CNC PCB milling

[NeverDie]

@dbemowsk said in CNC PCB milling:

@neverdie I can see that the left side looks like it cut deeper, which would cause that. I still think it is the thin traces though. Can you do any manual leveling of the bed? If so, I would attempt that. My Anet A8 3D printer, though not a CNC, is all manual leveling and I do have to check it from time to time.

I suppose I could shim under the sacrifice board with slips of paper to get the right height.

I would have thought that the autoleveling would have accurately compensated though. Not sure why it isn't, especially if I'm autoleveling at 2mm spacing. Maybe this is an area where a future version of the GRBL driver will get it right.

[executivul]

@neverdie cnc milling auto leveling is not like 3d printing bed leveling.
In 3d printing the firmware probes the bed, stores the values and automagically compensates at every move.
In cnc milling the host control software asks the grbl to move and probe each point, then it modifies the gcode accordingly and sends that gcode to the controller which in turn just moves in the xyz coordinate system.
Now, are you sure your gcode is updated to reflect the leveling?
Normal flatcam gcode has a few Z-0.1 moves, the rest are only G1X...Y... (without Z) and autoleveled gcode has almost every move with z, like G1X...Y...Z.... Chilipeppr also adds some comments at the end of the lines like "new Z" or "Z mod".
If you run the autoleveling probing but do not hit "send autoleveled gcode to workspace" in CP or "apply height map" in OpenCNCPilot the probing is useless.

[dbemowsk]

@neverdie I just wasn't sure if your CNC had any way of manually leveling the bed. My 3D printer has screws in the 4 corners of the bed for me to manually level. I have been sticking to manual leveling on it because I have heard of people that have switched to auto-leveling that have had a number of problems. Maybe you need to switch to a different type of auto-leveling sensor.

[dbemowsk]

@executivul Interesting. I did not know this. That is quite different.

[executivul]

@dbemowsk said in CNC PCB milling:

@neverdie I just wasn't sure if your CNC had any way of manually leveling the bed. My 3D printer has screws in the 4 corners of the bed for me to manually level. I have been sticking to manual leveling on it because I have heard of people that have switched to auto-leveling that have had a number of problems. Maybe you need to switch to a different type of auto-leveling sensor.

What firmware? BEST sensor ever = piezo sensor. I get 3 micron accuracy and repeatability. I use it on Marlin 1.1.8 with UBL on a corexy printer and also on a Smoothie delta.

[dbemowsk]

@executivul I don't have a sensor yet. If I was going to switch to one I would have to flash a different firmware on my board. It is a newer version of Marlin.

[executivul]

@dbemowsk said in CNC PCB milling:

@executivul I don't have a sensor yet. If I was going to switch to one I would have to flash a different firmware on my board. It is a newer version of Marlin.

Marlin=Arduino (usually Mega2560). So UBL. What type of printer?

[dbemowsk]

@executivul I do manual leveling and I rarely have to level the bed itself.

[andrew]

@neverdie for pcb milling a small height difference across the board's area could result in a big negative effect if autoleveling is not performed properly.
I would double check your autoleveling process, on the other hand, you could also try to use bed flattening with a bigger endmill tool first, then place the pcb to the flattened area. this could help you to eliminate or decrease the cnc assembly or the sacrificial board caused roughness.

for further details please see the following link:
http://flatcam.org/manual/procedures.html#bed-flattening

[executivul]

@dbemowsk said in CNC PCB milling:

@executivul I do manual leveling and I rarely have to level the bed itself.

"If it works don't break it!"

To end the offtopic, for anybody interested RepRap forum piezo discussion and https://www.precisionpiezo.co.uk/ the store of the guys who brought the piezo to the world, there are 1:1 clones on aliexpress, but the price is almost the same if you buy only the board and print your own holder and they provide warranty, fast shipping and are really great people so they deserve some support! (I'm not affiliated in any way to their store)

LE. @NeverDie if you read this please look up 4-5 posts and tell me if your gcode has the autoleveling

[NeverDie]

@executivul said in CNC PCB milling:

if you read this please look up 4-5 posts and tell me if your gcode has the autoleveling

yes. In Chilipeppr I send the autoleveling to the workspace, at which point it modifies the gcode.

