Old school, you have the love. Was that a Ferguson no...let me see, the British tractor motor that AH and MG used? Starts with an "A" I think. Well, I'm just 40 but I love vintage wheels. My first 'real' car was a 68 Alfa boat-tail Spyder set up for racing. Pops decided I'd do better in the cherry '70 Cuda somewhere around my 17th..

I drive an Escape Hybrid now...hehe...but been thinking about buying a pile of Norton junk to give me and Jr something to make a fire with.

BC, totally off topic but you might look at the Kawasaki W650s that were imported into the US around 2000. They kind of resurrected some old designs that they had in the Kawasaki vault and combined in with new tech, but it still looks like a classic ride and handles great too. I know its not a true classic ride but then you don't have to deal with the constant oil leaks that plague British motorcycle machinery and get the benefits of modern manufacturing technology and the resulting tolerances. Its still in production for sale in Japan/Asia but the market here in the states was a little slow they stopped bringing them over. I had a beautiful blue and silver version that got totalled despite my best efforts when a little doggie decided to step into the road and then follow my headlight. One year and 15,000 miles riding with no incidents and got taken out by a dog. That's life.

Yeah, my dad was into SCCA and I was around cool cars growing up. My dad gave Carol Shelby his first ride in a sports car (for trivia it was a TD MG.) Now my budget limits me pretty much to motorcycles. I do like old ones. Had a couple of Danish Nimbus bikes. a '37 and a '54 (interesting 750cc, inline four, exposed valves, "strap iron" frame) You probably have heard one of them. It was used for the sound only of the Indiana Jones sidecar rig in "Last Crusade" The one you saw on screen was another of my bikes (2 cylendar bike-4 cylendar sound, the Nimbus club in Denmark was in heaven when I wrote an article in the newsletter)

For cars we now have a Prius and an old Miata (reminds me of my Sprite).

Love LL. Y'al keep up your strength, eat good food, be happy and wrap up the micRo and it's spindel.

Just so happens that a spindle motor is on the way to me from Asia...this is the 'big boy' version of the 100krpm outrunner. I expect it within the week.

Very cool! I wait with interest.

Ok, no eye-candy today but I wanted to report on our findings about production. We tested casting the production pieces in urethane resin today with inspiring success. This will allow us to keep very high precision by casting off of "perfect" masters rather than the very slow process of machining the parts with the attendant error, wear and never-ending problems. Aside from the many desirable properties of the purpose-derived isocyanoacrylate polymer, the fact that we can cast the pieces is a huge step forward for our production effort...I'm almost emotional.

brainchild, I've done some small production volume casting with two part 1:1 casting urethane. I have run into a few situations where shrinkage was an issue, particularly with parts of varying wall thicknesses and thought that might be helpful to bring up. Out of curiosity are you using silicone or aluminum molds?

Hi Graham, By all means please elaborate on your experiences. The resins I have chosen are from Freeman, and are muxed with alum and (possibly, if tests show benefit) PTFE micronized powder. The cure shrinkage is at .0009/1". Our castings are over-sized and after several days' cure are machined to final spec.

PS, the micRo parts rely on massivity with only a few areas requiring very high tolerance.

I knew it would work I knew it!!

Very excited that you are going that route, I got a bit sad when you seemed to be planning to stick to machining.

Compensating for shrinkage is best done from experience but you can get good guesstimates by examining the datasheets. Make all kinds of nice casts (and accessories) with your CNC machines and cast all the blocks

Well, my enthusiasm was tempered a bit today by noting that the Acme screw would bind in the resin-cast block if the block were put in the freezer for ~30mins. This is not however our chosen resin, but rather the only resin obtainable locally. This resin exhibits minor foaming which is indicative of older stock. Numerous attempts to speak with an engineer failed me today, with much time on hold. The datasheets show properties similar to Delrin, but the datasheets really mean nothing compared to having the material in hand. With these issues in mind, today we worked toward a simplified plan for machining, as to me it seems best to work toward a production method constantly without becoming married to a single idea. I'm still very much desirous of casting the pieces, for obvious reasons. In the future I desire to cast RoGR pieces in alum, so steps towards that end are always desired. Regardless of the method, production standards will be established by Sat (ha, well, so long as the materials arrive. I do however have the materials to start the machining run now, so that is a natural course for the short term).

