Final doc is up.
PDF, 3.3Mb
http://www.box.net/shared/lszkke346z
Word Doc, 2.3mb, in case anyone wants to improve on it.
http://www.box.net/shared/8a9abbtr61
Do let us know how you get on with yours
My low profile clamps
File online storage sites are blocked where I work - I'll have a look when I get home
...and I've just heard from the wife that a 'large box' has arrived. That was mighty quick delivery, Jarod only posted it on Friday! (I also had to 'fess up to the missus how much I paid - I've got a few hours to think of a good excuse - wish me luck!)
Just tell her you are engraving jewelry
I've had time to photograph the machine as I was unboxing it.
http://s532.photobucket.com/albums/e.../Tonsen%20CNC/
I've just powered it up and manually moved the X, Y and Z axis and everything moved...
One thing I have noticed is that there is no play in the Z and X axis, but there appears to be a small amount of play in the Y axis - any ideas? Is there a way of tightening or adjusting the nut?
Just reading your PDF now, mr_w
Looks like yours was better handled by the couriers then mine
I have just checked mine, there does appear to be some movement.
That play may not be a problem, we will know for sure when you start milling.
Yep, the package seems to have been carefully handled.
Despite this, you can see on the photos where the spindle motor has rubbed against the side of the machine bed and damaged it slightly. This could have easily been avoided by wrapping the spindle motor in cardboard or bubble wrap. My other concern about the packaging is that the Y axis stepper seems quite vulnerable with only a 20mm piece of polystyrene to protect it from knocks, if you look at the other pic you can see where the connector has cracked.
Having seen the various images that have been uploaded (showing lots of rough edges and tooling marks) I wasn't expecting the machine to look as good as it does. Sure, it does have the odd rough edge and tooling mark but overall, but I'm very pleased with the physical condition.
So far, I've hooked everything up to the PC and jogged all three axies around just to check the control box and steppers are functioning. All the screws are well coated with grease and everything runs smoothly and quietly. I've yet to start up the spindle.
A word about the CD - for starters, it's all in Chinese! I already had a licensed version of Mach3 installed on a dedicated machine so all I was looking for was some config information. I found it in a PDF file, but I had to rename it before it would open. Worryingly though, AVG detected several trojans on the CD, so proceed with caution.
mr_westie - Thanks again for the PDF. It'll be really useful for the next stage too...
As soon as I get the chance, I'll try cutting some test pieces - it might be a few days off though.
After reading this thread I decided to take the leap and got one of these off EBay.
It arrived about 10 days later in what seemed to be a good condition; however once I unpacked the unit I found that it had a rough trip.
When I connected it, using the control unit, only the X-Axis and the spindle worked, after setting up Mach3 the problem persisted. After a few emails between myself and the seller, I got the OK to open the control unit. Through some debugging (plugging Y into X etc) it looks like the DB25 interface board is the culprit. Almost three weeks after receiving the machine and unfortunately having to leave negative feedback the seller finally shipped a new interface board.
My current impression of the machine is that while things are ok everyone is happy, but when a problem occurs the support is extremely lacking and quickly disappears.
Hopefully I will have the replacement control board by the weekend so I can start using the machine, and update with a more positive impression.
Adriaan
PirCapAdi, sorry to hear that.
I am eyeing up their bigger maching with the faster feed rate of 2400mm/min.
http://cgi.ebay.com/ws/eBayISAPI.dll...m=110434297243
Now if I can sell this one to cover that one i'll be set
The seller's careless packaging, haphazard quality control and dodgy documentation does make this purchase a bit of a risk. It's worth a punt at $580 though.
I tested the spindle and it works well. I engraved a few test shapes into some acrylic sheet and the results are easily good enough for the work I'm doing.
I noticed that the backlash issue that affects the Sable 2015 is present in this machine too - not quite as pronounced, but still visible in long diagonal cuts. Unfortunatley, the closeup pics I took aren't very good - I'll try and take some more.
Does anybody know if the nut in this pic can be adjusted?
Which nut? Do you mean the Y leadscrew?
That's the one!
Not sure how you want to adjust it.
Do you want to try to get rid of the backlash?
It has not been a problem for me so have not looked into it.
Looks like I have lost some accuracy after testing some Drewtronics bits.
The spindle/collet assembly may have developed some more runout.
My cuts are now 0.4mm wide vs 0.3 before.
It has also developed the noisy motor problem rayellam mentioned.
Only when moving in the -y direction with MACH3 strangly.
The problem is not present when using the control box.
Will try the fix prescribed and take some pics.
Be careful what you ask for.
I wanted some thin copper clad, and I got it.
Visited a local pcb fab place and asked for some scraps.
advancedcircuits.co.nz
They were pretty cool and didn't charge me for some sheets, 300x300mm
-
We have
0.15 Double sided, yes probably more copper than fiber glass. Used for multi layer PCBs
0.30
0.90 Double sided
1.60 Standard stuff we get from local electronic shops.
The 0.15, no way I can hold that down properly.
While the machine does have the accuracy to scrape off the right amount of copper the board has to be dead flat.
Will make a floating head.
>>If I can keep the costs below us$150 would anyone else be interested?
0.30, as above, too much flex.
0.90, now we are talking.
The other problem I have come across is copper thickness.
http://www.richter-elektronik.de/en/...anchor=1010047
Let's assume those are industry standard.
