[astromachinist]

Hi Folks,
New here, just starting out on a journey of CNC machining to build cool toys for my other hobby: astrophotography.

For the past couple of months, I have been researching various mini mills that I intended to eventually convert to CNC. Eventually it led me to consider the Tormach family of CNC mills. Unfortunately owing to space considerations, I have to rule out the PCNC 1100 and the PCNC 700.

The newly announced PCNC 440 seems like just the right machine for my purposes except for one consideration. The rolled lead screws are of a lower grade and consequently less precise than the ground ones used in the PCNC 770 and the PCNC 1100 mills. Per conversation with Tormach representative, they are accurate to around 1.8 thousands over the X travel.

I called Tormach to enquire why this decision was made to use a lower grade component in the 440, the answer they gave was that this was a purely cost driven decision.

Speaking for myself only, I would be willing to pay a bit extra for ground ball screws. Buying a mill of this level is already quite expensive with all the tooling. I would ideally have liked to get the 770 but it is too big for my workshop.

The good news is that it appears that Tormach takes this sort of customer input seriously.

If there is anyone else who is on the fence regarding the 440 and cannot justify getting the 770 for space considerations, it may be a good idea to call Tormach with your input.

Best Regards,

[wtopace]

I believe rolled and ground ballscrews come in lots of grades. Rather than jumping to conclusions, it would be nice if NYCNC or someone else that already has their hands on a 440 could do some real-world accuracy tests with the machine. The 3D machined parts he did last week looked pretty darn good to me. With a 0.001" stepover, if the precision wasn't there, I would expect to see a lot more "waterlines" around the part, rather than the smooth shiny surface he had.

Overall, it would be great to see some real data on the accuracy.

[Spinnetti]

Most people way over specify their accuracy requirements and there are so many considerations to accuracy that this one number can hardly decide it all. Thermal stability, fixturing, tooling, tooling hangout, speeds and feeds, blah,blah,blah.... You can also do things like ball screw mapping (could with Mach anyway) to improve accuracy. What will you make with it? what are the actual accuracy & precision requirements (which are not the same thing)? If you need true high precision, likely none of this class of machine will do anyway.

[astromachinist]

Thanks for your responses.
The Tormach PCNC 440 uses a C7 grade rolled screw. You may see the 440 specs from the Tormach site.

The PCNC 770 and 100 use the P4 grade ground screw.

Here is a useful document that describes the expected accuracy of the various grades of rolled screws.

http://www.cbmind.com/linear/thk/pdf...ifications.pdf

Grades C0 through C5 are considered precision grades. They have both travel distance error and fluctuation error fully specified. C7 grade only specifies travel distance error and not fluctuation.

One of the applications that I am considering is thread milling a worm. Worm accuracy needs to be specified in arc secs of periodic error which is a measure of fluctuation. That would not be possible to do with the c7 grade lead screw.

As another reference, consider Tormach's own document.

Engineering Documents - PCNC 1100 | Tormach Inc. providers of personal small CNC machines, CNC tooling, and many more CNC items.

Here are some quotes from this document:

Ballscrews and the associated ball nuts are available in a variety of precision levels. The screw portion is offered in two basic dimensions: the grade and the application. Applications are marked as P, T, or C:

P grade is generally taken to mean position or precision. P grades are intended for high precision work, such as that of CNC machines.
T or C grades are referred as transport grades, designed for utility applications such as lifting, pressing, or other non-precision applications.


The grades refer to basic accuracy of motion. With P applications, both incremental and cumulative accuracy are important. With T/C applications, only incremental accuracy is specified. As an examples:

P3 allows 12 um error within any 300 mm. In imperial units, this is 0.0005" per foot. Cumulative error at 900 mm is limited to 21 um (0.0008" at 35").
C3 allows 12 um error within any 300 mm (0.0004" per foot). Cumulative error is not specified, thus allowing as much as 36 um error at 900 mm.
P4 allows 16 um error within any 300 mm. In imperial units, this is 0.0006" per foot. Cumulative error at 900 mm is limited to 22 um (0.0008‖ at 35").
C7 allows 52 um error within any 300 mm (0.0020" per foot). Cumulative error is not specified.


The C7 grade is relatively low in cost and often used in retrofits or conversions of manual machines to CNC. The low tolerance allows the screws to be manufactured through a rolling rather than grinding process.



Tormach clearly believes that the C7 is a lower grade of lead screw and the only reason to use it is cost. Given this, I am simply exploring to see if there is a demand for a small envelope but high accuracy mill. It would be quite easy for Tormach to offer it if there was sufficient demand at a somewhat higher price.

[astromachinist]

It also seems to me that Tormach wanted to create a new machine with the PCNC 440 to satisfy two criteria 1. A smaller machine for those with limited space 2. A less expensive machine. It was likely not possible to offer a significantly less expensive machine than the 770 even at a smaller size. I guess this forced them to make the cost tradeoffs they did: using lower grade rolled screws, brushless DC motors instead of induction motors etc.

