That is a good looking belt reduction. :D
Looks like you have everything figured out :cool:

"4. sneak as many complex curves into the design as possible :D ..."
This is my favorite point... :D :D

So now get a mill bit and see if you can cut them out of steel on your table.

For sure could cut them out of aluminium.

I agree about loosing the steppermotor hole.

I am going to start a tech thread on pullies becuase I have a question and do not want to hijack your thread :D

Hi Mike

Hijack away, its the wild west in my thread.
A question for you, I have looked at your site http://www.cvsupply.com/servlet/the-...-Plates/Detail and it does not say what is the reduction.
Maybe put a note on your site for the easily confused like myself.

The pulleys look great by the way as there is no need for them to be all alloy.

I have recently done far worse than trying to mill 6mm mild steel.
I set up a pinion and using a HSS bit then later in a separate experiment a carbide bit, milled 14mm holes twice through 30mm of steel.
The HSS bit deflected and the hole was conical, 1mm out over the 30mm.
The carbide bit being stiffer was more accurate but unfortunately not good enough for nicely rounded hole.
The internal walls on the bore were however beautifully polished using .5mm DOC.

Using a 2 flute 6mm carbide mill at 1500 RPM and constantly wetting with oil, I believe the Mechmate would get the job done eventually. In the end buying from you or having your own design laser cut is just quicker !!

Regards
Ross

All the 1926 Chev kick plates are now on the car.
Here is the pic I promised.
The car looks like it has its Sunday best on now.

Attachment 13481

Regards
Ross

Beautiful! nice work.

Belt Reduction Drives
 

The pulleys and new pinions arrived today.
So I shot down to the local machine shop and asked the owner how much to machine the pulley bores out and do some pinions for me.
It was a simple formula he gave me apparently it was expensive as he had to use a 4 jaw chuck.
His final figure - $35 AUD that not in total ...each.

This happened not long after.

Attachment 13484

Not at all impressed with being asked to pay $490 AUD to machine $300 AUD worth of parts, the Mechmate was tasked with making its own upgrade.
Just did some final axis calibration, drilled 4 holes in each pulley and screwed them directly to the machine.
Somehow I managed to snatch a victory from the jaws of a defeat.

Once the videos upload to Youtube i will post the links.
The actual machining video is a bit slow so please don't watch it all.

Regards
Ross

I reckon the machine shop guy is engouraging you to play with your MM more :D:D

BTW, I'll do the same if I were you... BUT luckily I have more than a dozen machine shops just within a stone throw... literally :cool:

And some of us even bought a small lathe to help in future upgrades :)

Did you find the center with a touch probe?

Hi All

Yes David the probe finds the exact center by just touching the extreme inside edges of the hole on both the x and y axis then halving those distances.
The routine runs twice, once fast and then a second pass slower for more accuracy.
The probing facility is built into Gerry's 2010 Screenset for Mach3 along with a range of edge finding functions.
The probe tip I use is the shank (no flutes remaining) of a 4mm end mill.

Here are the videos

Center Finding
http://www.youtube.com/watch?v=En76h4c8ans

Boring 72 Tooth Timing Pulley
http://www.youtube.com/watch?v=mqfCwb65sIQ

Finishing Pass
http://www.youtube.com/watch?v=EfScbUfUPwA

Danilo - If I bought a lathe then i would be compelled to CNC convert it :).
Nearly did buy one just to bore these pulleys but the machine rescued of my wallet just in time.

The fit is of course not metal lathe quality but still perfectly adequate for the purpose.
I am let wondering now how well the machine will bore with the new belt reductions in place.
Any comments from belt reduction owners would be appreciated.
Presumably the cut may be a little better but other factors like slop, flex etc in the machine will also determine the absolute accuracy achievable.

Regards
Ross

You should try single flute cutters, recently I got one 6mm from a friend, it does wonders in aluminium. No melting, no bonding and no need to spray anything on it, pure fun!

