What's new
What's new

Micro Machining Milling Machine Option

Rcgiovannani

Aluminum
Joined
Apr 6, 2015
Hello,

I am currently in the market for a new machine. I would like to go in the direction of Micro Machining because most of the parts I make and quote can fit into less than a 1-2" cube and use tooling that is smaller than 1/16" in diameter. Mostly cutting non-ferrous materials, but I see 303, 304, and 410 SS from time to time (all small parts). Would like to have 3+2 capabilities, but would need at least 3+1. Tolerances I usually work with are +/-.0002" (this is usually only one or two dimensions on a print and are usually less than .1875").

I have looked at the following threads:



It seems like there isn't much information on micro machining on the internet. I'm trying to find someone with some personal experience with any of the machines below.

Here are the machines on my list:

Haas CM-1 - Honestly, I'd buy this machine rather quickly with the TRT70 if I could just see some more YouTube videos of it. I'm also concerned with the accuracy and repeatability of this machine considering Haas' thermal management (which is basically nothing except timers). I haven't really found anything online (except from Haas themselves) about it. This machine is well within budget even with 5-axis.

Datron Neo - I'm actually going to see a demo tomorrow for the Neo. The thing that gives me pause with this machine is that every opinion I've read about it is that it is mostly suitable for plate work. I've heard that it is very rigid and accurate. It's only available with up to 4 axis and I'm not sure if it is "flexible" enough for a job shop. Looking about in the $150k range with the options I'd like. This is doable from a budget standpoint. John Saunders from NYC CNC seemed to like it when he had one on demo. Thermal management is important and I will make sure to ask Datron tomorrow how the Neo deals with it.

Tsugami VA2 - Not much info on this machine is available. Currently awaiting a quote from Morris Group on one of these. Like the Datron, I believe it is only up to 4 axis. Awaiting on confirmation from Morris on that, though. Not sure how Tsugami manages thermals but I'm sure Morris group can give me that answer.

Considering:

Brother S300 - There's lots of information on this machine - just not sure how accurate it is.

Robodrills (Not very familiar with anything FANUC related)

Machines that are most likely out of reach lol:

DMG MORI HSC20
Makino V22
Yasda YMC430
Kern MICRO
Mikron HSM200

Used machines that would be okay, but seem hard to find:
Mori NVD 1500

Open to any and all suggestions!
 
I worked in a shop that had a number of Robodrills and they seemed to be decent machines.

However, I never had much time on them myself. I never heard any complaints from the operators about reliability or accuracy.

They do offer a machine with a 24,000rpm spindle, which is good for the tiny end mills and drills.
 
I got a CM-1 last year, with 36 tools, probing, 50,000 rpm spindle, wired for 5 axis, with an HA5C-T fourth axis. I cut a lot of small medical device parts, mostly Ti and 17-4, with some other stainless and plastics. I'm pretty happy with it so far. I think the small work envelope help the machine be more accurate and experience less thermal growth than their larger machines. I'm able to hold .0001" in X/Y if I need to.

No experience with the TRT70 yet; I might be getting one soon.
 
I am biased but the Brother Speedio S700X2 with 27K rpm spindle is remarkable. Stocked in the USA on S700 but can also be ordered on the S300 and S500 size machines. Great rotary options available up to 4+1. Where are you located? I can help you get in touch with local support. Attending IMTS next month? 7 different Brother models will be running there in the Yamazen booth with a few new surprises too.
 
Hi Rcgiovannani:
The answer to "What mill should I buy for micromachining" is a complex one and depending on what you plan to do, the answer can be all over the map.
If you primarily prototype and if you have lots of experience compensating for the shortcomings of the machine and if you get paid "Cost Plus" you can make a good return on a relatively low end machine like the 20 year old Haas Minimill that I have.

But as soon as you need to make time, and you need to rely on the mill to run production efficiently and for a whole day at a time for months, you need something far better.
My biggest limitations come from two places:
1) I don't have the spindle speed I need (I have only 6K and I should have at least 20K)
2) The kinematics of the machine aren't great, which limits me severely when I want to do operations like interpolate a hole for example.

