust shoe cause missed steps.? A series wired NEMA 34 with 68V should rip that dust shoe off the machine and not even know it's done it.!
If you wire the motors in parallel and set the current to 4.2a then with 68Vdc you should expect to get approx 900 to 1100rpm before torque starts to drop away at any significant rate. So if you have a 5mm pitch and are getting 6000mm/min without stalling then your about right, but if you have a 10mm pitch then you should be seeing roughly double that.!
The setting is per phase so just set protune to the spec sheet per phase rating ie: 4.2aQuote:
Originally Posted by PaulP
I'm running 4x AM882 drivers with a 68v/500W PSU (Leadshine) to 4x SYS80STH86 - 3008BF currently wired in series. As for vel/acc I'm running OK at 6000mm/min with 1200mm/s2 without stalling.
Thanks again chaps, I can see my soldering station coming out again soon but before I do that will I need another 68v linear power supply to deal with the extra amps required for parallel? I read many posts from those that went from series to parallel and have had better results but it was either laziness or amperage concerns that pointed me to wire them in series.[/QUOTE]
Neale,Quote:
Originally Posted by Neale
Have another look at this please. My assessment of the drawing with it's 'start of winding' marks would suggest connecting A to -C and -A to C to create an in-phase pair of windings. It seems an odd way to label the wires but I think it's correct and agrees with the standard drawing....
Attachment 29543
Absolutely correct - that was a weird way to label the connections! "Start of coil" ends of windings (marked with star) should go together. I confess that I was looking at the lettering and not the "star" marks - and the lettering does not seem to follow a logical convention. Nothing would have blown up my way - but nothing would have gone round either...Quote:
Originally Posted by Kitwn
Well spotted!
Thanks for the replies gentlemen,
OK so I've been popping on most of the day for short burst to see what's been said but was unable to reply as I hate using my phone to do any forum posts. I've got back to my laptop to read your replies properly without the mither of a sugar loaded 4 year old.
I ran the program twice more this evening with the dust boot on and got the same results in both Trespa (HPL) and MDF. I nipped up the grub screws on the carriages to see if they were causing any issues ran the programs and had exactly the same deviation in the tool path at the same 2 locations. I'm concluding that there are two issues at play here. One I expected the other I didn't. Firstly it's definitely losing steps in both the x and y, the x was out by 0.23mm and the y was out by 0.18mm, I'm now going to change the wiring to parallel. I'll include the wiring diagram that I translated into english, if someone could confirm I got this right I'll try to get start on doing that tomorrow.
The second issue which is a bigger worry is the flexing on the Z axis. Those bristles are 0.6mm diam. the layer of bristles is about 6-7mm thick and when pressing down and sideways when it's machining it's flexing the Z axis at certain points because of the geometry of the part it was cutting. Bristles are easy to bend if the layer is only 2-3mm thick but 2 rows of a 6-7mm channel was enough to cause the z axis to bend enough to show up in the finished part. The bristles are already 15-20mm past the bottom of the cutter and when it's cutting 20mm deep they are forcing the z axis in 3 different directions, plus sometimes they go into the slot and get caught by the cutter. I'm opting for a single 0.6mm layer of skirt printed out in a strip to be fastened on to the rim. This should eliminate the forces acting on the z axis and contain the dust/chip spray when cutting. It's also much easier to make as it's 3 layers of PLA printed rather than cutting bristles off an old draft excluder and glueing them in vertically. We'll see how that goes. I'll include the sketch of what I mean below.
I didn't know it until tonight but I zero'd the tool to the work piece on both the x and y sides and pushed the neck of the spindle and on the side of the spindle to see what was going on. When I could see light passing between the workpiece and the tool I knew what was up. It was evident in both the x and the y but more so the y. This was disappointing but at least I know what was causing the movement. I'll measure the flex tomorrow/today but I'd estimate it's about 0.3mm in the Y and a little less in X. I should say that this machine is built from 20mm thick Trespa TopLab and uses supported 20mm rails. I was never meant to be a brute of a machine and was only for light materials such as MDF, foam, balsa and any other material used in pattern making. It is pretty tough though it cuts ali pretty nicely but you can't go too deep with it. The gantry design does allow me to lower it for more rigidity but I'm unsure if the flexing is the gantry upright plates or the combination of the Y carriage plate and Z axis assembly. For now I'll be using it for light machining and drilling ops for the MK2 version which will be a totally different design, mainly ali frame and components.
Without the dust boot it does cut MDF, Trespa and Ali to within 0.02mm to 0.04mm so I'm happy about the fact I can make the parts for the next machine on it. Hopefully with the wires in parallel and the printed skirt it should do what it needs to do without any more cock ups.