[shred]

Tormach is all about engineering to the market sweet spot. If the average 440 user can get .002" accuracy at the end of the day, they'll probably be happy campers.

[kstrauss]

For your application of milling a worm screw I believe that the important issue is periodic error. It is not obvious that an error of 0.002 or 0.0004 or whatever over a foot says much about the error over shorter distances. That is, it seems a stretch to conclude that the error over an inch is 1/12 of the error over 12 inches. Or perhaps I misunderstand your concern.

[SCzEngrgGroup]

There are SOOOOOO many things other than the quality of the ballscrew (not leadscrew....) that define the actual, real-world accuracy of any CNC machine. Picking out a single component, and making projections of the entire machines accuracy based on that borders on ridiculous. Have you considered that even thermal expansion is often as large a factor as ballscrew accuracy? Or, even more important, the machines ridigity? Backlash/stiction will also typically be major contributors to error.

The reality is, expecting to get TRUE accuracy and precision of greater than, at best, +/-0.002" from ANY machine in this class is wishful thinking. You may do it on a given day, but to do it repeatably, day in, day out, you better get ready to spend a LOT more money.

Besides, it you're making worm gears, you want a lathe not a mill. A mill will limit you to making only very short worm gears, and seriously limit your choices in the tooth profile, unless you're equipped to grind custom cutting tools...

Regards,
Ray L.

[astromachinist]

For your application of milling a worm screw I believe that the important issue is periodic error. It is not obvious that an error of 0.002 or 0.0004 or whatever over a foot says much about the error over shorter distances. That is, it seems a stretch to conclude that the error over an inch is 1/12 of the error over 12 inches. Or perhaps I misunderstand your concern.

Yes, exactly. The issue is that C7 grade rolled screws arent considered precision screws and dont have a periodic error specification unlike the P4 screws used in their other mills

[astromachinist]

There are SOOOOOO many things other than the quality of the ballscrew (not leadscrew....) that define the actual, real-world accuracy of any CNC machine. Picking out a single component, and making projections of the entire machines accuracy based on that borders on ridiculous. Have you considered that even thermal expansion is often as large a factor as ballscrew accuracy? Or, even more important, the machines ridigity? Backlash/stiction will also typically be major contributors to error.

The reality is, expecting to get TRUE accuracy and precision of greater than, at best, +/-0.002" from ANY machine in this class is wishful thinking. You may do it on a given day, but to do it repeatably, day in, day out, you better get ready to spend a LOT more money.

Besides, it you're making worm gears, you want a lathe not a mill. A mill will limit you to making only very short worm gears, and seriously limit your choices in the tooth profile, unless you're equipped to grind custom cutting tools...

Regards,
Ray L.

Hi Ray, Thanks for your response.
Could you explain why Tormach chose to use P4 grade ball screws in their other mills and make a big deal about it (see my earlier post) if you believe they have negligible impact? Not trying to start an argument but was interested.

BTW, I spoke to Tormach already. They did state that their larger mills are more accurate for this very reason.

[adamvs]

I think Tormach made a wise business decision doing this to meet a price point and fit it into their product line. They wanted a lower barrier to entry into the CNC game; face it, this works for a lot of people. If you have limited space, or are a school wanting to teach the principles of CNC, or want to get your feet wet in the game, this works great. Also they have somewhat of a razor/razor blade business model, knowing no matter what machine you buy at what level of precision, you're going to be spending a lot on tooling and wanting more. It's easy to double the cost of the machine on all the goodies they offer. The smart thing about the way they implemented this is, if later you upgrade to a better machine, almost all, if not all, of the tooling still applies.

A thought about thermal expansion. Iron at roomish temps changes about 6 millionths per inch per degree F ( think that's right ). The ball screw on the 1100 is what, about 2 feet long. So if your shop changes temperature over the day by 20 degrees F that's .000006 * 20 * 24 = .0029 .. almost 3 thou.

[Gilmax]

To me, the accuracy provided by the C7 Class screws is not the issue at all on the 440. I'm concerned about preload, backlash and longevity, which are generally better for higher grade ground screws. From a business standpoint, Tormach probably made the right decision. I personally would pay extra for higher grade screws though.

[astromachinist]

Gilmax, Adam,
Thanks for your comments.
I fully agree that Tormach has a winner on their hands. I am however a bit befuddled by their reasoning to use some lower grade components after they went all out on their bigger 770 and 1100 machines.
Perhaps I am obsessing a bit much about the accuracy, but then again, I would personally pay more for higher grade components.

[Spinnetti]

Well, you still can. Buy the machine, order up the bits, modify to suit and Bob's yer uncle.

[Moggot]

Gilmax, Adam,
Thanks for your comments.
I fully agree that Tormach has a winner on their hands. I am however a bit befuddled by their reasoning to use some lower grade components after they went all out on their bigger 770 and 1100 machines.
Perhaps I am obsessing a bit much about the accuracy, but then again, I would personally pay more for higher grade components.