[NeverDie]

@andrew said in CNC PCB milling:

@neverdie for pcb milling a small height difference across the board's area could result in a big negative effect if autoleveling is not performed properly.
I would double check your autoleveling process, on the other hand, you could also try to use bed flattening with a bigger endmill tool first, then place the pcb to the flattened area. this could help you to eliminate or decrease the cnc assembly or the sacrificial board caused roughness.

for further details please see the following link:
http://flatcam.org/manual/procedures.html#bed-flattening

What's the best type of sacrifice board to use if doing bed flattening? Regular wood, or should I stick with particle board or MDF?

[NeverDie]

It seems the z-height only needs to be off by a small amount (say 0.02mm) to have a noticeably negative effect on the trace width.

[dbemowsk]

@executivul said in CNC PCB milling:

"If it works don't break it!"

Amen to that

[andrew]

@neverdie said in CNC PCB milling:

What's the best type of sacrifice board to use if doing bed flattening? Regular wood, or should I stick with particle board or MDF?

I never had to try it, so I don't know. for me the regular wood seems to be logical...

[zboblamont]

@neverdie Sorry @andrew, wood is not a stable material, it is not inert and changes shape, it breathes if you will.

Artificial materials such as particle board or MDF have a stable matrix, but of the two, MDF is the more uniform and least reactive, hence it's popularity for kitchen cupboards, loudspeaker cases, etc..
It is easily machined, but is hard on the typical router bits, although nowhere near that of HDF.

The bigger problem with mdf is the very fine dust produced when milling or cutting it with power tools, it gets everywhere, including your lungs, and can even get through vacuum cleaners without hepa filters....

I would suggest either a thick MDF or layers glued together, cut the sides to requirements, lock it in place and set a 6 or 8mm router bit for MDF in a collet, and set it to work, ideally outside where you can hose it down after sweeping up and vacuuming.

[NeverDie]

I think there may be some kind of coating that comes on these PCB's. Maybe it's to prevent corrosion of the copper? It seems to come off using IPA. Here one is before cleaning with IPA:

And here is the same PCB after cleaning with IPA:

Not sure whether it interferes with soldering. Anyone know? For now I'm cleaning it off.

[dbemowsk]

@neverdie Looks like you have that board dialed in...

[dbemowsk]

So a CNC question as I build my prototype machine. I have an arduino with CNC shield on order that should arrive tomorrow. My question is, does this allow for a display and keypad for local control or is everything just done through GRBL?
https://www.amazon.com/SODIAL-Arduino-Compatible-DRV8825-StepStick/dp/B074FVTTR7/ref=sr_1_7?s=electronics&ie=UTF8&qid=1517499916&sr=1-7&keywords=arduino+uno+cnc

[ben999]

@dbemowsk

All done via GCodeSender and similar i'm afraid... which is not a problem as far as i can see

Your Arduino is fully populated with this board :

• driving stepper drivers
• end-stops
• coolant (or wathever) enable
• spindle control (speed and rotation) (V3 shield can do PWM control if your spindle driver can). Be aware that this board doesn't drive any kind of motor. Just sends signal
• cloning axis of your choice
• emergency stop, pause and resume
• i might forget some details but i'm sure you get the point

But it's a pretty cool board for veeeery reasonable price. I love it !

[dbemowsk]

@ben999 Thanks can't wait to try it out.

[dbemowsk]

@ben999 So reading through your specs, everything seems pretty straight forward. I get that the spindle motor will need it's own controller which can be driven by the CNC board. One thing I am confused on though, what is "cloning axis of your choice". I am new to CNC.

[ben999]

@dbemowsk this board can accept up to 4 stepper drivers

your machine is probably a 3-axis CNC, so in most cases one stepper per axis could do the trick.

BUT sometimes some CNC designs require a 2nd stepper for an axis (usually the x-axis). Then add a 4th driver and some jumpers and you're done. You end up with 2 steppers doing exactly the same job

[dbemowsk]

@ben999 That makes sense. My 3D printer has 2 Z axis steppers.

[dbemowsk]

So I received my CNC board today. After looking over the board, I see that it has 2 endstop points per axis. Logical assumption would be that there would be one on each end of travel of an axis. I was a bit confused because my 3D printer only has 1 per axis. Do I need to use both for each axis when running GRBL?

