Thanks all! My daughter Oli is perfect.Kellie's in great shape too. Be back asap...

Awwwww, what a lil angel
Congratulations to the Sigler fam!

congratulations she looks just like pricess comanawannaholdu from the snugwali tribe come down from the full pink moon., when the green grass comes pushing up out of the earth and the new eggs laid when the shad run up river.
What a blessing congratulations to you both.

What a cutie! She looks only a little bigger than that mega-servo!

Congratulations!

Phew!

Thanks guys...man what a couple days....my eyelids are made of lead, but I'm very happy.

Best estimate I'll be back in full swing Wednesday, with more pics and some tests results late in the week. RoGR is exciting stuff...way ahead of the curve, and so many possibilities! I love thinking about it, can you tell?

Cheers!

bc

Good God! Man!
Take a breather for a few! You've earned it!

He just doesnt quit...

SJ

Sincere congrats to you and Kelly, G, on the new addition to the fam.

Yea, that's what I was doing when I posted!

Congrats Brain.

We are all very excited for you.(the baby man the baby)

Thanks! Phew! It's been busy around the old household lately...

Termites swarmed the house requiring me to "delete" an entire wall of one bedroom. Now I have a giant hole in the house awaiting rebuilding...great timing with a house full of relatives and a newborn! To boot, I picked up a nasty bug from the hospital.

That said; I've still managed to work on RoGR. The servo motor kits should go up for sale this week when the encoders get here. I'm awaiting one more sample for the drivers and then I'll decide which driver is best, or whether we'll manufacture our own design. Chinese ballscrews are being sampled for RoGR as well. They will be delivered with the bearing kits. The micRo design is pretty much finished too...a few small details remain.

Will micRo just need different lengths of bearings and screws? I am not having second thoughts about RoGR, but rather that it might be good for me to build smaller scale initially as far as the home front is concerned. I want to have all the parts for RoGR, but I wouldn't mind a few more details on the overall construction requirements for micRo to see how much of one I can use for the other.

I hope the newborn is being good for you ! You know we're all interested in seeing progress on RoGR, but the little one's are what "counts"... well; them and beer ! (oops... and the wife. of course. ) I like demo, but under your circumstances, probably not what you were hoping for.

Cheers !

micRo is tiny, but large enough to machine the largest aluminum piece on RoGR (10" x 12"). Originally I had wanted to use the parts from micRo "upstream" to RoGR, but I realized the utility of a much smaller "PCB" machine; for model making, rapid prototyping of small parts etc.



Cheers!

Ah, so that has changed; No modularity between the two any longer, so in essence the micRo is a totally separate design.

I'm actually feeling the magnetic pull towards this project but I'm concern that I might be in over my head. If I understand the entire project will or may be broken down in several kits comprising of bearings, motors (servos or steppers), gantry and then software? If you don't my asking G, but what kind of ballpark figure, roughly, would I be looking at the end upon completion? BTW, I tried purchasing the CNC book from store but they were out.

This is the first I heard of the micRO being a separate project. Is there a thread on that topic?

Indeed. RoGR evolved to large servos and has the frame, rails and power to be a "big boy". micRo kept getting bigger to use RoGR pieces until micRo was nearly as big as RoGR. I still want micRo to be useful to RoGR, so I made the cutting area large enough to make all of the machined pieces of RoGR. micRo uses these tiny NEMA 17 steppers:

Click to view attachment

micRo is based on high positional accuracy and excellent holding torque, but won't go fast. Perfect for small-scale machining and PCB routing. The entire micRo kit will retail for $250-300, which includes everything but your computer.



The servo/ballscrew version of RoGR is creeping up to ~$1600 and the stepper version ~$1200. At the end, all of the major pieces will be available as a kit, much like the DIY projector, you can buy as little or as much as you want.



I have been trying to start a thread for a week now; been a busy time for me! Here's to finding time this week!

Brain...Question...do you think your current robot can cut brass?

Sure.

brainchild, you said that one advantage of servos is that you can use ball screws. From the tiny bit of info I've seen it seems like the big difference between ball screws and normal screws (ACME?) is that ball screws turn much more easily and also wear less. Would it be possible to use ball screws with steppers? Would it be a waste in some way?