The spindle speed thing is obvious...little cutters = high speed spindle, but not just how fast it will go.
It's just as important to control for thermal growth and to have a duty cycle that can tolerate running balls out all day every day with no runout and no performance degradation.

The kinematics is a whole 'nother rabbit hole...there is a poster on here named "cameraman" who adores this stuff and has posted frequently in the past on the benefits and shortcomings of many different designs.
If you care to dive in, do a search for his user name...his posts can be arcane but they are well reasoned and very knowledgeable too.
In his search for the ideal machine he's written a doctoral dissertation on probably every high end brand out there.

Most people just assess what they think they need and rely on the machine maker to deliver it for them.

If money were no object there's be a Kern Nano or a small 5 axis Hermle on my floor...they are dead nuts accurate machines and will stay that way practically forever and have the "Brag factor" in their favour too.
But you really pay for that "Brag factor" unless you're making parts to sub tenths levels of precision in which case you cannot get by without.

Going with "inferior" brands takes you only a small step down unless your needs are truly unique...the flavour of the day seems to be Brother.
I've never driven one but compared to my Minimill they are so far above what I can muster that for all practical purposes, they are plenty good enough for 99% of what you will ever quote.
Also you probably don't need to get too anal about what you buy, so long as you stick with a reputable machine (no Tormach please!).
Buy for the service you get, buy for the features you get, buy for the whizzo control, buy for the pretty colour of the machine...it's all probably not going to matter enough to matter.
Whether it's a Speedio or a Robodrill, or a Mazak or an Okuma...it will be just fine and you'll make lovely parts...all have benefits and all have shortcomings.

Of course if you do go high end, you also have to splash coin for everything else too:
The toolholders need to be premium and balanced.
The cutters need to be the high end 200 dollar brand.
The climate control must be excellent.
The machine foundation must be excellent.
The power must be clean and stable.
The list goes on.

If you neglect these things you've just wasted your money on the super cool toy, so budget as much for the stuff around the machine as you do for the machine itself.
You might not spend it all, but a million dollar Kern Nano in a drafty garage with a dirt floor is not going to cut it.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
One of the machines I run is a creative evolutions. Im happy with it and the support has been great. Ours is set up for graphite but they will customize and option them out to your liking.
 
I got a CM-1 last year, with 36 tools, probing, 50,000 rpm spindle, wired for 5 axis, with an HA5C-T fourth axis. I cut a lot of small medical device parts, mostly Ti and 17-4, with some other stainless and plastics. I'm pretty happy with it so far. I think the small work envelope help the machine be more accurate and experience less thermal growth than their larger machines. I'm able to hold .0001" in X/Y if I need to.

No experience with the TRT70 yet; I might be getting one soon.

Thanks for the information. Maybe the CM-1 is the right machine. You definitely cannot beat the price.

Hi Rcgiovannani:
The answer to "What mill should I buy for micromachining" is a complex one and depending on what you plan to do, the answer can be all over the map.
If you primarily prototype and if you have lots of experience compensating for the shortcomings of the machine and if you get paid "Cost Plus" you can make a good return on a relatively low end machine like the 20 year old Haas Minimill that I have.

But as soon as you need to make time, and you need to rely on the mill to run production efficiently and for a whole day at a time for months, you need something far better.
My biggest limitations come from two places:
1) I don't have the spindle speed I need (I have only 6K and I should have at least 20K)
2) The kinematics of the machine aren't great, which limits me severely when I want to do operations like interpolate a hole for example.

The spindle speed thing is obvious...little cutters = high speed spindle, but not just how fast it will go.
It's just as important to control for thermal growth and to have a duty cycle that can tolerate running balls out all day every day with no runout and no performance degradation.

The kinematics is a whole 'nother rabbit hole...there is a poster on here named "cameraman" who adores this stuff and has posted frequently in the past on the benefits and shortcomings of many different designs.
If you care to dive in, do a search for his user name...his posts can be arcane but they are well reasoned and very knowledgeable too.
In his search for the ideal machine he's written a doctoral dissertation on probably every high end brand out there.