[dbemowsk]

So today I decided to do a little bit of testing with the CNC shield and my Y axis drive. To start off, one of the issues I have is that I do not know the specs of the motor that I have. Here is a pic:

Did quite a bit of searching over the past week to try and find any information, and I can't seem to find what I need. I decided to do some tests anyway. I was able to get GRBL loaded on to the UNO board fine and connect to it through PUTTY. I then installed just the Y axis driver and connected a 12 volt power supply for starters. I was able to get the Y axis screw turning by sending some Y axis commands such as "Y20" or "Y-20". When I ran it initially I didn't have the Y axis platter mount secured to anything, so that would just spin with the screw. The problem I have is that if I hold the platter mount giving the screw even the slightest bit of resistance, the motor will spin for half a second and then start making a whining/buzzing noise and will just vibrate a bit. Once the motor starts to slow down while stopping, the screw will stop whining and spin slightly at the end of the stop. I then thought that it might be that the voltage was too low as I had read on some forums about some Epson steppers being higher voltage steppers, so I connected a 30 volt supply. Running the same test gave me the same results. Does anyone have a clue as to why this would happen? Basically the motor doesn't have any torque.

[executivul]

@dbemowsk current gets you torque, voltage gets you speed. What driver do you use? See the current setting for the driver you use, how/what voltage you measure to determine the current and the formula that links the two.

When I want to set the sweetspot (assuming driver and stepper are somewhat matched) I start low and increase the current little by little until the stepper gets warm after a few minutes. Warm, not hot! But again, the driver and stepper must match, I wouldn't try that with a Pololu A4988 and a Nema 34 stepper

[dbemowsk]

@executivul said in CNC PCB milling:

What driver do you use?

DRV8825

[executivul]

@dbemowsk well then check the current set by measuring Vref, plenty of tutorials on the web. If it's lower than 500mA try bumping it up little by little. Also try reducing the microsteps to 16 or 8 by removing jumpers.

[NeverDie]

The adjustment screw on the stepper module allows you to set max allowable current.

Also, in software, you could try setting a lower max acceleration parameter (as discussed earlier in this thread) and lower max velocity parameters, in case either or both are a factor.

You're brave to be attempting this with unknown equipment. If there are multiple factors... it's trickier to troubleshoot with no working baseline to start with.

[NeverDie]

I've changed my modus operandi so that now I use the largest pad and trace sizes that I can get away with. That way, if they get whittled down during the milling, I still hopefully have enough left.
Example:

[NeverDie]

That way, I can also cut deeper and wider, which helps with soldering.

[dbemowsk]

So good news. I was able to get the motor turning with some torque. Initially the board was set at full step resolution (no jumpers on M0, M1 or M2). I thought I'd try it at it's lowest microstep resolution which is 1/32, and the motor ran good. I checked the max current and it was set quite high at 2.035 x 2 = 4.07 A. Running some Y axis moves got the motor a bit warm, but I don't think I ran it long enough to do much damage. I now have the current down to 0.254 x 2 = 0.508 A and it seems to be running quiet and good. With it being at 1/32 microstepping it took a lot of revolutions to move the platter mount any distance, so I bumped it to 1/16 microstepping. This is mainly due to the 4.8:1 gear ratio that I have. I think this should give me some pretty good precision on the Y axis at least. Haven't built the X or Z yet, but that is next.

[dbemowsk]

@neverdie said in CNC PCB milling:

You're brave to be attempting this with unknown equipment.

I'm not too worried. Worst case I may have to get some other steppers. Being that this motor is salvaged, it's not like it costed me anything.

[NeverDie]

Here's my first attempt at copper removal between the traces/pads:

I used a 2mm endmill to do the work pretty fast. It actually came out pretty good, but the remaining un-removed copper is not what I wanted. I guess it's inevitable unless I use a supper narrow end-mill, or probably a carving bit. But that will take a lot longer.

i.e. some kind of hybrid approach would be best, I should think. Is there some way to do the detailed cleanup with a small bit, and then switch to the large diameter bit for the fast cleanup where precision isn't needed?

[dbemowsk]

@neverdie Looks like you could raise your end mill a bit (no pun intended). Shouldn't it be the same depth as your other bit?

[NeverDie]

@dbemowsk said in CNC PCB milling:

@neverdie Looks like you could raise your end mill a bit (no pun intended). Shouldn't it be the same depth as your other bit?

In theory, I suppose maybe so. However, the trouble is I can't preserve and then re-apply the same auto-leveling results. And, after the initial isolation milling, I can't do another to support setting the copper removal depth, because the touchpoints might hit the already removed material, which would seriously skew the results. So, I just set it deep enough to cover the possibilities, and that's why it's deeper.

If anyone has a better solution to that, I'm all ears.

[NeverDie]

I did get better removal using a carving bit:

[dbemowsk]

@neverdie My apologies for near hijacking your thread. I am going to create a new one so I don't mess yours up.