I tell ya: you drive a hard bargain! $200 for those servos sounds like a phenomenal deal. I'm "this close" to biting, but I think I better set my sights a little lower for my first machine...

If you were thinking of the micRo to start with, then given that it will be designed for the wee-bitty NEMA 17 motors, I'd imagine you'd have to do some shoehorning to make the servos fit.

Heheh, the servos are easily 10x the size of a NEMA 17.

BTW: Over time, it has become obvious to me that servos are wasteful for micro-machine sizes, since positional accuracy is more easily obtained in small spaces with steppers/step-encoders, rather than the "dancing~breathing" analog-loop servos. The servos shine on a RoGR-sized machine since they offer better power, accuracy and speed over longer distances with 100% duty cycles....they can stretch their servo-legs as it were hehe. The servos can move RoGR four times faster (X,Y) than NEMA 34s and track their position (well, they can stop at 128 pulse count error..not quite CNC GPS yet.. ). Note RoGR uses a stepper for Z; a servo would be wasteful and finicky here.

It is a complex subject involving "stiction", dynamic load and torque. Since a lead screw is so inefficient, it also resists being "back-turned" by opposing force. This means the motor only needs enough power to screw-drive the tool to the position, plus a little more, but that's it...the screw won't move when a "back-load" is applied. A lead screw is regarded as about 20% "efficient".

A ball screw has about 98% efficiency and can "back drive" easily due to its low "stiction". If a stepper is paired with a ball screw, the maximum current/torque (holding power) a stepper can make is within just one "field" position between its windings, and this "fractional power" is defeated easily by a ball-screw; whereas a servo/analog loop can increase its power to "hold" the position reported by the encoder...to the point of death, if so programmed!

Sweet, finished motor testing yesterday. We tested the motors to 93VDC without issues, and without capacitance to test for "singing". These motors kick-ass! Smooth like buttah and power to spare, plus, closed loop! It's alive!!!!

Anyway, the motors are now labeled and rated:

Click to view attachment

Motor kits will go on sale tomorrow!

OK, motor kits did not go on sale yesterday... I got lots going on here that is eating time worse than the Langoliers: but I got another 10 hours of amp/motor/machine-controller testing/programming done today! DIY will not be undone!

Phew, if I can stay awake long enough I'll show some of the pics I've taken during the process. Several encoders have been chosen for testing, as different situations require more or less resolution from the machine. For the basic RoGR, a balance of fast speed and reasonably tight accuracy was chosen, with a tolerance of .001 delivered from a 1200ppr encoder. A 128 pulse-error-fault (following error) is common on the low-cost servo amps, meaning a missed count of 128 pulses stops the drive. When large loads and motors are driven at high velocity and are then asked to reverse, there can be "slippage" if torque/accel variants are not programmed correctly. This results in "overshoot" (it's alive!!!!!). A faster system has a greater likelihood of error, so a large pulse-count-encoded machine must by necessity either a: go slower or b: be made more powerful; neither of which is desired from the (broke-ass) machine designer point of view.

The pic below shows a partially installed quad-encoder. If you choose your own encoders, be very careful to calculate every aspect of the speed, resolution, power and torque requirements to make an informed choice. Many believe very high resolution encoders have more "refinement", but unless you have a 10-ton knee-mill, you ain't ever gonna need or see .0001! What most of these people find out, after having invested a small fortune, is that their machine moves slower than plate-tectonics, well...that is if they don't want error-faults every second.

Click to view attachment

Here is the finished encoder; it is a delicate thing. The design I am working toward is having the amp, encoder and a 120vdc bridge-rectifier housed on the back.

Click to view attachment

Several commercial "amps" have been chosen for testing. The amp provides both the counting and PWM driving force to the servo. An amp of our design is also being tested, as I seek to simplify the process further by looking for the modular approaches. So much of closed-loop robotics is "black-magic", but it shouldn't be!

Here is a Gecko 320. This is a very fine amp with some quirks, but the company is highly regarded. The small size is phenomenal for the power it can switch...

