So I originally joined up and posted the following thread in the welcome forum pretty much 2 years ago ...
Since then I became overly cautious about diving in to CNC, and went for a 3D printer instead, and I must say I've really made (and continue to make) good use of it, making automotive brackets and so forth, and really feel like I've got my money's worth out of it. So my mind recently wandered back to my original plans for a home-made CNC so I started reading up again. And watching various YouTube videos, trying to find someone who had approached it how I'd like to approach it. Then I stumbled across this guy :-
And *that* right there is exactly what I'd have envisaged as the perfect machine, and it also shows a build how I had figured it would have to be. And so it's confirmed for me that it's just too big a job to take on at the moment, I simply don't have that level of machinery around to make a decent enough job of it, and I see no point in doing it unless I'm going to do it properly.
Soooo... I'm thinking of a stop-gap learning exercise. Downgrade my expectations somewhat. And I'm finding it really hard to talk myself out of a cheap, crappy Chinese-made 6040 (I know, I hear the groans already). Thing is, even for their faults, I'm finding it hard to knock them in terms of outright value as a learning exercise. The 6040's have the 1605 ball screws, the high-speed (ish) motor for PCB work and by all accounts they'll do some soft aluminium as long as speed isn't a prime concern (it's not). I'll look to the VFD models with the better spindle, although won't fuss to much over the kw rating, they seem to vary anything from 0.8 to 2.2 on eBay.
Best I can make out, I can do some choice upgrades as time permits (or as they fail) :
- Replace wiring when it starts to fracture
- Fit a nice bed plate and level it myself
- Upgrade the electronics if/when they fail (sound like more "when")
- Swap the water pump pretty much straight away
- Do they all come with limit switches now?
- Various different representations of the control boxes on eBay; the 'newer' ones seem to be the black box design; is any of it really important or are they all equally as bad on the inside of that box?
- Am I right in assuming the ballscrews are C7 grade so about the best that can be reasonably had at that size/price point?
- I'm fairly au fair on the electronics/fabrication side of things, are there any cheap (open source/DIY) approaches to the motor drivers and control boards that I can look at to save money and get a higher quality solution? Would want something compatible with the main software, either LinuxCNC or EMC3/4.
- Can't help but think while I'm at it move to EMC4 and USB
- Any other choice upgrades/modifications I should consider? Thinking this is just a learning exercise for a few years, I'm happy to concede it won't deliver the quality level I originally intended in my welcome thread, but like I said, I'm just lowering my expectations and putting it down as experience.
- Am I right in understanding the main limitation are these TBS drivers? Uprating them will allow the motors to be driven at higher V&A reliably? Are the TBS drivers just substandard/badly matched to the motors or something? It seems a common criticism... but then some get on fine with it "from the factory"?
I'm put off by the 3020 as I know they are cheaper but their use of trapezoidal screws concerns me for the accuracy/repeatability issues, particular with any PCB stuff I do, and also their slower, less powerful motor (again, my understand is the PCB milling will need some real high speed - I mean, ideally 60k+ but I can settle with 24k).
Am I being overly picky? Is the 3020 not significantly worse than the 6040 and I should just go super-cheap? Will the 3020 realistically do aluminium at *any* level, even if I accept the 6040 would only scratch at it :) ?
Then hopefully in a couple of years when I have more space and equipment (and knowledge!) I can contemplate doing something awesome like This Old Tony :)... but realistically, I just don't have the time right now :(
Thanks in advance for any thoughts/opinions. Happy to be shot down in flames ;)
edit: Thought - go 3020 and upgrade the ball screws to 1605-C7's? Worthwhile? Then uprate the spindle over time to a VFD?
Last edited by brumster; 19-01-2016 at 10:06 PM.
Be afraid Very afraid.!!!. . . . . Nah but seriously now I wouldn't buy one unless it's purely for learning purposes and I wouldn't waste a penny trying to upgrade one. . . . Can't make Silk purse out of sows ear.!
Q1: Not seen any of the Cheaper ones come with limit switches or Home switches.
Q2: Mostly junk and the better ones are still Cheap Analog drives best avoided.
Q3: Yes and C7 is more than good enough for router. C5 or higher screws will cost more than the machine.