Have you considered casting the "main" block of the piece using this resin, then machine precise openings to accept inserts of delrin for the nuts and guide rail bearings ? With this setup you could easily change/try various insert and block materials w/o having to replace the entire unit. Perhaps only the nuts would would require being an insert, reducing the complexity, while addressing the main wear part ? (I'm assuming the nuts will be the main wear area, vs. the guide bearing ?)

Robin has attempted to convince me to try this. The outcome would be the same as it relates to thermal instability. Any resin that has detectable delta in a relatively narrow temperature differential is unacceptable, as this change would affect the alignment of the entire machine. Additionally, to maintain high tolerance it is desirable to keep the number of parts to a minimum. Any mechanical union is a weakness and a source of error when compounded throughout the machine.

Brainchild: Assuming your final design meets your personal criteria for thermal stability in order to maintain what you would consider nominal acceptable operating alignment what might the requirements be for the enclosure in which the micRo (or RoGR for that matter) resides?

In other words - would you personally recommend I keep my shop at 20°C (or whatever temp I desire) +/- 0.1°C or what?

Yikes, I dislocated my thumb today, it's hard to type. 110f in the shop, stomach bug...the days you live for!

Robin got through to a polymer chemist today who seemed to confirm our CoE tests, but we haven't analyzed the data enough to be 100%. Based on the side-by-side tests with Delrin and HDPE, PU resin is losing on therm. It is a little disheartening if the resin isn't usable as its other properties are highly desirable. (BTW The HDPE performs excellently, so injection molding may be a good solution.)

Dan, I wouldn't issue a machine that had those characteristics, so no. The machine must operate in all reasonable ranges, I'd say beginning at freezing and going to about 140f.

Really? I would have said "reasonable" would have been about 50-100F. I think 0-140 is extraordinary, which is not a bad thing, but for a design criteria, is pretty tight. No?

As I think about it, you want to create something that is going to be manufactured for a large range of temperatures, while an individual user will likely only see a much smaller range.

I was thinking injecting molding, but setup costs don't put this in the realm of "inexpensive"... but you can get a lot of small parts quickly. ... and relatively cheaply, aside from setup costs.

Well, I said freezing and I'm from the South, so that's 32f to 140f, not 0f to 140f; big difference!

Currently in our tests, HDPE is an excellent performer in this range; well within spec and spins smoothly after an hour in the freezer with no detectable difference. The higher temp range reflects the motor-heating safety factor; not to be ignored. There are extremes represented in common use of micRo, and loss of calibration in the range of common use is not acceptable. Materials that have passed the tests are HDPE and Delrin; both are injection-moldable, but as you said have high up-front costs.

Brain,

I gotta say, the temperature test is good! I would not have thought of that. Are you going by a product development map or are you making it up as you go along?

Osmosis: one observation leads to the next...and eventually the critical factors become amplifications of the 'minor observations'. Though I was crestfallen by the result, it is better to know. As any durable good, micRo must live in reality plus the occasional extremes common to our world. We have a ready solution at hand but it isn't the easiest path; yet the choice must be made primarily on 'best choice', secondarily on economy.

hmm... my beer freezes well below 32F.

(ooops. )



Ah, I hadn't thought of the motor heat being introducted into the part. How hot do those suckers get? (when your not trying to blow them up!)

BC, the shrinkage factor that I was referring to will be adequately dealt with when you machine the casting down to final specs. I was using a mold to create parts without a machining pass following.

Maybe small batch injection molding with less expensive aluminum tooling would allow you to make sub 1k part runs in house at a significant discount. The tooling could be machined in house and the heater/hydraulic injection press could be had for a few thousand/possibly designed yourself from eBay parts. Easier to say when its not my money directly buying the tools. Just an idea, I know that's what Protomold does with their production runs.

I've never not tried to blow them up, but at 2x their rated v/a the delta is ~75f.

All of the testing and research for a production method is now concluding with an inevitable result; machined technical HDPE. The material is hands down the best yet tested for this purpose. Wide operational temp, easy machining, very smooth power transfer, very low friction, excellent damping, 100% recyclable.

We've had to brainstorm the design to make changes that are better suited to production. As you know, I was concerned about our ability to produce enough as the machining is a very long process. So we streamlined the design to be 'jig friendly' allowing us to improve precision while simplifying the 'ops'. While not the fastest method, it's the only one that works, so we have to improve upon it!