[NeverDie]

@dbemowsk No worries. Not really my thread. It's @andrew 's. I'm just a disciple.

[NeverDie]

Answering my own earlier questin, I forgot to remove the PCB coating using IPA before soldering. I think it probably did make the soldering come out a bit funky looking:

Next time I'll remove it first.

Appearances aside, though, it should still work.

[NeverDie]

I made the breakout board a bit smaller, and then, since the proof of the pudding is in the taste, I connected it to a pro mini and ran it. Voilà! It works like a charm.

[NeverDie]

@andrew What is the widest diameter end-mill bit that this CNC machine can accept and handle? I'm thinking in terms of the mill leveling discussed earlier, and going wide to speed up the process.

[zboblamont]

@neverdie Pardon my interruption, but I seem to recall my mentioning this near the start of this CNC adventure, and that there was a collet available for that device to accept 8mm standard wood routing bits...
I realise that US traditionally holds to 1/4 and 1/2 inch shafts, but the bigger the size the greater the mass to spin up and motor requirements... Pretty sure you would still be able to find 8mm bits...

[NeverDie]

@zboblamont I have an ER11 though, so doesn't that preclude using a bigger collet?

[zboblamont]

@neverdie Google it... The ER11 8mm collet indeed exists, although the ads on Ali seem to sell it partnered with a shaft extension.
I'm not familiar with your machine, but well acquainted with wood routers...
Old size bit shafts were 1/4" and 1/2" and probably are predominant in the US, metric were 6mm and 8mm etc... Have only used 8mm personally, but with 1500-2000w routers they make short work of mdf.
For the job you have in mind a 1/4" is probably meaty enough if your motor has the power to handle the cutting rate...

There is no reason to go big on the bit diameter, a 1/4" would be plenty strong enough to do what you require and should be readily available stateside... If you have the collet of course...

[NeverDie]

@zboblamont So, indeed, it looks like there is quite a range of collet sizes available for ER11: https://www.maritool.com/Collets-ER-Collets-ER11-Collets/c21_56_60/index.html

It's harder to judge what end-mill diameter the motor can handle though. The motor looks very light duty.

[NeverDie]

Wow, the ER11 collets on Ali Express are incredibly inexpensive--just a little more than $1, with free shipping: https://www.aliexpress.com/item/6MM-SUPER-PRECISION-ER11-COLLET-CNC-CHUCK-MILL-CNC-Workholding-Mill-Lathe-VEC05-P0-25/32795483676.html?spm=2114.search0104.3.44.64385e319AOvIk&ws_ab_test=searchweb0_0,searchweb201602_5_10065_10344_10130_10068_10324_10547_10342_10325_10546_10343_10340_10548_10341_10545_10084_10083_10618_10307_5722320_10313_10059_10534_100031_10103_441_10624_442_10623_10622_10621_10620_10142,searchweb201603_2,ppcSwitch_5&algo_expid=3eb984d6-bba6-4935-8f90-0f6a6da33214-6&algo_pvid=3eb984d6-bba6-4935-8f90-0f6a6da33214&priceBeautifyAB=0

Unless someone here has theory or experience to contraindicate, I guess the only way to answer the question is to buy some end mills and collets and try it out.

https://www.aliexpress.com/item/5pcs-4-Flute-HSS-End-Mill-Diameter-2mm-6mm-Router-Bit-Set-straight-shank-milling-cutter/32619808737.html?spm=2114.search0104.3.15.3b3f1d457qqIFL&ws_ab_test=searchweb0_0,searchweb201602_5_10065_10344_10130_10068_10324_10547_10342_10325_10546_10343_10340_10548_10341_10545_10084_10083_10618_10307_5722320_10313_10059_10534_100031_10103_441_10624_442_10623_10622_10621_10620_10142,searchweb201603_2,ppcSwitch_5&algo_expid=f46173c9-8f6c-4633-9e93-26d8769c6ac0-5&algo_pvid=f46173c9-8f6c-4633-9e93-26d8769c6ac0&priceBeautifyAB=0

[NeverDie]

@neverdie said in CNC PCB milling:

try the Shurtape GG-200 that was recommended in the Hackaday article.

I finally received it, and I tried it yesterday. Definitely has more grip than the regular double sided Scotch tape.