My bristles came from a draft excluder but your milk bottle idea made me think of the printed skirt idea, it's hands off and much less tedious to do. As for the wiring I think Kitwn confirmed what I thought when I read that so below is the schematic I'll follow when I rewire my motors...Quote:
Originally Posted by Neale
The printed skirt...
Attachment 29544
The wiring schematic...
Attachment 29545
That's good news as I'm now reassured that there's enough power there.Quote:
Originally Posted by Clive S
I did the tuning but wasn't too sure what I was doing or what the Kp and Ki settings were about but found a decent guide in a Leadshine ProTuner manual, out of curiosity I'll try it again before I go parallel. They do squeal a bit which is noticeable when the ENA is not active (as in powered down) but I'll try them again when I'm back on the machine.Quote:
Originally Posted by JAZZCNC
These motors are NEMA 23, I got them from Zapp a few years back but either they've discontinued them or for some reason, I may be wrong they seem to be listed as NEMA 24. Either way the frame size is defo 23. I might have used the wrong diagram but I wasn't brave enough at the time of buying to get the 34s. I bought all the electronics before I'd designed the machine. I was 10 years younger back then and didn't know exactly what I was doing or getting into.Quote:
Originally Posted by JAZZCNC
As for speeds I think they stalled when rapiding at 7500mm/min so I brought it back down to 6000mm/min which is pretty quick over the size of bed I'm on, it's 915 in x and 515 in Y so not that far to go but I cut MDF at 4-5000mm/min, Trespa at 3000mm/min but shallow cuts only. I knew when buying the ballscrews that I should go with 10mm pitch but corona had started to go mental this time last year and the ballscrews were a bit of a panic purchase. The next machine is having 10's where needed and changeable pulleys for high res work.Quote:
Originally Posted by JAZZCNC
As for the setting per phase or for motor total your last comment has really settled my mind. It's been bugging me for a while now with regards to what it meant. Not just on the router but on any stepper system. Anyway I can rest my mind on that one.
OK I'll see what happens with the new skirt tomorrow or Tuesday. Rewiring motor plugs, checking all the fasteners and re-tuning should see me good for a couple of days, I'll then report back.
Thanks again for your responses, I'd be goosed without this forum.
You're not the only one mate!Quote:
Originally Posted by PaulP
Those motors sound very like the ones I bought from Zapp around 10-11 years ago - described as Nema 24 as the case is just a little larger than '23. To reassure you, mine were used on my mk1 router and are now in use on mk2, still running just cool enough to hold (= almost but not quite too hot to hold!)
Mk1 router was MDF, built in a hurry for a special job. All the structural integrity and stability of cold-rolled cow dung. Mk2 is welded steel. I suspect that your mk1 will also lead to a much-improved mk2!
I would change the motor wiring one motor at a time (obviously the twin screws will have to be done together )Quote:
OK I'll see what happens with the new skirt tomorrow or Tuesday. Rewiring motor plugs, checking all the fasteners and re-tuning should see me good for a couple of days, I'll then report back.
After some time away from the workshop I have finally got some info to report back with.
OK first off one by one I rewired the motors in parallel and set the amps/phase according to the chart given for the motor after that I tuned them all upto remove the resonanse issues. Now they sit idle and are much quieter and I'm now confident that their torque is much more suited to my setup. I tried a cut at 5mm deep in Trespa (HPL) at 3000mm/min without any problems other than the expected noise from the cutter. It didn't skip a step and the final part was as expected.
As for the dust boot I tried the 3d printed skirts in PLA but if you're thinking of replicating it I'd say don't bother, it didn't work. PLA might be stiff but it's still too brittle, maybe another material might work but mine went into the bin. I work in a business that does plastic fabrication so had a decent offcut of clear flexible PVC sheet, the sort you see hanging on doorways. I ended up using that with 8mm strips cut in to act like the bristles would have done. It works well enough but I'd still say bristles would be better but maybe I'll do that at a later date with some 0.15mm paintbrush bristles.
If anyone is interested the machine allegedly pulls just over 3 amps from the socket which is slightly confusing given the amps being pulled by the spindle on the VFD is showing as 3.3a. All devices were powered from the same socket so that might need a bit of investigating. Without the spindle running it shows 2.3a when all 4 motors are running at rapid speeds.
So all issues are now resolved, for now. There will be more to follow no doubt but until they show up I'd like to thank everyone for their help and assistance on the all of the above.
Getting sensible measurements of current with the sort of pulsed waveforms you find around VFDs and stepper motors is not a simple matter. Many meters are designed to work with either fixed DC or purely sinusoidal AC and will give missleading results with anything else. Have a look at the spec for your VFD and exactly what the displayed current is meant to represent. As long as the numbers are in spec for the device and nothing is getting uncomfortably hot I would suggest not loosing too much sleep over the exact readings and how they relate to each other.
Kit