Hi Danilo

It was a 6mm single flute alloy bit i used !! They work very well.
Like you I was given 2 by a fellow CNC user but only one was long enough to drill the full 25mm of the bore.
Considering there was eight pulleys to do, I flooded them with oil constantly just to be sure.
The 4 big pulleys done are now and the hold down for the 18 tooth pulleys is ready to test.

Thanks for your tip on the cutter.

Regards
Ross

Cutting 18 Tooth Timing Gear Bores.
 

Probably machined them too slowly but all are well and truly usable.

Here is how I clamped them to the CNC bed for machining.
The timber clamp is pinched around the pinion hubs with a screw or bolt then screwed to the machine.

Attachment 13485

Regards
Ross

Good job! That is a lot of effort!
Better keep it slow than to redo the job.

Definitely not on the fast track Ken but learning how to made do with what I have.
As you say, soon I will be done and the effort forgotten.

Regards
Ross

..next thing you know, Ross will have a working 4th axis and start making those gears from scratch! Well done Mate.

Hi Sean

Am I that predictable - guess so :D
What I did find was this http://gearotic.com/ from Art who makes the Mach3 machine controller application.
If one needed lots of different gears then it would be an easy path to pop them out probably in a hard plastic.

Belt Reduction Drives

Sent the DXF files off to the laser cutters.
I can report that MetalHead Mike's prices are very competitive.
It was $215 AUD for my design to be cut here.
Will will send a few photos once i gather up / make the rest of the bits for the drives.

I have a few questions for the forum so next post will set the framework, hopefully I might get some valuable input - more to follow.

Regards
Ross

Optimising for Belt Drives
 

This Youtube video is a great visual reference for anyone who ever wondered about the practical difference between full step, half step and 10x microstepping on a stepper motor.
http://www.youtube.com/watch?v=8zVp1Ivs7Vs

The next part of the formula is the reduction in torgue as microstepping increases. This table explains the relationship very clearly.

Microsteps - Holding Torque
1 - 100.00%
2 - 70.71%
4 - 38.27%
8 - 19.51%
16 - 9.80%
32 - 4.91%
64 - 2.45%
128 - 1.23%
256 - 0.61%

The most vague part of the formula is the way torque decreases with increasing RPM of the stepper motor.
Looking at the Torque vs RPM curve of a motor somewhat similar to mine (maybe) here
http://catalog.orientalmotor.com/ite...prodid=3001048.
In summary the torque falls away rather quickly at increased RPM but does not seem to fall away as much as the additional torgue the belt reduction provides.
Makes sense as why would you do it otherwise.
Detent torque (the bearing, mechanical and magnetic stiction characteristics) further muddy the water.
It is hoped that any changed microstepping values chosen are less than where these forces might significantly alter accuracy, something like a 2x value at most.

My question is..

In a like for like voltage, amperage and driver scenario, is it worthwhile giving away some of the remaining increase in torque a belt drive may provide for extra resolution provided by increased microstepping.

In some ways the question could be rephrased as..
Do we require all the extra torque the belt drive provides to improve cut quality or could a balance of the two be beneficial.

Finally I realise that this all depends on the job in hand and that increasing microstepping values for jobs that don't need torque is possible but not really that practical.
So is it torque or resolution we need ?

Thanks
Ross

Not enough resolution, the surface may be rough and you have to sandpaper
Not enough torque, the motor may stall and you have a gash (or fire)

:)

Hi Gerald

I laughed when I read your response.

The numbers initially point to me increase the microstepping from 10 to 16 (a 1/3 increase) after fitting a 4:1 reduction.
That yields an electronic improvement in resolution, a mechanical improvement in resolution and a marginal mechanical increase in torque at the pinion.
More from each parameter, that is by definition optimization :)
And I will buy some sandpaper just to be safe !!

Regards
Ross

I noticed the forums standard stepper motor driver the Gecko 203V is native 10x microstepping.
That might explain why most are silent as to how changes to microstepping might change performance in conjunction to belt drives.
The other more obvious explanation is that I'm a PITA as my mates so often point out :D

One last thought to record on the forum before I leave this one alone.

