3d printing gotchas, tips and tricks

3D printing, additive manufacturing, fused deposition modeling…it’s a bit of a minefield at times. So a few random tips, collected as recalled or experienced…

In your slicer, find the first layer speed setting and cut it to 50%. The first layer is crucial so make sure it gets laid down properly. PrusaSlicer and Cura both offer this so take advantage of it.

I hope this link doesn’t go stale: it’s a much better way to dial in a first layer, especially for the Prusa with its live Z offset. Takes just a few minutes and is better than the supplied first layer calibration or any home-brew array of squares at different offsets.

Adhesion issues on metal beds can often be resolved with soap and water…a drop of dish soap, a good scrub with a scouring pad, rinse and dry. It’s not a bad thing for glass beds either. Get in the habit of doing that at the start and solve adhesion issues before they start.

Bed leveling is easier with receipt paper, that thin thermal stuff you get from most stores these days. Thinner than copy paper and easier to work with…(h/t my much more experienced 3D model making pal Ken)

Metal beds > glass beds, for me anyway. Easier part removal through faster cooling and the ability to flex the bed and pop the pieces off.

Slower/higher quality is better. The time for a print to complete happens once. The moments when you look at something rough and raggedy-looking last as long as the part does. Go for quality.

As a test print/project, find the tool holder for your model on whatever model site you like and make it. Having those tools handy is good and it’s a good test of the machine’s capabilities. More useful than a benchy…

A camera to monitor the process is good and if you are going that far, may as well go with Octoprint. See an earlier post on that.

On purchasing choices…

Having worked with machines from AnyCubic, Creality, and Prusa (loaned as it wasn’t working), I would buy AnyCubic again before the others. Dual Z-axis machines > single Z, for stability and alignment. More expensive, as you need another motor as well as the hardware, but one less thing to worry about. Only the Prusa MK3S  has bed levelling (a misnomer: it doesn’t level the bed so much as map it out and work from that map) but that doesn’t mean I have never had to mess with the setup. As much or more than the other two; in fact I would say more than the AnyCubic i3 Mega S, the least sophisticated of the three.

The Prusa and Creality Ender 3 are quieter: the Prusa is silent when it’s not actively working, as as the Mega S now that the internal fans have been replaced with Noctuas. Noisy while working but silent once they cool down. The fan on the Ender is all you hear and the internal fans cannot be turned off with gcode commands so replacing them may be on the agenda.

AnyCubic and Creality both offer machines that can sell for  as little as $100. You don’t get automatic feeding or runout detection at the low end but you can live without those. And they can be added: all these machines are hackable. Bed leveling/mesh mapping can be added as well.

The Prusa MK3S looks and sounds like a great machine and when it worked, it was great. Lots of control/visibility into the process, adjustable speeds and temps and Z-offset, plus active firmware development, all good things. But I have some to think of it like an old European sports car…great when it works but plagued by gremlins when it doesn’t. A naff thermistor meant the bed temps were low and uneven which made adhesion touch and go. The extruder has issues with blobs and uneven filament deposition which either would leave turds on the work pieces or worse, knock the pieces over by being unable to clear the layer at which it was working. Took awhile to work that one out but I haven’t found a solution or anything beyond that symptom.

The Ender was beset with manufacturing issues. The Z-axis had to be shimmed with cardboard (helpfully provided in the packing materials) before the extruder would move freely. The adjustment of the X-axis is a bit subjective which can be confusing. The bed needs regular leveling, sometimes between jobs on the same day. Not sure it that the X-axis sagging or if the bed is actually drifting. A dual Z-axis machine would be better but also not $100.

The AnyCubic is ugly, has a crummy touchscreen interface (a natural fit for Octoprint), unsupported/has old firmware, but with the replacement of the glass bed for a magnetic metal one and the replacement of the internal fans, it’s actually pretty solid. Noisy — the steppers and controllers sound almost musical — and the UI is annoying: you still have to use it to load/unload filament. But it does have automated loading and runout detection, as well as being a dual-Z machine. You can upgrade firmware but my one attempt at that rendered it inoperable for 2-3 weeks until I was able to revert whatever happened. I think I will leave it alone.