[zboblamont]

@neverdie If you go 6mm or 1/4 inch only to flatten the bed, you might be better off just selecting the collet after buying a plain router bit locally, probably neither will be used for much if anything until you need to replace the bed.... They are pretty cheap usually and ideal for mdf, just keep the cutting rate cautious with the light drive motor, and beware of the dust, it goes everywhere, and you can't spray it as water affects the mdf...
A 1/4" face cutter would put less load on your motor than a 1/2", just take it easy on the milling rate....

[NeverDie]

I'm not sure how to analyze it a priori. If I have to run it at a much lower feedrate to compensate for a light duty motor, then the supposed speed benefits of a larger end-mill might not materialize.

[zboblamont]

@neverdie said in CNC PCB milling:

a priori

? You mean as earlier, sorry, I didn't use Latin after 60BC, it rather went out of fashion don't you know....
The power you require to bring the mass up to speed is one thing, the speed of approach and the area to be cut is the greater puzzle I suggest. The bit will not snap for sure, your motor might object if you take big bites though...
If you rip the surface in a 0.5mm top slice and listen to the motor you will get a good idea of how it reacts, then either increase the approach speed or the depth I would suggest.. You are using the device outside it's parameters, better to be cautious, the bits are a lot tougher than the motor...

[NeverDie]

I ordered the collets and end-mills that I referenced above, so after they arrive I'll post an update as to how it went.

Meanwhile, now that my CNC is in the garage, it's hard to know when it has finished running. I'm thinking of making a node to monitor it and then sent me a wireless alert when it finishes. So, how best to know when it has finished? I'm presently toying with the idea of monitoring its current consumption, on the theory that if no significant current is flowing, then it is done. However, maybe someone here knows of a better way?

[zboblamont]

@neverdie I find it hard to believe you would just walk away from it with no curiosity or desire for direct observation. Have you considered using the highly technical door to access and use eyes to see and ears to listen ? You are breaking new ground here, or mdf actually, where is that American spirit? Did Galoka think the Ulus were too ugly to save?....
Beyond the giggles, I suggest a strategically placed vacuum cleaner intake as close as you can get taped down and just watch what happens, you are in a better position to cut power than 20m away diverted by Youtube....

[NeverDie]

@zboblamont
Watching it do an auto-level on 300 points is pretty boring.... I have better things to do than watch paint dry.

[zboblamont]

@neverdie Hey, your choice, your motor.....
Watching paint dry is most under-rated really, except when you are on the other end of the brush...

[Yveaux]

@neverdie why not put a webcam/ipcam at it, like many people do with a 3d printer?
Will show you when it's finished and when something goes wrong.

[NeverDie]

@zboblamont I think maybe you're misunderstanding. I would watch it do the bed-level (well, at the beginning anyway), to monitor drain on the motor. However, with all the auto-leveling and etching and milling and drilling I just let it run and come back when it's finished. That's really where I want to go with this in the end anyway. As far as protecting the motor, there should be (but maybe isn't?) something to shut it down if the current jumps above a certain limit. I should think that would be all the protection that's needed. So, maybe that's yet another reason to task an arduino with monitoring the current flow. One reason would be to alert me when the present job is done. The other reason would be to shut it all down if the current got too high (indicating a failure situation of some kind).

In fact, the beauty of having it in the garage is that it can run in the background without being noticed. If I were to run it in the house, the noise is rather overwhelming, and it would be hard to ignore it.

[NeverDie]

@yveaux said in CNC PCB milling:

@neverdie why not put a webcam/ipcam at it, like many people do with a 3d printer?
Will show you when it's finished and when something goes wrong.

I just don't see the allure in that. It requires my polling it, rather than being interrupt driven. To me it's like putting a webcam on your dryer to observe when it's done. I mean, yes, it is faster than walking over to it, but it just seems far from optimal.

[NeverDie]

So, anyhow, maybe another way to check for job completion would be to monitor the DATA+ line from the USB cable feeding the WoodPecker controller. I'm guessing that if it shows no activity, then the job is done.

[dbemowsk]

Is there something like, or does octoprint work with CNC machines? That might be an option if it does. Octoprint can monitor with an attached USB webcam

[NeverDie]

Yesterday I was doing copper removal with a 2mm end-mill, and several times when plunging it brought the motor to a complete stop. The motor recovered and continued in the x-y dimension at the new z.

So, obviously, the z-feedrate was too high. Is there a feedrate specific to the z-axis, or is it just the same as the general feedrate used by the x-y? I know there are velocity and acceleration parameters that can be different for x,y, and z. Do I control it using that instead of the seemingly general purpose "feedrate"?