Changing the specification of the NEMA 34 ungeared motor to a higher torque motor when used in conjunction with a 4:1 reduction may help to restore some of the torque loss from my proposed increased microstepping. The detent torque unfortunately also rises along with the increased motor torque so with each change does not come without hurdles.

The reason I went down the path of belt drives in the first instance was because of this ...
"One of these inexpensive motors (MotionKing 34HS9801), with a belt reduction added at a later stage, will rival the performance of the "first choice" geared Oriental Motors mentioned right at the top"

I'm simply trying to test those assumptions initially made in 2006 and see if in 2012 tweaking a few things can give us a little more for roughly the same investment.
I have already run my machine at 16X microstepping and nothing bad happened.
Now I want all my torque back too ;).

Regards
Ross

Do you know that almost all drives morph speeds? above certain rpm all drives morph to full step to regain torque.
Even geckodrive. at 267 RPM it starts to morph and at 565 RPM it goes to full step :)

here it is explained:

http://www.geckodrive.com/how-morphing-works/

Hi Danilo

My DM856 Leadshine driver manual makes no mention specifically of morphing.
I believe the feature your refer to as morphing is contained in the DSP or Digital Signal Processing algorithms within the driver but it makes no mention of the any micro stepping corrections at RPMs it might make.

In the end it is pointing to the difference between an analog and a digital driver.
I remember specifically choosing a DSP pathway assuming it would be more flexible over analogue but this does leave totally in the dark on specifics.

Thanks for your link as it all goes towards helping me build a complete picture.

Regards
Ross

I posted the question about morphing to the Eric from Leadshine here in this topic, as I am interested too.

At my friends (Pajka) in macedonia we used DM856, and they are more stable and quieter than my 2M982. I also blew one off with a short circuit in a cable jack and was able to fix it quicklu with a 0.1ohm resistor and a IRF540N mosfet :).

LED Light made with the Mechmate
 

Today I made a 40 Watt LED Light with the Mechmate, it was a great exercise in what the machine can do.

The Machine made 2 x acrylic lens, an alloy mounting block for the LEDs and for the first time I made a PCB to mount the 4 LED drivers.
It is extremely bright and uses a measly total of 37 watts to do so.

Front
Attachment 13490

PCB inside uses the exclusion method with the traces drawn in CAD
It was fast, clean and easy.
Attachment 13491

The laser cutters are slow with my belt drive parts, so that project remains on hold for the moment.

Dry Machining of Aluminium
Forever the optimist, I had another go a dry machining.
Big fail and two dead 6mm end mills later, it all got thrown away.

My RP7 oil method works very well, however after my latest dry machining fail it was more than ever obvious that two things were needed 1. cooling and 2. lubricity.
The reason the spray oil method works is that it provides the lubricity to reduce the friction heat build up.
Compressed air alone (and lots of it) as a coolant was a poor second to oil.
The best cutting fluids however also contain water as when the water evaporates the change of state from water to vapour produces significant cooling.
This time as I dry machined I kept note of the workpiece temps.
Initially all goes well but a the heat builds eventually the aluminium starts to melt to the cutter.
In the end what this means is that if using fast feed rates and very shallow passes the heat input has time to dissipate and the aluminium remains machinable.
What this also means is an overall slower machining time.

So while the oil was working their is also further room for improvement using a glycol concentrate (I am assuming) and water mix.
All the water borne fluids tend to use a 20:1 dilution yielding enormous amounts from just a few liters of concentrate.
The diluted fluid is very slippy, so I can't wait to have a go with a squirt bottle.
While the pros all know this it takes time for someone like me with no previous exposure to run through all the options and find the best method for our machines.

Droplet Mister
The best delivery method for these water based fluids is a droplet mister setup.
I have experimented with how this is best achieved and short of having a good needle valve to control the coolant flow, all is working well.
The secret to forming droplets rather than mist is to have two separate air supply pressures in the system.
The fluid reservoir is held at a higher pressure than the final air delivery pressure.
The internet is full of pressures that people state work for them, however as I experimented it became clear that the difference in pressures was the real factor.
Using the same setup with two variable air pressure regulators I was able to have the system move from no coolant, to droplets and finally to a mist just by varying the pressure differences between the two lines.

