[peteeng]

Hi Ralph - Agreed you can't please everyone. That's been a common thread in developing the Tetra Motion biz. So I try to please many people so as said, there will be a non rigid tapping version (unless 750W servo is very good price which is unlikely), 750W & 1kw servo and the 1.5kW water cooled spindle for high speed freaks. Will a 750W servo with 1:2 pulley drill a 1/2" hole in steel? I use UCCNC so that's my starting position. I'm familiar with stepper set ups so that's easy I usually source steppers and drivers from stepper on line. One client is going with Buildbotics as its a one box affair. Everything is in pencil at the Mo... but a picture is emerging Peter Oh yes everyone wants a Datron... Perhaps I should do a Datson version?? My family used to have a Datsun car...Peter

[peteeng]

Morning All - I'm looking at reflexing the Y motors into the base using belts and pulleys. Are these too compliant for a mill? Is a stiff short belt stiff enough to maintain accuracy vs a coupling? I'm looking for coupling compliance figures at the moment so I can compare. Peter

[klaas123]

A stiff short belt will be very accurate and will stop any motor vibration to the axis, giving a better finish on your work piece.

[peteeng]

Hi Klaas - Vibration isolation is a good thing. BST have HTD belts and pulleys so will look at those. I usually use AT belts from local but they are twice the cost... Cheers Peter s

[peteeng]

Hi All - I found a Ruland data sheet that provides torsional stiffness of their jaw couplings. If the urethane jaw is the same hardness I assume it's the same stiffness as others. If 1Nm is applied to a "held" load with a 10mm pitch screw (and the screw is rigid) the coupling effectively moves the system 0.0164mm. Or you can envisage this the opposite way if you like. If I use a 16mm wide HDT5 belt with 30mm pulleys (100mm centres) and apply 1Nm it stretches 0.0129mm so a little bit less which is good... So happy at the moment to use a belt drive coupling. I'd like to use 25mm belts but this depends on if the pulley can sit on the motor shaft...

A HTD8 belt is stiffer then a HTD5 and the AT5 is same as HTD5. Seems size to size the HTD and AT series are the same stiffness. I have a Metrol manual with all the common belt data.... Keep at it out there Peter

[spumco]

Belts - you aren't going to need 8mm pitch belts for those Y-drives. 5mm is fine. The difference between HTD & AT is that AT tooth profile has lower backlash, and HTD are cheaper and more widely available. HTD was designed as a power drive system, not a positioning system.

More stiffness comes from wider belts for a given tooth profile. I'd suggest going with 15mm wide HTD5m belts/pulleys for the first test build. If stiffness and backlash are not acceptable under load it's extremely easy to change to 25mm wide, or try an AT (or GT2/GT3) tooth profile. Just make sure the test machine has long enough ballscrew ends to accept wider pulleys in the event you need them.

Spindle motor - I don't know if a 750w driven 1:2 would be able to drill steel. My feeds & speeds calculator estimates 500RPM with a cobalt drill in 1018 steel (250rpm at motor). Cut requiresbetween 4 and 7 ft-lbs of torque (~0.5hp). Check the torque curve of the DMM servo and see if this is within the continuous torque band. If it's between the continuous and peak lines, the motor might handle a shallow hole, but a deeper one would overheat the motor before the hole is completed.

[peteeng]

Hi Spumco - I agree with your comments. For positioning, ATL belts are even better then AT or HTD but are 2x cost. I think I'll go AT to start with and use the 5mm. I think 25mm pulleys will fit the Nema23 motors so that's the starting point. 16mm is the std width here and my first router used AT16mm belts for XYZ axis and was very good. I like belts they are under-rated I think.

I'll check the torque curve of the spindle. Ta Peter

Hi Ralph - The 1kW motor can produce up to 12Nm intermittently. They do not define intermittent. 7ft.lbs is 9.5Nm so this is possible. I'll ask Stephen what the def of intermittent is in terms of duty cycle or time based. The 750W can produce 7Nm max intermittent so is under the 9.5Nm. So unlikely to be able to drill 1/2" as you say. Is there are drilling rate? Would not power needed also depend on feed speed? I'll have to clock my little cheap drill press with a 1/2" drill to see what its doing. Thanks again Peter

Its power day today The lights and outlets are in and cabled! They have dug the trench and today they connect the juice, excellent.

Just got word from DMM - They have a time overload chart. At 200% overload it will last 10secs. So you can't drill a hole in 10secs. So the 750W at 2.4/7.0NM means it can pull 4.8Nm for 10 secs. So 1/2" is out unless the feed speed is important for torque. The 1kW 4/12Nm means it can pull 8Nm for 10secs so again won't get there. Will need some real numbers I think.

Also I looked at asian servos and they have similar series but frame and holes are different.

Asian 80mm frame PCD90mm 4x6mm holes 400W, 750W and 1000W
Asian 90mm frame 750W 3000rpm 750W 2000rpm more torque
DMM 86.6mm frame PCD100mm 4x6mm holes 750W and 1000W, so two motor brackets will be needed...

[peteeng]

Hi All - This discussion is good and its pushing me back to the 1kW size. Better get the No9 model running to see how it fits.... Peter

Just to take the overload data a bit further at 150% the motor overloads in 12sec; 200% overload the motor thermally overloads in 10sec, at 300% or the nominal max it thermally overloads in <0.5secs nearly instant so don't be hopeful about the intermittent values, they are for true overloads! Peter

see chart supplied by DMM below

[spumco]

Peter,

The cut parameters were at 0.0085IPR, or about 4IPM feed. If the IPR is reduced to 0.006IPM then feed drops to 3.08IPM. Torque required is also reduced to 3.3ft-lbs.