Q4: Best advice is don't bother only ends with you buying proper drives. You can buy Lead shine Digital drives for less than £40 on ebay and nothing you'll find or build open source will come near in performance. No point trying to re-invent the wheel
Q5: Never heard of EMC4. Thou I think you may mean Mach4. In which case No stick with Mach3 for time being and Don't go with USB it's not stable enough. Ethernet is the way to go.
Q6: Upgrade what.? if 3020 or any of the Chinese machines then Nope don't waste a penny ugrading just use and sell when learnt and ready for better.
Q7: The TB based drives are Low voltage which if pushed have habit of letting out the Magic smoke. Again just avoid them they are bad news.
To learn with the Cheap chinese machines are great Entry into CNC but don't expect the world and Don't waste money trying to make better and you'll get value for money.
The Following User Says Thank You to JAZZCNC For This Useful Post:
Ah yes, sorry, you're right, I meant Mach4. Ok, that's cool then, I'll stay with the parallel port. Glad I kept that old IBM Thinkpad now :D
On Q6, well, I didn't know - I was just after any thoughts/opinions. Sounds like it's just not worth it, so that's cool.
Digital drivers is a new one on me - guess some more reading to do - do they need different motors, presumably? No problem if it's a complex answer, I'll do some reading on the forum, just curious.
I guess any upgrades to the electronics can be considered an investment, assuming the drivers and motors remain matched..... maybe I can keep the motors and spindle, if nothing else, if/when I do make my own.
Am I right in thinking the main thing I need that's really lacking is the rigidity in the gantry, the quality of the runners, lack of support on the Y and just general strength/"trueness" in the aluminium extrusion approach...
Here's the crux question then - for the PCB work and aluminium work (slow, hobby, prototyping stuff, low volume, 0.1" through-hole and SMD designs) is the extra £400 on a 6040 worth it over the 3020? The only real differences I can make out are the crappier spindle, the crappier ball screws and the smaller working size. The latter I couldn't care about :)
They work just same regards motors you can use.
All duly noted. Arrrgh, you're making me think maybe I should just build a scaled-down steel DIY jobbie instead :D !
Last edited by Clive S; 20-01-2016 at 08:37 AM...Clive
Aluminium profile and Aluminium plate still build a very strong and more than capable machine and can be used with minimal tools.
Use correct components along with good electronics and you'll have a machine that will knock all those Chinese machines and several other commercial offerings into tin hat.
I think my idea is to build it in steel just to test the theory and process of construction (I also have loads of 20 and 30mm box, 3mm wall, kicking about), although I fully appreciate it will be overkill at the size I'm looking. So idea is to build a small, desktop build but with the same concepts as the larger one I'd *like* to do one day (in another property, no doubt). But conscious I don't want to get too tied up with the construction/planning that I lose interest and it all dies a death. Having said that, I'm pretty determined in most things, so.... :-S
I'll do some more reading, mull over some ideas - and of course always value other opinions (no offence Jazz, I totally get what you're saying!)...
It would help if you decided on the physical size so that you can get some advise on the frame.
20 - 30mm box doesn't seem big enough unless you are going to build a 3D printer...Clive
Truth is they sooner or later realise the advise was good, unfortunatly it's usually cost them money to realise this. . . . Called the price of learning.!
Sad thing is it needn't have done if they'd only listen.
To be honest it's a double edge sword advising people because those that do listen never really fully realise the bullet they dodged by listening so you never really hear from them because things tend to go fine and without issues.
Those that don't are often embarresed so never really say anything publicly and just go with what was suggested first time. (Although do often get PM or email thanking me and asking again.!)
By magicniner in forum General DiscussionReplies: 15Last Post: 07-01-2015, 07:59 PM
By shapebusters in forum Machine DiscussionReplies: 17Last Post: 12-03-2014, 05:21 AM
By Fivetide in forum General DiscussionReplies: 4Last Post: 28-09-2012, 11:06 AM
By C.AlveSilva in forum Linear & Rotary MotionReplies: 1Last Post: 17-04-2012, 10:53 PM
By ali hedi in forum General DiscussionReplies: 1Last Post: 25-12-2010, 05:05 PM