For all that, it’s pretty reliable. The bed doesn’t move/need adjusting, it came with a spare hotend and a better set of tools that other two machines, and it mostly just works.

a better argument for basic income would be hard to find

Andrew Goodacre, the chief executive of the British Independent Retailers Association said the cost of living crisis had made people “think of alternative ways of sourcing items that are essential to them”.

Yes, essential items…like food. This verbiage could have come from a business plan or some other document no one is expected to read. But this was a quote provided to a newspaper so here we are.

What if — instead of failed schemes like the derisory  “eat out to help out” — HM Government simply gave its hungry citizens money…money they wouldn’t keep, wouldn’t hoard or send offshore but would spend on those essentials the anodyne Mr Goodacre mentions above. Anyone who has any experience in the real world of economics, which excludes most of HM Government or the US congress, money at that level doesn’t have owners, only spenders ( © Omar Devone Little). Any money you give to the people mentioned in the article will be in someone else’s account, some store’s till, within 24 hours.

Not like I expect Labour to pick up this ball and run with it but maybe after they get in, they can start a scheme like this for the neglected cities of the north and elsewhere.

 

 

Octoprint, octolapse and friends

A time lapse from Octoprint/octolapse:

To make the magic happen, you need a Raspberry π board with PSU, an SD card (8Gb is plenty), and a camera. All of that can be had for well under $100, closer to $60 if you are careful. The Raspberry π cameras are fine but other options exist from Arducam and Inland. I have had three Raspberry π cameras fail in a month so I am not inclined to advocate for them. An Arducam model I bought to test worked first time and has given me no trouble. They both fit in the same size case and as bonus the ArduCam has a status light to let you know it’s working.

If you go with Raspberry π, the Camera Module 2 is what you will most likely want, as it is well-tested and supported. If you find the camera does not work or that the Raspberry π board doesn’t see it as connected, reboot the board and see if that clears it up. The command line methods for detecting the camera are not available if OctoPrint is running, so lsmod and friends are not there for you. But by the same token any Linux or other FOSS experience you have will come in handy.

You’ll need a case for the Raspberry π and the camera…so many options. I chose this one which has versions for the 3 and 4 boards. For the camera mount, I recommend something that mounts to the X-axis and can be adjusted to frame different points of view. For monitoring active jobs you want something that will cover the full print bed but for time lapses you want to shoot straight across the X-axis. Aiming straight back on the Y-axis or at 45° also work, if you can find a mount that allows it.. You might want to choose an articulated model in that case since the time lapse process will push the bed into the shot on each layer change. So something like this if you don’t want to make your own. Look on thingiverse as well…the kind that clip onto the QR code on the Ender are a good choice.

Which Raspberry π should you get? I have tried the 3 and 4 and have no preference. A model 3 with 1Gb of RAM is perfectly fine and at $35 (currently) it’s pretty friendly. The 4 can be had at the same price, if you can get one. The one drawback for the 3 is the USB micro B PSU, where the 4 and newer models use USB C. It was an issue for me as the case I initially chose for the 3 didn’t allow the PSU to be plugged in easily. The 4 is about $20 more at this writing (with 4Gb RAM: not a lot of choices sometimes, they are often in short supply).

You will need a USB cable to connect your Raspberry π to your printer. Many of them use the A/B type like a conventional printer so you likely have one. The Ender also uses the Micro B 🙄 so you may need to get one of those. Monoprice is a great place for cheap cables and has all kinds. Here’s one for $1.49.

Getting the Raspberry π imaged and setup and octoprint installed is documented at the Raspberry π site and at Octoprint and will be more up to date, but the short summary is to download the imager from Raspberrypi.com and let it set up your SD card. You can set up your hostname for the Raspberry π, your login info, WiFi credentials, and all of that in the imaging process. Put it in the Raspberry π board, and fire it up, do all the setup that is documented at Octoprint. Once it’s installed and you have logged in, install the Octolapse plugin. Hit the wrench icon to set up OctoPrint for the printer it will be controlling.