Also, not sure why I experienced the problem this time and not previous times, as I was running the same g-code as before. Perhaps the end-mill acquired too much plastic gunk clinging to it? What's the best way to clean that stuff off?

[NeverDie]

@neverdie said in CNC PCB milling:

Is there a feedrate specific to the z-axis, or is it just the same as the general feedrate used by the x-y?

I ran some tests and proved that the general feedrate does apply to the z-axis. However, at the moment, the rest of the answer remains unknown to me.

[NeverDie]

@neverdie said in CNC PCB milling:

So, anyhow, maybe another way to check for job completion would be to monitor the DATA+ line from the USB cable feeding the WoodPecker controller. I'm guessing that if it shows no activity, then the job is done.

I tested this, and it won't work. It seems that USB is constantly sending data frames (or something) regardless of whether actual serial data is being transmitted.

So, to take the idea any further, I'd have to tap into and monitor the Serial Rx pin on the Woodpecker atmega328p chip itself.

[ben999]

@neverdie just a thought
One is able to customise a program header and footer of GCode programs
Maybe you could use a “coolant off” and “coolant on” (M07, M08 and M09 if I remember correctly) to achieve your goal?

[NeverDie]

I tried measuring the current with an INA219, and wow, the current measures much higher than what my bench power supply had been telling me. Doing just auto-leveling consumes about 0.5a at 24v. Turning on the spindle and just carving air consumes over 1a at 24v. So, I'm guessing that actually routing PCB material (not just air) might well exceed the 2a limit of the INA219, at least in worst case scenarios where the bit may bind (or the feedrate too high) enough to slow the motor down.

So, I'll simply slam a couple more shunt resistors in parallel with it, and then it should be good for measuring up to 6 amps.

All this assumes that the Woodpecker contains a proper snubber diode, so that I don't get voltage spikes when turning the spindle off. Indeed, it looks as though the woodpecker does have an SS54 schottky diode installed just below the spindle power plug, and I'm guessing it is intended to serve that purpose.

[NeverDie]

Virtually every screw on this CNC has come lose, so I'll be applying Loctite on all of them to hopefully avoid a repeat in the future.

I think there will just inevitably be some amount of vibration during the milling process (which obviously gets worse if there are lose screws). So, beyond Loctiting everything, I wonder if it makes sense to also rest the whole thing on some vibration dampers such as:
https://www.amazon.com/Anti-walk-Silent-Feet-Anti-Vibration-Machines/dp/B00536VQE0/ref=sr_1_8?ie=UTF8&qid=1518798934&sr=8-8&keywords=vibration+laundry

I mounted our clothes dryer on these particular feet, and they work fantastic, at least for that purpose. Maybe not the right choice for this CNC though.
In the case of 3D printers, I notice that the Prusa I3 Mk3 sits on rubber feet, presumably for a similar reason.

Anyone found good vibration damping feet for their CNC?

[zboblamont]

@neverdie Crazy though it may sound, have you considered cheap engine/gearbox mounts (for vehicles) for the base frame onto some baseplate or U-mounts?
All excess energy in a machine has to be dissipated to maintain base accuracy, the more rigid the structure is, the more that undamped energy is transferred to somewhere to dissipate, the base is probably the easiest to resolve...

[NeverDie]

@zboblamont
Not sure how that would be setup exactly.

I'll try this:
https://www.amazon.com/dp/B00B84FNBS/ref=cm_sw_r_cp_ep_dp_H.1HAbVV7GBAF

That way the aluminum frame will be supported all the way around.

[NeverDie]

Here is the Sorbothane installed:

[dbemowsk]

@neverdie Out of curiosity, I thought most CNC machines used ball screws. That to me looks like a lead screw.

[NeverDie]

@dbemowsk Yes, it's not a ball screw.

[dbemowsk]

@neverdie Good, then I don't feel so bad using plain 5/16 threaded rods in my build. I would assume that there is some sort of anti backlash where it connects to the carriage though, correct? There is not a lot of play in mine to begin with, but I am using two threaded rod coupler nuts with a spring in between on mine to take up any little bit of backlash it might have.

[NeverDie]

@dbemowsk said in CNC PCB milling:

I am using two threaded rod coupler nuts with a spring in between on mine to take up any little bit of backlash it might have.

Yes, mine has a similar spring, and I assume for the same reason.