Click to view attachment

Here is the "death motor" workstation. You may notice it is set up on another CNC-robot that we are trying to "frighten" into behaving...This poor motor has been subjected to dropping, repeated disassembly, running for long times at 4x the rated RPM, running at very high voltages, torque stalling at high current, running off a stick welder, being made to reverse direction at maximum accel for long periods and being run off of AC voltage: OW! It works fine, though I have driven it off the platform several times with its own torque, and purposefully made it oscillate wildly for long periods (I do this to women too ;p).

Click to view attachment

Wow. This entire thread is amazing - being able to actually watch the first RoGR come together is invaluable information.

I'm drooling over this build and am trying to get a general concept of how much its going to cost to build one - I know you've said in the past that you're shooting for between $1200 and $1600 but I've got some questions - I'm not about to purchase these items right now but am trying to get a handle on them.

In the list of 6061 aluminum parts you've got (in this post)

• Are they extruded, plate or channel aluminum (or some mix, as I hope/suspect).
  
• What are the .75" x 8" x 12" pieces for?

On the servos, encoders, drivers, amps, etc.

• I got the impression from this post that the servo kit would include encoders, is that correct or did I misread it? I only ask primarily because your discussion of choosing an encoder in post #325.
  
  
• Also, is an amp the same thing as a driver? Looking at Gecko's website, it appears to be, but I just wanted to be sure.
  
• Given that we'd need 4 G320 drivers, would something like the G-Rex G100 (http://www.geckodrive.com/product.cfm?pid=19) be more suitable? And perhaps more cost effective?

Sorry for the numerous questions and thanks again for your dedication! Its amazing to see the decision making put into every aspect of the design.

Thanks! It's been a good experience for me. RoGR-type technology is entering the public mind, very much like the internet of the past...



Well, that depends on where you shop, but we spec'd only "flat bar" alum (yes it's somewhat huge flat bar!) because the price is a fraction of "plate", owing to the ease of cutting "bar" over "plate". I can currently recommend http://discountsteel.com as a good supplier for nearly the best price.


AGH, those pieces are now .5" x 10" x 12"...bad Brainchild!



The encoder is included, and it seems as if I can include the power supplies (1 per motor) at the same price as well if the tests show reliability. One of the problems with this (and why the kits aren't on sale yet) is that I want a higher voltage than is commonly made with a switch-mode power supply (48VDC). The difference in price for the higher voltage analog supplies (>75VDC) is significant, so I've been resisting the spec until I have a good solution. :/




You'd think, right?

A servo "amp" is like any other analog power amp, and while terms may be mixed within the industry, a driver is not an amp, but rather the device that can deliver the "pre-amp" processing which occurs post-pc g-code interpretation over the parallel port. If there was a computer port that could deliver a proportional voltage or amperage (DAC) that corresponded to appropriate analog motor behavior in the amp, you'd have the "driver". A G320 is a hybrid driver/amp, as it has a microprocessor and DAC on-board.



As per above, you'd still need the amps; the G100 is a "motion controller" (computer built specifically to deliver machine control signals). The G100 is superfluous to the RoGR.


Anytime! Cheers

Alright...got a lot of the small stuff spec'd and ordered today...time to finish the beast! raWr RoGR (nearly) lives! Picked up some 3/4-6 Acme leadscrews for the "minor" spec'd RoGR, and to use with the (luscious) servos while I find another supply of ball-screws somewhere...(deal fell through, doh, go China! Come on USA, drop the price!!). No worries...hey, I love Acme screws! They have a lot of stiction letting you get away with all kinds of naughty-naughty things that no one should ever do to a spindle! 3/4 screws are some chunky monkeys too...mwuhaha. Lastly ordered pillow block bearings to receive the machined (mini-lathe:P) screw ends, and ordered "love-joys" to couple the screws to the motoRs (hehe).

~cheers!

Oh yea heheh, picked up a smaller servo to close-loop the Z..stay tuned!

I'm tuned!

Blast! I began the last machining ops needed to finish the RoGR and broke my last .125" carbide making a simple squaring cut....

I was really looking forward seeing most of RoGR completed by Monday. It's possible the local tool supply is open tomorrow...I'm definitely gonna find out! The worst thing is that the redesigned gantry is ridiculously easy to make, I just forgot about the tendency for carbides to break when cutting anything that leaves a long thin edge-shaving. The shaving always binds in the bit at some point, and the hard, brittle carbide shatters.