Most people just assess what they think they need and rely on the machine maker to deliver it for them.

If money were no object there's be a Kern Nano or a small 5 axis Hermle on my floor...they are dead nuts accurate machines and will stay that way practically forever and have the "Brag factor" in their favour too.
But you really pay for that "Brag factor" unless you're making parts to sub tenths levels of precision in which case you cannot get by without.

Going with "inferior" brands takes you only a small step down unless your needs are truly unique...the flavour of the day seems to be Brother.
I've never driven one but compared to my Minimill they are so far above what I can muster that for all practical purposes, they are plenty good enough for 99% of what you will ever quote.
Also you probably don't need to get too anal about what you buy, so long as you stick with a reputable machine (no Tormach please!).
Buy for the service you get, buy for the features you get, buy for the whizzo control, buy for the pretty colour of the machine...it's all probably not going to matter enough to matter.
Whether it's a Speedio or a Robodrill, or a Mazak or an Okuma...it will be just fine and you'll make lovely parts...all have benefits and all have shortcomings.

Of course if you do go high end, you also have to splash coin for everything else too:
The toolholders need to be premium and balanced.
The cutters need to be the high end 200 dollar brand.
The climate control must be excellent.
The machine foundation must be excellent.
The power must be clean and stable.
The list goes on.

If you neglect these things you've just wasted your money on the super cool toy, so budget as much for the stuff around the machine as you do for the machine itself.
You might not spend it all, but a million dollar Kern Nano in a drafty garage with a dirt floor is not going to cut it.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

I will check out "cameraman". As I said above, I think the CM-1 is probably a "good enough" machine based on what mhajicek said. I've thought a lot about the costs that come along with higher end machines. I was worried about having to buy anything HSK haha. That would make me go broke. Huge investment.

I just got home from the Neo demo. It is a very cool machine and the dynamics of it are insane...but I feel like that's it. It's "cool". I'm sure it has it's place but I don't think it is in my shop.

Thanks everyone for the replys. They are super helpful. I'll let you guys know what I get.
 
Hello,

I am currently in the market for a new machine. I would like to go in the direction of Micro Machining because most of the parts I make and quote can fit into less than a 1-2" cube and use tooling that is smaller than 1/16" in diameter. Mostly cutting non-ferrous materials, but I see 303, 304, and 410 SS from time to time (all small parts). Would like to have 3+2 capabilities, but would need at least 3+1. Tolerances I usually work with are +/-.0002" (this is usually only one or two dimensions on a print and are usually less than .1875").

I have looked at the following threads:



It seems like there isn't much information on micro machining on the internet. I'm trying to find someone with some personal experience with any of the machines below.

Here are the machines on my list:

Haas CM-1 - Honestly, I'd buy this machine rather quickly with the TRT70 if I could just see some more YouTube videos of it. I'm also concerned with the accuracy and repeatability of this machine considering Haas' thermal management (which is basically nothing except timers). I haven't really found anything online (except from Haas themselves) about it. This machine is well within budget even with 5-axis.

Datron Neo - I'm actually going to see a demo tomorrow for the Neo. The thing that gives me pause with this machine is that every opinion I've read about it is that it is mostly suitable for plate work. I've heard that it is very rigid and accurate. It's only available with up to 4 axis and I'm not sure if it is "flexible" enough for a job shop. Looking about in the $150k range with the options I'd like. This is doable from a budget standpoint. John Saunders from NYC CNC seemed to like it when he had one on demo. Thermal management is important and I will make sure to ask Datron tomorrow how the Neo deals with it.

Tsugami VA2 - Not much info on this machine is available. Currently awaiting a quote from Morris Group on one of these. Like the Datron, I believe it is only up to 4 axis. Awaiting on confirmation from Morris on that, though. Not sure how Tsugami manages thermals but I'm sure Morris group can give me that answer.

Considering:

Brother S300 - There's lots of information on this machine - just not sure how accurate it is.