[NeverDie]

For now, I've settled on this hardware for monitoring the CNC when I'm not in the garage:

The bottom unit monitors the current that the CNC consumes to determine whether or not the CNC is active. When the CNC transitions from active to inactive, it sends a wireless signal to the battery powered node (above it in the photo), which rings a buzzer to let me know that the CNC has finished.

Since my 3D printer that's on order is also 24 volts, I think there's a good chance this hardware may work with it as well.

[NeverDie]

The sorbathane helped a little, maybe, but there's still a lot of vibration. Found this video on how to dampen a 3D printer. It has some Interesting ideas on how to dampen vibrations:
Silencing the Prusa i3 MK2 & Horrible Vibrations – 05:13
— CNC Kitchen

The video is aimed at Mk3 i2 owners, but I imagine similar tricks might work for a small CNC such as that discussed on this thread.

[NeverDie]

Based on a file test, it appears that the smooth rod is not hardened. Not sure how much improvement I might see if I upgraded to hardened.

Unfortunately, this thread seems to have fizzled out....

[NeverDie]

Looks as though upgrading to hardened and chromed rods is fairly inexpensive:

https://www.aliexpress.com/item/Optical-Axis-300-380-400-500-mm-Smooth-Rods-8mm-Linear-Shaft-Rail-3D-Printers/32838846750.html?spm=2114.search0204.3.2.7fc04405d43d4M&ws_ab_test=searchweb0_0%2Csearchweb201602_5_10152_10151_10065_10344_10068_10130_10324_10342_10547_10325_10343_10546_10340_10548_10341_10545_10084_10083_10618_10630_10307_5711220_5722320_10313_10059_10534_100031_10629_10103_10626_10625_10624_10623_10622_10621_10620_10142%2Csearchweb201603_2%2CppcSwitch_5_ppcChannel&algo_expid=04efdb04-9516-4d2e-a8d0-ec277a9f66b3-0&algo_pvid=04efdb04-9516-4d2e-a8d0-ec277a9f66b3&priceBeautifyAB=0

So, less sagging in the middle, and maybe less vibration also.

[dbemowsk]

@neverdie As part of my CNC build, I have gotten 8mm rods that I pulled out of old scanners and old ink jet printers, I guess I never checked to see if they were hardened or not. How much of a difference would it make?

[NeverDie]

@dbemowsk Well, let's put it this way: the set screws on my rods had come lose, so I noticed considerable sagging when the spindle got to the middle of the rods. When I tightened up the set screws, a lot of that sagging went away. So, I think that means the spindle is now, in part, being supported by tension.

Hardened rods should flex less. I'm sure there are formulas that could tell you by how much. Also, there are different degrees of hardness.

[NeverDie]

On the other hand, according to this thread:
http://forums.reprap.org/read.php?1,113177
deflection isn't affected by hardness. I had thought hardened would be stiffer.

[dbemowsk]

@neverdie So I have a few questions regarding your spindle assembly. First, do you know how many RPMs your motor is? Next, how many collets do you have and use for your setup. Third, is your collet and chuck assembly like this one?
https://www.amazon.com/Adealink-Spring-Collets-Extension-Holder/dp/B078JQQ5X7/ref=sr_1_2?s=hi&ie=UTF8&qid=1519193790&sr=1-2&keywords=5mm+shaft+collet

[NeverDie]

@dbemowsk said in CNC PCB milling:

@neverdie So I have a few questions regarding your spindle assembly. First, do you know how many RPMs your motor is?

I don't know. You could try asking the seller: https://www.aliexpress.com/store/product/Freeshipping-spindle-775-with-ER11-High-speed-Large-torque-DC-motor-Electric-tool-Electric-machinery-12/424291_32809235881.html?spm=2114.12010612.0.0.47291db35wj5jd

Next, how many collets do you have and use for your setup.

One.

Third, is your collet and chuck assembly like this one?
https://www.amazon.com/Adealink-Spring-Collets-Extension-Holder/dp/B078JQQ5X7/ref=sr_1_2?s=hi&ie=UTF8&qid=1519193790&sr=1-2&keywords=5mm+shaft+collet

Your link says its an ER11A, whereas I think mine may be just an ER11.

Though my spindle eventually gets the job done, I think a more powerful spindle could maybe cut faster

[dbemowsk]

@neverdie said in CNC PCB milling:

I don't know. You could try asking the seller

Thanks, I wasn't sure if you bought yours as a package deal where they had the specs on all the parts.

Your link says its an ER11A, whereas I think mine may be just an ER11.