USPS delivered two more servo amps to test; both look promising. For now, I have decided to leave the power supplies out of the motor kits, mainly so that I can finish the spec and put them up for sale.

According to China, we should be seeing the RoGR-BK (landed) in 3 weeks. :/ Last minute spec changes hurt the time frame some, but more so the factory didn't tell me that some of the bearings were not in common production, so the lead time increased (when asked why I wasn't told, I get; "I already blame production line!! I already quarrel them!!" -- yikes!) This is "fixed" now, and the kits are in transit (again). It is better for us though, as the final design is much better than where we started.

Bummer, I was hoping they would be included in the kits.

Well, not a bummer per se; it's actually best for now. While the SMPS is affordable, it is under-volt (@48vdc) to make the servos really fly (to be fair, my 48vdc/servos are far faster than my current NEMA23 steppers).

I've been working on a plan to use rectified 115vdc at the motor, having the PWM done at an HBridge on the motor. If so, only one micro-controller is needed for all 4 (5,6) of the motors, and no major PS are needed at all; just rectified AC (to DC). This can save a lot of money and time; and for the performance @115vdc, would save at least $2000 over anything comparable. (Try getting a 115vdc PS with 30a!)

So, that said, you can run them from your own linear/sm power supplies @48vdc and top out at around 120IPM, or spend for a very expensive linear supply to deliver something like 80vdc @ 12a to each of 3 or 4 motors; or you can hang tight for the plan that I'm finishing now..

The motors and encoders are now for sale!

BTW, this was a single cut through 1/2" aluminum @ 50mm/min.

The RoGR could do this all day long if it weren't for those meddling side-slivers!

Side-relief has become a study for me...the easy solution is to cut the workpiece with an: (algorithmic/interval_derived/interpolative) pattern allowing an "ejecta-notch" at numerous prescribed "relief intervals" that would easily solve the "issue". The "relief interval" is a function of tool size.

Alright, got the single servomotor/quad encoders kits up for sale. I'm reluctant to bundle them into the full RoGR kit yet, as I still have to test a closed loop solution for the Z that will be completed by a few day's time (I"m hard-headed, what can I say?). If you spoke up for a RoGR kit on 4/20/08, you should PM me so I can apply your discount now. Judging from the constant email we're getting about these, I think they will not last long. When they are gone, the party is over. The cheapest motor I've found that can replace the RB-Servo is $400 wholesale for ONE motor without an encoder, ouch!

How many of these motors do we need? What is this discount?
I ordered 2 of the servos and I already put in an order for the bearing kit and for in on the early bird price

Phew, got RoGR JR moving again and machining the last (finally!!!) pieces of RoGR. If I can keep Robin from blowing up router bits, we'll be seeing RoGR come to life very soon....mwuhahahahah! Of course, the whole job is going way slower since Robin's paranoid about ruining bits...but don't worry, that won't last long!
RoGR JR is a dinky thing compared to RoGR, but it can cut through huge billet all day long! (Ask me how I know ) (Hey, time is everything when machining metals. Next I'm over at the metal-scrapyard I'll pick up a piece of 1" thick stainless and cut it to pieces with micRo, just to show once and for all what can be done...)

Here RoGR JR is finishing the final gantry design, which is a simple as can be. A few mounting holes for the bearings and motor, and a big hole for the motor-screw coupling.

Click to view attachment

The plug that came out of the hole is very finely finished for a piece with no finishing cuts!

Click to view attachment

A quick vid showing RoGR JR putting it to some 1/2" alum billet.

Hi Styro....If you are building a RoGR you'd want 3 "RoBLOKS-SRV800 w/Quad" servos; two for the X axis, and one on the Y. The Z axis motor spec is still pending some testing I am finishing, hence the lack of a complete RoGR-MOTO kit yet. The discount for the complete kit was proposed on my daughter Oli's birthday, 4/20/08, when I was still stupefied from "baby-flash". I offered the RoGR-MOTO kits for $549 all-inclusive to anyone who committed to buying a kit on that day. I may eat some money on that deal, especially if I go closed-loop for the Z, as I'm basically giving that whole part away (oof), but hey... I love to hook people up! Especially hard-core DIY junkies like myself.