Robodrills (Not very familiar with anything FANUC related)

Machines that are most likely out of reach lol:

DMG MORI HSC20
Makino V22
Yasda YMC430
Kern MICRO
Mikron HSM200

Used machines that would be okay, but seem hard to find:
Mori NVD 1500

Open to any and all suggestions!
Last place I worked our UMC500 shipped without a thermal sensor at all and even after we figured this out and service installed it the factor it uses to calculate spindle growth was way off and my part dimensions were all over the place lol
 
I'm a new Brother S700X2 user. I came from a little hobby router made from aluminium extrusion. The interesting thing is that even a cheap thing like that can turn out (aluminium) parts that are repeatably within 0.02mm of each other. Meaning that the ballscrews and machine are nowhere near that accurate, but you can tell the machine to make something, measure it, then adjust the program to "aim off". It will make it's errors repeatably and you can correct for those

My Brother machine has Blum probing and that will repeatably tell me that once warmed up, my tools are within 1-3 microns of the length it measured "yesterday". If I ask it to probe some feature on the table, it repeats it's position each day similarly. BUT. The machine is in an unheated, non climate controlled space, with the sun on it, and when first powered on, it's generally measuring about 0.015mm different than the day before. After (some time), this goes back to showing yesterday's kind of measurements, and seems to stay there for the rest of the day. So warming up makes a difference (given you are chasing tolerances of 0.005mm)

Note: I'm not clear if the machine is literally warming and becoming longer, or if it just takes an amount of time for the built in temp compensation controls to kick in and apply an adjustment (everything still on defaults, feel free to jump in and advise me?)

Also I find end mills are not accurate in dimension to the micron. eg my 2mm nominal 4 flute, seems to be cutting about a 1.94mm slot in aluminium. However, if I stick that value into the CAM, then my parts come out within 0.01mm of desired size. So measuring your end mills carefully will be important to be able to hit your tolerances

One thing which is nice about the Brother vs the Datron (and other machines) is the speed of tool change. At 1 second ish, you can flip tools as much as you want without a real penalty. I can speed up my jobs by using a 6mm to rough out a "square" hole, then use a 4mm to tidy up the corners, then finish with the 3mm. When you look at how much harder you can push a 4mm than a 3mm, this saves quite a bit of time. Plus you can be pushing that 6-8mm roughing tool really rather hard, compared with what would be possible on the Neo. Then consider that this huge speed allows you to do tool break checks virtually for free (nice if you are using micro tools). For production work there would be a big difference in speed between the Neo and a Brother

The Brother can have a 16K spindle with CTS (which is nice as you can also have it running through the collets to ensure your micro cutters aren't recutting chips. However, the 27k spindle doesn't have CTS. So if you are mainly doing sub 6mm end mill work, you would need to make a tough decision on spindle choice...

Everything I do is plate work or small parts that are 1-2" ish. You can see a bit more here, and feel free to ask any questions?
https://www.instagram.com/hi_fly_cnc/

What I would say is that hitting tolerances of your 0.005mm (0.0002"), is going to need some fettling and dialing in. Sharp tools, good metrology, repeatable and precise work holding, etc. However, it's quite hard to be out by 0.02mm on this machine, so you aren't a million miles from what you want
 
I recently did a hardend D-2 tool steel project on a Mkron HSM200U. Project required cutting with the end of the endmill on opposite sides of the part which is tough becouse all five axes come into play. The width of the part had to be +/-.0001" so a little tough to account for tool wear. Probing the semi finish pass and tweaking the TLO from the program was easy peasy. Hit that tolerance every time (on the first try) it was run on machine. Yeah, rotaries to zero, tool change, probe, rotaries to zero, tool change, position b90 co, cut, flip C180, cut. So Z/B axis had to sync well within 50 millionths per side.
Pretty cool trocoidal milling channels at 30k where the machine is just vibrating instead of visibly swinging arcs. It's a badass piece of kit.