I'll post drawings when Robin sends them. This also means we're going into production. I'll be ordering the plastic tomorrow.

Woohoo!!!!

Well, as you may have read, the Urethane's big problem is its coefficient of expansion, which is large enough to alter the spec. Injection molders won't touch anything as thick as our pieces and casting resins aren't suitable.

Appreciate the follow-up!

BC, good points, even a heavily ribbed and skeletoned version of the solid "massive" side supports would have significant sink areas with injection molding. I think HDPE is the best bet.

On another note, with all this worry about stepper slippage going through my mind, what is the highest end load that the current NEMA17 motor and driver setup can achieve? Have you tested this yet? I'm looking for a number that makes me think "this is pretty high and I'm sure not going to hit it with a 1/4in end mill machining a block of ABS".

If we used quadrature encoders then we'd basically be using alot of the circuitry from the yet-to-be-completed RoBlocks daddy electronics minus the servo drive circuitry. Is that correct?

What would be the drawbacks from going to NEMA23 if you could source 166oz*in bipolar steppers with a 1.8° step angle for about $30 each in small quantities?
I'm assuming the microstepping acceleration and deceleration ramps would have to be slowed down to get similar response with the higher rotational inertia. How much would our max end force go up? Would the stiction problems decrease? Any other impacts?

For interest sake I have been doing some research on injection molding (a little knowledge is dangerous). One of the newer trends is servo driven injection systems. Large shops are going to them when precision is an issue and with no messy hydraulics they lend themselves well to smaller shops.

With your current CNC know-how it seems to me that developing a low cost cnc driven injection molding system might be an idea. It is after all an enabling technology. It would also make developing future tools easier.

Some problems are actually opportunities waiting to happen.

Low production volume injection molding in aluminum mold tooling is a great way to make more than 20 parts fairly consistently. It would definitely allow people to copy parts more cheaply if they need them quicker and don't want to tie up the CNC mill. The servo drive portion of the system could supply the pressure, but any ideas on how to supply the heat that will be necessary for melting the plastics?

Hi Graham, Who's worried about slippage? There's no reason for concern. Finding the sticky parts is just my obsession. Remember that torque and speed in a stepper are inversely proportional? If you find missed steps you are trying to go too fast. Also with high microstepping the signal data is far denser which leads to count error at high speeds. The benefit is high precision at normal speeds.

Nema 23 is too large for the micRo. There are more powerful nema 17 available, but I question the benefit, as micRo is designed for 5mm or smaller tools. Again, micRo can cut anything, it's only a question of how fast. Currently the motors have ~60oz/in each, with 120 total on the X.

For using quad; you can simply count with emc2 over the parport inputs, so long as the count isn't really high.




I'd like to know how it works?

I guess we will find out what is a good "safe" speed to machine at for a given material and bit combination by experiment? If the instrument does not throw an error when it skips a count then what is our QA/QC on whether the pass was performed correctly or not? Will there be a calibration routine where the platform counts from one edge to the other or to a set point so we can mentally convince ourselves we did not have slippage during the previous routine?

Yes there are still unknowns, like spindle capability, so data is forthcoming.

For QC, you can measure the piece #1. You can make the tool use a reference before and after the job. Mostly though, you'll know from the sound of the spindle. It will either scream at you, or be 'too quiet'.

Ok.. so here's the preliminary re-design. rough-n-dirty rendering, but the important stuff is there. Distances for the Y axis are incorrect. No couplers or fasteners are drawn in. And... it's (groan) rendered with perspective on. Maybe have a more complete version online this weekend.


Click to view attachmentClick to view attachmentClick to view attachmentClick to view attachment

Damn... that blows my micRo logo design to hell.
Guess I have to come up with something else.

Just for the sake of render-porn: here was the beginnings of a design for the injection mold makers...

Click to view attachment

I kinda had a feeling that the Y support frame was going to square off, but I didn't expect the X axis to be the short direction. Is that right or is the drawing presenting the image wrong? Is X now the 10 inch dimension?

Erase, erase, erase

I didn't read everything.

I was about to gripe about the isometry!

Quick update:

Got a crude DEPCAP happening:



I've also been poking around Asia for some nice finishing tidbits for micRo (umm free upgrades to those who have already ordered!)