We now have another problem, even if we can get the board flat, the variation in the height of the board can be as much as the thickness of the copper layer.
As can be seen here. Left board.
One way around this is use a standard end mill instead of a v cutter.
http://www.precisebits.com/products/...sebit-stub.asp
But that does not help if your board is 0.15mm thick and not flat.
For that you will need a floating head.
With my runout problem the track for my latest board were a little thin for hand soldering.
What was supposed to be 0.4mm turn out to be 0.2mm.
The only way was to reflow
http://www.sparkfun.com/commerce/tut...utorials_id=59
My quick test setup. I will get a hot plate for this later.
It was beautiful to see it in action.
I am no longer scared of all those fine pitched components, bring it on.
-
-
I finally got the “replacement boards”, however it is the wrong boards! Instead of the DB25 breakout board, I got two more stepper drivers; no complaints thou since I am sure they are worth more, however I am disappointed since I still can’t use the machine 100%.
Any recommendations on a DB25 breakout to get?
Searching on eBay mostly comes up with full systems, and since I now have 6 stepper divers it does seem somewhat redundant.
There is this one from the seller self http://cgi.ebay.com.au/ws/eBayISAPI....=STRK:MEWAX:IT
or
http://cgi.ebay.com.au/ws/eBayISAPI....=STRK:MEWAX:IT from a different seller,
or alternatively a completely new system:
http://cgi.ebay.com.au/ws/eBayISAPI....=STRK:MEWAX:IT
Also I am looking at getting http://cgi.ebay.com.au/18-1-2-10-ACM...d=p3911.c0.m14 with matching bearing mount http://cgi.ebay.com.au/3125-Bearing-...d=p3911.c0.m14 to upgrade the Z axis lead screw.
Any advice?
Take a look at the MB02 from www.homanndesigns.com - They're excellent, and cheap. Peter Homann is a member of this forum.Originally Posted by PirCapAdi
Regards,
Ray L.
PirCapAdi, you are not having much luck there
At least I may have found the problem with my poor cutting resolution.
Got hold of a Dial Test Indicator.
Looks like there may be some debries in the collet I may have missed.
It's still with my engineering friend so will test and report back when I have checked it out.
fyi
http://www.precisebits.com/tutorials/spindle_runout.htm
edit:
Ground mod done, just need to get the machne back to test.
Last edited by mr_westie; 09-28-2009 at 05:27 PM.
Looks like there is a problem with the jaws on the collet?.
One of them is extending slightly pass the other 2.
Now I am not sure if it was like this from new.
Has anyone else measured the run out at the bit?
Point A there is nelgible runout.
B there is about 0.1mm, enough to kill a 0.1mm end mill
Hopefully it was not this bad so should be under warranty?
Only had less than an hour runtime, and maybe 10 tool changes.
What is in the head.
The chuck is only pressed on, not screwed like larger ones.
Looks like I have a bad chuck.
It does feel gritty when turning it.
Intintally I went to a general engineering place for a new chuck.
www.tradetools.co.nz
They offered me this unit.
4mm JT0 ROHM KEYED CHUCK #111-72807
nz$100
I also learn how to take the old one apart.
Will take it apart later and see if it can be revived.
The shaft has a "Jacobs Taper", JT0 to be exact.
He could not find any precision information on it. Runout etc
Suggested I visited jewler suppliers if this didn't work.
So I did.
And look what I find,
http://www.regal.co.nz/category.php?sub_id=279
Ah ha, so they used a Number 30 style handpiece as the spindle head.
At least I now have a local option for parts.
The motors were not quite as reasonably priced.
nz350-600
In I went with my box of parts to see if I could buy just the chuck.
None in stock and I could not wait weeks.
I ended up buying a whole new handpiece.
They had the option of Basic one and premium one.
At $100 vs $125 I went with the premium.
No precision data available either.
But the cheapie felt worse than the tonsen one.
It had sticky bits when I turned it.
I also got a solder paste disepnser solution.
nz$60
Much cheaper than the other low cost option available.
http://newzealand.rs-online.com/web/...duct&R=5057540
nz$166
Runout is now much better, 0.03mm, workable but could be better.
This does mean the smallest recommended solid carbide end mill I can use is 0.15mm/0.006"
http://www.precisebits.com/faqs/cnc.htm
"Spindle Runout (TIR) - should be less than or equal to 2% of the tool's diameter (absolute maximum)."
Might have to look at one of those china spindles.
Morel of the story?
Keep it clean !!!
PS.
I will be testing a very simple floating head design shortly.
Last edited by mr_westie; 09-30-2009 at 04:05 AM.
Most larger chucks use a taper mount, not a threaded mount. The 5/8" capacity chucks we have for the Bridgeports at work use a JT3. JT0 is common on smaller chucks as is a 3/8-24 thread. Taper mounting is very precise, and should not be confused with "simply pressed on". It's also a good way to save a chuck when you've bent the arbor. Press out the bent arbor, and tap in a new one.
Understand that with any drill chuck you are not going to get good runout numbers. This is simply not the technique used to hold tools precisely. When you want to hold a tool with low TIR, the correct choice is a collet.
There is only so good you are going to get a drill chuck to be, you can keep chasing smaller numbers, but only blind luck is going to get you anything with any accuracy if you keep using one.
escott76, thanks for the tip.
What is the typical lower limit for a chuck?
Reply with Quote