So a 1" deep hole will take 20 seconds at 0.006IPR and 500RPM. If peck (or deep-hole full retract) drilling is done, there there will be a short period of time for the motor to catch up and cool off slightly. Chips will be stringy, but it'll work.

-Ralph

[peteeng]

Hi Ralph - So if we reduce the feed the torque drops to 3.3ft.lbs which is 4Nm which is in the continuous rating of the 1kW. So more evidence for the 1kW. A year ago when I first started this a friend said 1kW is the go so maybe just fly with that.....Use a 1.5:1 maybe thus limit top speed to 4500rpm nominal 7500rpm tops. Owner can fiddle with ratios to their hearts content....Peter

asian 750W 80mm $420AUD 1000W $450AUD so no biggy really. The 110mm 1.8kW (6Nm 3000rpm) is cheaper $381AUD Hmmmm so it cleans up the 1/2" hole easily.... The 110mm goes 0.6kW to 1.8kW so good range there...

[spumco]

I think the 1kW at 1:1.5 would be a good compromise. As I mentioned earlier most of the BT30 spindles I've seen have an 8kRPM bearing set in them and the costs rapidly go up when you want 10k+ with cermic bearings. The 7500RPM top end would keep the spindle from being contantly run at its max rating.

There are a number of folks (on CNCZone and other places) who have fitted a DMM servo for thier spindle motor with various drive ratios. Suggest browsing around and asking questions from folks with real-world experience. I may have been under- (or over) estimating how well they work as a spindle motor.

[peteeng]

Hi Ralph - Inertia matching is important. Even the strongest athletes have trouble with Sumo wrestlers... so more evidence for 1kW plus.... Done 110mm frame it is.. My workshop is now lit and powered YEH>>>>>start moving stuff in. Peter

[pippin88]

Peter, its hard to find what intermittent means on a lot of servos.

The Delta B3 manual has a chart (takes some digging to find it). It's not long. Not 20 seconds I don't think.

More like the time taken for acceleration rather than any form of actual cut.

[peteeng]

Hi Pippin - Yes the DMM chart is logarithmic so in a blink of the eye its cooked!! So best to keep to the 100% band and don't have hopes that the overload will be OK. Peter

Let there be light!!

[klaas123]

Nice shop Peter!

[peteeng]

Good Morning all - - I'm off to work on YaG (Nee Brevis-HD router) and I have been working on Milli's No9 base. It has to be very stiff. The 30GPa Tetrium-S Material needs to be very thick to achieve its required rigidity so I have been playing with the idea of an Al/Tetrium/Al honeycomb sandwich base. I want the honeycomb to be parametric and this is an interesting problem. The current base is 150mm thick solid Tetrium and weighs 780kg. If I do it in FG it weighs 351kg but I'd like lighter say 200kg. So some sort of egg crating in the bottom would be good or the sandwich will be stiffer. Probably set it up as a generative model next time around as well... Have a great day all you out there. Peter

[maxspongebob]

I love your new shop.

What do you think of starting with a granite surface plate for the base?

[peteeng]

Hi Max - I think granite is a great material but I can't cast it. There is a stonemason 5mins from me and he has a wet cnc mill for doing gravestones, counter tops etc. For none appearance grade granite or synthetic granite the cut prices seemed very fair. Nearly went down that path last machine. But putting granite machines in the post is a bit of a problem...

The current base is 150mm thick Tetrium-S or infused FG. If I use 10mm thick Al plate each side I could reduce this to 130mm thick. If I use high density foam and 10mm Al plates it would be 165mm thick. I wanted to get to 100-120mm but can't see how at the moment unless I use steel... YUK. So I'll work on my honeycomb for a little bit then start building Milli No9. Got the last pieces of YaG's machiune base built today so now onto reviewing its metal parts and waiting for word from BST about my parts order.... Peter

[maxspongebob]

Hi Max - I think granite is a great material but I can't cast it. There is a stonemason 5mins from me and he has a wet cnc mill for doing gravestones, counter tops etc. For none appearance grade granite or synthetic granite the cut prices seemed very fair. Nearly went down that path last machine. But putting granite machines in the post is a bit of a problem...

The current base is 150mm thick Tetrium-S or infused FG. If I use 10mm thick Al plate each side I could reduce this to 130mm thick. If I use high density foam and 10mm Al plates it would be 165mm thick. I wanted to get to 100-120mm but can't see how at the moment unless I use steel... YUK. So I'll work on my honeycomb for a little bit then start building Milli No9. Got the last pieces of YaG's machiune base built today so now onto reviewing its metal parts and waiting for word from BST about my parts order.... Peter

Understood. I was just thinking that, since this is to be a kit, you could have the purchaser source their own base. Also since the envelope X and Y are primarily defined by the size of the base, a purchaser could get a stone that is a bit smaller or larger. For that matter you could have a series of kits that are based on standard surface plate sizes.

Would it be an advantage to start with a base that is absolutely flat and requires little machining?

Just me thinking out loud.

[peteeng]

Hi Max - The "kit" market is in two halves. Halve A wants to build as much as possible and is majority $$$ driven so you would think that your suggestion would work. But it doesn't. When I present a design, parts, a shopping list and instructions that sort of Maker looks at it and thinks they can do better. Maybe they can. But they don't see the 100's of hours that go into establishing such a shortlist and sorting a design. Then there is side B that want a machine, they want it now and are happy to pay for it. They have done their homework and if the $$$ are about right, the machine is about right and its there, they buy. Party A haggles and talks and gets educated and gathers more info then they go away and buy something else, usually a chinese machine (that's in routers). So for the mill I've decided to make a very good little basic mill, take a step forward and make it out of composites like all the top builders are tending to do now and see what happens. I have a few enquiries for mills, a bit of a spread of requirements but they can be simmered down to a basic spec. Having my workshop space now will be helpful in achieving all of that... Peter