Doug Phelps sent in his Argon servo amp for testing, thanks Doug! This amp is very evolved, especially considering that Doug makes these with his salvaged SMD pick-and-place robot here in the USA:

Click to view attachment

I ordered 2 servos and then ordered 1 more. Can these 2 orders be combined into one and is there still a chance ot get in on the discount?

Worked late but finished the gantry blocks. These are 1/2" alum, and you can see how simple the machining is..

The first piece:

Click to view attachment

And both pieces:

Click to view attachment

Woooohoo! No more major machining, and no more major spec changes for awhile.....{bc exhales}.

Next was to modify the ol' mini-lathe to pass through the 3/4" Acme screws for machining the ends. The 3-jaw chuck could only pass 1/2", but the spindle-bore on the mini-lathe is something like 7/8". First was to remove the chuck:
Click to view attachment

I had a piece of circular alum from the scrapyard, like 20 years ago. It is one of those things that you carry around, trying to justify having. It (finally) had use as a new backing plate for the lathe:

Click to view attachment

Robin did an excellent job of locating, scribing and boring the holes to make a good match. He also tapped the holes and mounted the piece...sweet! It is a relief to work with capable people!

The mini-lathe has a one-thou dial:

Click to view attachment

Tapping...

Click to view attachment

The face-plate is beautiful:

Click to view attachment

Last was to simply add a RoBLOKS 20mm way-support to clamp the 3/4" screw for end-machining:

Click to view attachment

Of course we can combine shipping!

The discount is over though...but don't worry; I am prone to spontaneous insanity!

I haven't visited this forum in over a year, but I decided to check it out again today and damn this is awesome! I've always wanted to build a diy cnc machine. Is the micro still going to be available in the next few days? I think I want to pick up one from the first batch. Oops, I probably should have asked this in the micro thread

Wow, exciting stuff...really, really exciting stuff! I think this project is just great. If I can scrape the money together I'd love to buy a kit.

Can't wait to see more. A few questions though.

On the RoBLOKS homepage....

"The Drive (motor and screw) kits and the T-Slot base kits will be available mid-May, 2008. "

But the servos are not currently shipping with screws are they?

Is the mid-May a good estimate for steppers and screws?

I understand that there is a huge time difference between steppers and servos, but what is a realistic estimate of the precision trade off?

I know it's been covered, but I'll ask again in the hopes that somebody nice 'll reply. What'll be included in the full kit?

EDIT: ohh... one other questions....Do the sale prices end when the kits go up?

Gotta say it again, this is great stuff.

A blood sacrifice was offered to RoGR (typing w/ one finger); RoGR's final gantry shape: the football (go long!!!!):

Click to view attachment

I measured the screw holes and they were within a few-ten-thou'..... WHOA! It is a beautiful thing!

Click to view attachment

Everything intact I trust???

Click to view attachment

Wot dat fwum Da-da?

That's Robin...apparently at some time he had to sew his arm up while camping...

I'm intact, just nicked (well gouged) a finger.

Hi Sabrex. micRo parts sourcing is nearing completion, about 95% done. A full micRo kit is ~ 2 weeks off.


Hi mbotumbo. I really need to update the RoBLOKS page...

My delivery of 3/4 Acme screws is late, but in transit. I will test the screws with the RoGR being completed now, and if they prove satisfactory I can offer them for sale both kitted and loose. Mid-May is err now.... so not quite accurate!

A step-motor is plenty accurate so long as it doesn't miss steps. Asking for too much torque can cause lost steps. The servos are more powerful, but also track their position via a "closed loop" encoder and can vary their power when more torque is needed.

A full RoGR kit would ideally contain every piece, but so much of the machine is handmade that time will be required to develop the manufacture of the pieces. Therefore a full kit would simply consist of all of the pieces currently available. As I put more RoGRs online, parts availablity will increase. My guess at when every part is available would be about 90 days. I am offering the hardest to make or find parts first, since many people can make the other pieces.