At another shop, we zeroed the probe on top of a part. Trainee was pretty impressed with the decimal places display. He blew some very weak compressed air on the probe for 3 seconds and re-probed the part... .000030"

100% duty cycle 50,000 rpm spindle with all the scales and coolant through the granite based / bells an whistles.There is certainly cheaper out there but hard pressed to get better CPk repeatability down to the micron
 
Last edited:
I'm a new Brother S700X2 user. I came from a little hobby router made from aluminium extrusion. The interesting thing is that even a cheap thing like that can turn out (aluminium) parts that are repeatably within 0.02mm of each other. Meaning that the ballscrews and machine are nowhere near that accurate, but you can tell the machine to make something, measure it, then adjust the program to "aim off". It will make it's errors repeatably and you can correct for those

My Brother machine has Blum probing and that will repeatably tell me that once warmed up, my tools are within 1-3 microns of the length it measured "yesterday". If I ask it to probe some feature on the table, it repeats it's position each day similarly. BUT. The machine is in an unheated, non climate controlled space, with the sun on it, and when first powered on, it's generally measuring about 0.015mm different than the day before. After (some time), this goes back to showing yesterday's kind of measurements, and seems to stay there for the rest of the day. So warming up makes a difference (given you are chasing tolerances of 0.005mm)

Note: I'm not clear if the machine is literally warming and becoming longer, or if it just takes an amount of time for the built in temp compensation controls to kick in and apply an adjustment (everything still on defaults, feel free to jump in and advise me?)

Also I find end mills are not accurate in dimension to the micron. eg my 2mm nominal 4 flute, seems to be cutting about a 1.94mm slot in aluminium. However, if I stick that value into the CAM, then my parts come out within 0.01mm of desired size. So measuring your end mills carefully will be important to be able to hit your tolerances

One thing which is nice about the Brother vs the Datron (and other machines) is the speed of tool change. At 1 second ish, you can flip tools as much as you want without a real penalty. I can speed up my jobs by using a 6mm to rough out a "square" hole, then use a 4mm to tidy up the corners, then finish with the 3mm. When you look at how much harder you can push a 4mm than a 3mm, this saves quite a bit of time. Plus you can be pushing that 6-8mm roughing tool really rather hard, compared with what would be possible on the Neo. Then consider that this huge speed allows you to do tool break checks virtually for free (nice if you are using micro tools). For production work there would be a big difference in speed between the Neo and a Brother

The Brother can have a 16K spindle with CTS (which is nice as you can also have it running through the collets to ensure your micro cutters aren't recutting chips. However, the 27k spindle doesn't have CTS. So if you are mainly doing sub 6mm end mill work, you would need to make a tough decision on spindle choice...

Everything I do is plate work or small parts that are 1-2" ish. You can see a bit more here, and feel free to ask any questions?
https://www.instagram.com/hi_fly_cnc/

What I would say is that hitting tolerances of your 0.005mm (0.0002"), is going to need some fettling and dialing in. Sharp tools, good metrology, repeatable and precise work holding, etc. However, it's quite hard to be out by 0.02mm on this machine, so you aren't a million miles from what you want

One of the best posts ever on any fora for machine accuracy !
 
Hi hi-fly-cnc:
The point you make is a good one but of course, there's much that is missing.

In a nutshell, a cheapo machine like the one you describe cannot be relied on to make the most basic geometry to a reasonable standard of accuracy.
So round holes ain't round, angles are not like what was commanded, finishes are not great, surfaced features are gouged etc etc.

Even if you can compensate, who would want to when you are running a business, especially a micromachining business where minor errors often have an outsized impact.

Ponying up for a proper machine as you've obviously done is a no brainer...the question for the OP is: how good does it have to be for his kind of work.

If he needs a Hermle and buys a Haas, he's in exactly the same boat you were in when you only had the router...yeah he can do some of it (as I do) if he's really experienced at sidestepping the shortcomings, but oh how wonderfully nice it would be not to have to even think about overcoming them, and how nice it would be to be able to take on the hard, tight tolerance stuff too.