I suspect that if you're new to machining the grade of ballscrew is not going to be the primary issue with machine accuracy. There are many other things working against a small machine like the 440 that'll come into play before the screws do. As the 440 is marketed towards hobbyist and 'makers' (whatever that term actually means) I'd wager that Tormach decided that higher grade screws wouldn't make a meaningful difference for most users.

If you're worried about the accuracy difference between the C and P graded screws you can probably justify the added cost and size of the 1100. A machine that fits in the size and budget envelope of the 440 is always going to be a compromised tool, just as the 770 and 1100 are in their own ways.

[astromachinist]

As the 440 is marketed towards hobbyist and 'makers' (whatever that term actually means) I'd wager that Tormach decided that higher grade screws wouldn't make a meaningful difference for most users.

If you're worried about the accuracy difference between the C and P graded screws you can probably justify the added cost and size of the 1100. A machine that fits in the size and budget envelope of the 440 is always going to be a compromised tool, just as the 770 and 1100 are in their own ways.

I am not sure I buy this argument. The 440 weighs 450 lbs with a work envelope of 10 X6.25X10 so around 0.70 lbs/cubic inch.

The 770 weighs 662 lbs with work envelope of 14X7.5X13.25 = 0.475 lbs/cu inch.

So basically the 440 is MORE heavy than the 770 for each cu inch of work envelope making it potentially even more rigid than the 770 when both machines are pushed to their limits, assuming the weight is in the right place. There is no reason why it will need to be a compromised tool.

There are lots of folks: watchmakers, jewelers etc. that will appreciate higher precision and dont need a large machine such as the 770, let alone the 1100.

[LeeWay]

I don't think the screws matter as much for MOST of the target market. There is absolutely one thing for sure, you can't please everyone. My biggest beef with it is the enclosure design and lack of flood cooling design. That means that all chips stay in the pan until you pick them up. Not wash them down. No provisions for a tank below. Okay, maybe it's geared more toward mist coolant. That still doesn't get rid of the chips in any kind of modern way. That system is too much like my home made mill. Far too much wasted time doing clean outs. If you neglect it, then you will start to see deterioration in the tray, so you have to clean it out. That tells me everything I need to know about the 440. That said, I do have another stand and tray here that ain't doing anything at the moment. It could use something to do. Like support a bare bones 440 without a stand or tray.

In fact, most of the items in the basic package I would not need. They could get it down to under $6000 if they offered a real bare bones mill with just the controller. Closer to $5000. Then I might bite.
That being said, they had no trouble selling the first lot of machines and are now into the second lot of preorders, so if you want one, get in line.

[Gilmax]

Well they will eventually sell it for 5000 bare bones. You can build a very similar controller they sell for 300-400 bucks with a mesa card. I personally would want the stand and enclosure. Also if you go look in their page about availability of accessories, it says flood coolant. Not sure if that's referring to the fog buster or not. As for the C7 screws, like I said before the accuracy is not a concern for me, if they offered the same low backlash and life as their higher grade counterparts but at less accuracy I would be quite satisfied with that, but I suspect that's not the case. I'm considering buying one in the future for second ops and small parts. Some suggest that the "maker" crowd don't care about rolled screws, but they simply might not know any better (no offence). I would rather see a compromise on upgrades rather the core product. The main reason I bought my 1100 over a Novakon for example was first the support, second the perceived quality. Either way I will be eagerly be waiting for reports on the 440.

[SCzEngrgGroup]

The 440 weighs 450 lbs with a work envelope of 10 X6.25X10 so around 0.70 lbs/cubic inch.

The 770 weighs 662 lbs with work envelope of 14X7.5X13.25 = 0.475 lbs/cu inch.

Wow! That is the most nonsensical statistic I've ever seen applied to a CNC machine. There is absolutely ZERO meaningful correlation between the weight of a machine and its work envelope, nor can they, individually or in combination, be used to tell you anything of any value whatsoever about the quality, accuracy, or precision of the machine.

Do you also calculate the volume to weight ratio when shopping for cars? Or perhaps power to volume?

Regards,
Ray L.

[astromachinist]

Wow! That is the most nonsensical statistic I've ever seen applied to a CNC machine. There is absolutely ZERO meaningful correlation between the weight of a machine and its work envelope, nor can they, individually or in combination, be used to tell you anything of any value whatsoever about the quality, accuracy, or precision of the machine.

Do you also calculate the volume to weight ratio when shopping for cars? Or perhaps power to volume?

Regards,
Ray L.

Really? Ray?

Please see this link.

Engineering Documents - PCNC 1100 | Tormach Inc. providers of personal small CNC machines, CNC tooling, and many more CNC items.

When comparing different machines, you can get a rough idea of rigidity by comparing the volume of the working envelope (X, Y, and Z travel) and dividing by the machine weight. The result is cubic inches per lb, where a larger number indicates a less rigid machine. This comparison cannot be fairly made between a bench top machine and a floor standing machine, where much of the weight is in the base does not contribute to rigidity. A fair comparison also assumes the machine designer has put the iron in the right place.


It is quite easy to arrive at this figure of merit if you knew that weight goes up as the volume of a part. But that is something only taught in 5th grade