Prototype settled, I will send a few pics once the parts for the droplet mister have arrived.
Ultimately I am trying to have a system that works to cut aluminium without flooding the area with excessive moisture.

Long post I know so thanks to anyone who read through to the end.

Regards
Ross

Ross it sounds like you might have already found the plans for the DIY fogbuster but if you haven't a quick google search will pop them up on Pratical Machinist forum. Work great, mine was set at 10-15 psi.

Great post! I enjoy reading every single word.

Hi Russell

Yes had a look in fine detail but thanks for the link, always appreciated.
This is a good setup using the same water filter cartridge holder as the original.
http://www.youtube.com/watch?v=xQENewWM68Q.

My implementation of the needle valve is no where near as good.
If anyone knows what that neat brass coloured valve block is called / where they can be found, I would appreciate it.

Cheers Ken, really I just have a poor memory and have to write it down before I forget it :)

Regards
Ross

Great post Ross, very thorough and very informative as usual :D you certainly do your homework !!
Well done, and a nice addition once you get it all up and running.

Cheers
Tony.

Gidday Tony.

Ross

Cutting Fluid Results
 

Did a quick test of this concentrate (for you Aussie people available at Hare and Forbes).

Attachment 13497

The result was excellent. The machined part remained cool.
Significantly less odour than spray oil and an easy clean up.
Machining rates were significantly increased.

Attachment 13498

I applied the coolant with a hand sprayer typically used in kitchens for cleaning.

Regards
Ross

Ross,
How often do you spray? Do you have to keep pumping at the bit throughout the cutting action?
I brush on little kerosene on the cut path (on the alum) with a small brush for kitchen use. but when you have to do that for 5 hours continuously (time to cut a 4'x8' 1mm Alum sheet... it gets a little frustrating with muscle ache as side effect...
I've been thinking of trying out hand sprayer for a while but procrastination gets int the way. :o
The Zero Fogger seems the ultimate solution. Thanks again to Russel for bring up this topic.

Hi Ken

Once every "Lap" seemed to be quite enough.
Used 100ml of coolant or 5ml of concentrate costing 5 cents.
Still smiling.

Regards
Ross

Ross

That turned out very nice.

:cool:
I'm astonished ! what incredible possibility has the mechmate ... masterly driven by a creative owner !
just curiosity which mill did you used for "helen" ? it looks a very small sign.
Thanks
Sergio

Hi Sergio

3mm 2 flute solid carbide end mill at 30mm/sec, 22,000 RPM and .3mm depth of cut.
Bevel (done first) 90 degree 3mm V bit at 20mm/sec, 22,000 RPM and .4mm depth of cut.

The smaller the end mill the faster the RPM is a general rule.

Regards
Ross

Ross, I wish I was able to remember all the links I use to have before IE crashed on me. I use to know where to get it but I want to say it was for a small OA torch. Also here is the link to the Patent to for Fogbuster. http://www.google.com/patents?id=khs...390854&f=false
Reading the abstract will give you the important numbers like nozzle size and length. For the nozzle we use a Mig Tip. We did make our own Valve out of .75" Brass. Turned down the end of a 1/4 28 socket head bolt in a drill press to look like the picture. Took no time with a angle grinder and light touching.

I think precision liquid metering needle valve is the generic name.

Coolant Unit
 

Thanks Russell and Ken

Tried tapping an Acetal block as per the diagram - failed so will try again later.

Got the bottom end of the system done though.
Cross threaded the plastic on the canister input side as I was getting tired.
Think I got away with it.

The design can deliver air only, coolant only or air and coolant in any ratio due to the independent valving.
The input regulator allows full pressure supply to other air powered devices like the vac pump.

Can't post any more photos into the forum at present so I will get on to that.

Regards
Ross

Attachment 13501

Regards
Ross