That, to me is the real point.
Too bad I'm old and can't justify getting a good one anymore.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Yes.
No.
The hermle needs 12.000 $/mo payments and the haas needs 3000$, and a decent cash 25k payment at startup drops the haas below 1000$.
But not the hermle.

It is quite unlikely the biz will never have a downturn when money is tight during 5 years of paying for the machine.
It´s quite unlikely, but possible, the biz can make 30k/mo profits with the hermle, to get over the hard times -- and keeps the cash in the bank.

Many shops go down with machine payments.
Cycle time, productivity, accuracy are usually lies from mt salesmen.

I know --
I was the sales mgr for the best HAAS outfit in the world - Hitec Spain - nr 1 dealer in the world 2013 iirc.
Trained all the sales guys -- and 9 years later they still work there and do really well.

I sold, and they probably still sell, the Haas machines as "good enough".
They aint a hermle by no means, but whatever you make it´s probably better than you need.
We never ever lied to a client and always had a 100% customer satisfaction rate.
I no-quoted lots of deals, and sent some to Mori.

There are things I can think of where a Mori lathe or a hermle would be my choice.
Nuclear or jet turbine blades, valves for power stations, things like that.
Impellers for the jet fuel pump at 200.000 rpm for spacex.
Where failure has 100m++ costs and 10x in liability and parts costs mean nothing.

As You said it would be wonderful to not have to sidestep the shortcomings --
but the 12.000$/month for a long time that will bankrupt your business if you lose a client or 2 ..

Of course some clients (mine) had 1.2M€ prepayments in the bank and a huge order backlog at 90%+ marginal profit.
So they could afford the hermle out of ready cash.
Not at all sure -- I might have bought the hermle, myself.


Ponying up for a proper machine as you've obviously done is a no brainer...the question for the OP is: how good does it have to be for his kind of work.

If he needs a Hermle and buys a Haas, he's in exactly the same boat you were in when you only had the router...yeah he can do some of it (as I do) if he's really experienced at sidestepping the shortcomings, but oh how wonderfully nice it would be not to have to even think about overcoming them, and how nice it would be to be able to take on the hard, tight tolerance stuff too.
 
There's no $3000 Haas payment that can be made to be competitive with what a premium controller/builder can do that would require $6k-8k payment.
Price is what you pay, values is what you get.
Paying a guy $5000 per month to dic around with fixture offsets due to 5 axis inaccuracies is certainly part of the puzzle that doesn't show up in the 'build a quote's.
 
Hi Hanermo:
You make some excellent points and I have made the calculus for myself exactly as you describe.

But, and this is a BIG but, I have to turn down work because my equipment is not up to it.
I know when that is, as do many or most experienced businessmen.

Taking on a job you cannot deliver on is as sure a way to kill your business as getting killed with payments you cannot make.

Notice I never said, "Buy the best...buy ONLY the best...never compromise, never surrender"...what I said (or at least what I hoped to say) was, "match the machine to the jobs you intend to bid on".
This is a basic rule that many forget, either because they get stars in their eyes or because they're cheapass dumbasses, or just that they're young and naive.

Another point that many youngsters and crusty gomers seem unaware of, is that when you pay the premium coin for the premium kit, you also have to lay out for everything else that will enable you to do the best the machine can do.

How many shops have you been in where it's sweltering, the employees are wilting and there's a high end mill-turn baking in the heat, making out-of-spec parts all day.
It's common enough still that it's noteworthy, even if thankfully it seems to be slowly dying out.

So yeah...a Haas is a great choice for Haas type parts, so if that's what you mostly intend to pursue, I agree with you completely.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 
Too bad I'm old and can't justify getting a good one anymore.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
Marcus, I read thoroughly every word you type here on this forum as I have tremendous respect for you and what you do. This however I can't let slide. My grandfather has "accidently" made it to 90, he survived polio as a child and has a tiny leg to this day but he's still around rebuilding old cars. You can't take it with you and if you leave behind to much for your descendants you'll only stifle their ambition. I don't know if you'd rather a boat or something but I dislike the idea that we age out of deserving the fruits of our labor. Respectfully.
 








 
Back
Top