I've been going over the design to be sure it's going to work as expected; I always feel like I'm forgetting something, but I feel everything coming together now and I'm really excited to be producing micRo because well, it's just too cool.

What the heck is a DEPCAP? It's not entirely apparent from the image... or maybe I'm just up way too late.

BC posted about it in the RoBLOKS 3DP **ALPHA** topic, with lots more pictures:

After lurking around for much of the last year I can't put this off any more, preorder made. I will be looking forward to the spindle kit to finish this off, and likely an extended Z for some of the many projects I will finally be able to attempt. The effort to needed stave off more plans is increasingly difficult. Now to study and design so the tool can immediately be put to use!

In my preperation I am looking for raw material sources; primarily brass and aluminum. Locally there is a scrap metal yard which has a nice selection and the prices are Al (any alloy) - $5/lb and Brass - $8/lb. How does this stack up? I have looked on line and this seems to be good prices. Besides yard sales and your friend's side yard junk piles, are there better deals out there?

Dave, I typically buy my raw materials for work from McMaster-Carr because of the quick delivery and that they supply about 90% of what I need for a given project. 1/4" aluminum sheet from them ranges from $6.8 to $12+ per pound. I also regularly prototype in a 48lb/ft^3 urethane foam material called Butterboard that runs about $10 per pound. I'd say those prices are pretty good if you can find material that won't cause you to have to waste a ton of time and material prepping it for your desired application/part.

$5/pound! Yikes!!
I buy my Aluminum and steel as 'remnants' from my local metal supplier. Al is $2.50/pound and steel varies (depending on the type; tubing is more expensive than sheet) from $0.55 to $0.75/pound. Try going to your local supplier and ask to see the remnants, which is the "scrap" left over from when they cut stock for an order. Remnants come in all shapes and sizes, and have met all my needs for the last several years (I am building a welding cart right now, and I'll probably have to order the top which will be 30"x40"x0.25 and pay normal list prices), but I know that the base for my micRo will be available as a remnant.

-Brian

You can get it for less than $3/lbs shipped from eBay pretty easily. Usually they'll cut it for free too.

Aluminum = 0.0981 lbs per cubic inch

http://www.principalmetals.com/utilities/primecalculator.htm

Thanks Graham. I'll put them on my list.




Those sound like good prices. I'll try and find a local place.

I don't know why I didn't think of eBay. Thanks

So, any news? No posts for a couple days gets me all itchy...I've been doing weird things...like watching under-aged girls doing gymnastics and feeling all dirty about it - till they said some were 18-20, whew much better.

But seriously, any news on the micRo front?

Yes. I spent the last couple of days finishing the BOM, sourcing various items, and designing the production jigs. Yesterday we began setting up our mid-sized gantry bot for the first run. The bot wiil carry a 1/2" drill spindle and perform all of the drill operations. Since a drill is the maximum D for any given hole, the hole is straighter than if using a router. I'll take a pic for you today.

Please do and post quickly ... Don't just do it for us, think of all the innocent young gymnasts who will avoid feeling heebie-jeebies

Phew, long day. To boot I forgot to take my camera! You have missed the most "boring" thing on Earth; the hole drilling tests! Test after test was made today to get the perfect hole; I'd say we're 50% there...but the fact that we're working on a couple "thou" 'error' is very inspiring to me...having taken pains to make it as perfect as possible. The plastics supplier informed me today of a delivery time of 14 days for the stock, so I've got a little time to be sure every piece comes out like the last, but in just 4 days' time a lot has been accomplished.

I know I'm not giving ya'll eye-candy, it just that I gotta work hard right now: crunch time!

I plan to try for some "perfect" holes myself. I hope, when you get time to breathe, that you can describe the techniques and best choices you have learned from your tests.

I also have a question about the parallel port cable. Have you noticed any increase in errors due to length of the cable? I keep working on the space where micRo will go (I've got to do something!!) and I'm trying to place the computer where it will stay clean.

I can see that the priority of the moment is getting the manufacturing process down for micRo, which is as it should be. But if any information about the spindle happens to surface, that would be great as well.

If I remember right...I think the analog cables all have a maximum operating length of 25 feet. In comparison USBs have something along the lines of 6 feet. But I could be off.

Glad to see an update BC. I figured you were hard at work.