I did find this link (http://www.cnczone.com/forums/cnc-tooling/319670-er11-type-vs-b-type.html) which states that the A should designate the style of the clamping nut. Looking at the different styles they have, it just looks like they just use different tools to tighten the nut. A looks like a standard wrench type.

This is the only information I have on the motor that I plan on using. I googled all the numbers and couldn't find anything.

It is a German made motor, so I am assuming it is a pretty good motor. It feels like it has pretty good torque. It is one that I had in my junk parts bin. it has a 5mm D type shaft. In the pic I just have one of my 5mm to 8mm couplers on it. I think I may order that collet assembly as it sounds like it should work. It says that it uses ER11 collets anyway.

[NeverDie]

@dbemowsk Maybe you'll want to try using TMC2130 drivers for your stepper motors?

[dbemowsk]

@neverdie I do already have the DRV8825's that came with the controller. I guess I'll try them and see how they work and if they don't seem to do well, I'll consider changing.

[NeverDie]

As it turns out, the smooth rods on my CNC are 10mm in diameter, not the more common 8mm found on 3D printers. Definitely not hardened: I can see grooves where the ball bearings have scratched into it.

Sanladerer strongly recommends using precision tolerance hardened chromed rods for 3D printers, so I can only assume the same would apply to CNC.

Since it's a relatively cheap upgrade, I may do it.

What tolerance should I get? h6? Also, what spec for surface roughness?

[scalz]

@neverdie said in CNC PCB milling:

Definitely not hardened: I can see grooves where the ball bearings have scratched into it.
Since it's a relatively cheap upgrade, I may do it.

What tolerance should I get? h6? Also, what spec for surface roughness?

not surprised it is not hardened for the price, cheap price->cheap quality
pity is sometimes even more expensive kits cheat on this too..

I would say as long as you use h6/h7 with roughness around 60 or more, so, precision hardened, chromed nice too, you should be very fine. day&night vs cheap rods which, in first place are not made for this job+precision..

[NeverDie]

I made some enhancements to the remote used with the CNC Monitor. Among other things, it's now more compact, and the top now has a ground plane (which may help the RF):

The bottom makes most of the connections and will hold the two AA batteries:

[dbemowsk]

@neverdie So I have an odd CNC related couple of questions. Next, when you put a bit in the chuck, do you always bury it in to the base of the bit? When you have your bit in for working on PCBs, what is the distance from the end of the chuck to the tip of the bit that you use?

[NeverDie]

@dbemowsk Is this for calculating deflection?

[NeverDie]

@dbemowsk I haven't been very consistent, but I'd say roughly 7/8 inch.

[dbemowsk]

@neverdie Actually, I am working on my Z axis and am trying to gauge how long I should have my rods. They are currently 8.25 in (209.55 mm). Wondering if there would be any benefit to leaving them that long or shortening them up some more.

This is the spindle frame as I have it designed so far. The blue piece on the motor shaft is just one of the couplers that I bought for the steppers to connect the threaded rods. It is just for looks until I get the chuck.

[NeverDie]

@dbemowsk said in CNC PCB milling:

@neverdie Actually, I am working on my Z axis and am trying to gauge how long I should have my rods. They are currently 8.25 in (209.55 mm). Wondering if there would be any benefit to leaving them that long or shortening them up some more.

In that case, what matters is whether you can raise the spindle high enough on the z-axis to remove the used bit and insert the next one. I wish my z-axis had a bit more height on it. Sometimes I have to move the spindle away from the workpiece in order to get enough clearance for a tool change. I suppose it doesn't help that I'm using a 3/4" waste board.

[dbemowsk]

@neverdie said in CNC PCB milling:

I suppose it doesn't help that I'm using a 3/4" waste board.

Do you need 3/4"? Wouldn't 1/4" or 1/2" work? Do you drill in that far?

[NeverDie]

@dbemowsk 1/2" is probably ideal for my particular setup. 3/4" is simply what I tried first. 1/4" might work, but with less margin for error.

[NeverDie]

@dbemowsk Shouldn't you have some linear ball bearings gripping those z-axis rods?

[NeverDie]

Here is the above PCB after assembly:

As you can see, it offers more potential functionality than the earlier version. It still has a buzzer, which is now mounted underneath the pro mini.

[dbemowsk]

@neverdie FOr starters I am trying 3D printed rod bearings. I';ll see how that goes. The beauty of designing this myself is that I can alter the design whenever I want.

[zboblamont]

@neverdie Di the relocation of the buzzer increase audibility as you were seeking?