In the OptoLapse plugin these settings should deliver a timelapse video with the extruder out of the frame, letting the workpiece appear to grow out of the build plate as in the video above. Lighting and framing might be bigger challenges than the technology.

You can get more sophisticated with snapshot commands embedded in the gcode but I have not explored that. You will need to step through some options for OctoLapse (in the OptoLapse tab) based on the slicer you use: the automatic setting doesn’t actually work. OctoLapse will cancel any job you submit if you don’t fill out all the information about layer height, retraction distance/speed, etc. Save yourself some time and hassle by doing that.

If you use different slicers (Cura vs PrusaSlicer, for example) it might be worth setting different printer profiles that take that into account. So you may need separate Ender/Cura and Ender/PrusaSlicer profiles as the printer destination. That way OctoLapse knows when to trigger the snapshot function regardless of how you created the gcode file.

 

jobs are not work: a proposition

I have been seeing these monthly jobs reports for years — decades — and have always questioned what they really mean.

Total nonfarm payroll employment increased by 199,000 in November, and the unemployment rate edged down to 3.7 percent, the U.S. Bureau of Labor Statistics reported today. Job gains occurred in health care and government. Employment also increased in manufacturing, reflecting the return of workers from a strike. Employment in retail trade declined.

Did 199,000 people suddenly join the workforce from somewhere else? Were they in school or somehow out of the market? And are the new jobs a net gain or are there some losses that offset them? There is a lot of detail in these reports and perhaps some of these questions can be answered. For instance I have often wondered about earnings…are these “new jobs” good paying jobs or are they lower-wage replacements from jobs that are no longer available…

In November, average hourly earnings for all employees on private nonfarm payrolls rose by 12 cents, or 0.4 percent, to $34.10. Over the past 12 months, average hourly earnings have increased by 4.0 percent.

12¢ an hour over 40 hours isn’t a lot but it’s more…more is better.

But the bigger question I find myself drawn is, what is a job anyway? Maybe I have read too much David Graeber — or maybe not enough — but that’s where I think this is coming from.

Jobs are not work. Let’s establish that. We all do work that isn’t paid, whether it’s household tasks or childcare or meal prep, for others, it’s work that has to be done, just like the work we are paid for. There is a lot of work that needs to be done, public service like education or park maintenance or caring for those who are unable to take care of themselves — the very old or very young. Is that work being done to the extent it’s needed? Ask anyone who works in the paid workforce in those sectors and they will tell you there is more work than they can find workers for, and even with willing volunteers, there is work that is left undone.

So what is a job? A job is a tranche of work that can be monetized and sold in the market. This often means it comes with a performance requirement, like some number of hours to cover benefits and taxes, the costs of doing business which is how these jobs are created — to enable businesses, in most cases. So jobs are commoditized labor, not unlike the collateralized debt that was at the root of the 2008 housing crash. In that instance, risky loans were underwritten by the finance industry and wrapped up in bundles, then sold as assets themselves. No one knew what was in the bundle, but the underwritten value was taken at face value with results we all remember.

I see jobs as commoditized bundles of work, in much the same way, but mostly without the risk. We have often heard someone talk about wanting to hire staff but the hours are not there; in other words, the costs associated with adding someone to the payroll are too high in comparison to the revenue gained by adding them. It might breakeven at 20 hours a week or even 10, but most jobs are 40 hours week, with anything less than that coming with reduced benefits and the general sense of not being a full member of the staff. Part time comes to mean a partial person, as far as status goes.

So how we more clearly delineate the idea of work from our concept of a job? You could work all day long collecting trash or cleaning public spaces but if someone asks you if you have a job, you would have to say no. You could volunteer at a school or senior center, tutoring or managing appointments or any number of really useful tasks but do you have a job? Again, it would be no.. since you didn’t get paid for it and because it was not counted in the bureau of labor statistics jobs reports, your work has no value. You will likely be as tired as someone who does any number of the bullshit jobs David Graeber discusses, some of which actually do harm. But you don’t have a job.

I recall a passage from the Millennium Whole Earth Catalog, where someone posits a worker making land mines…imagine that being your contribution to the economy, making reliable ways to make land unusable and impassible using death or severe maiming as the means. And the amount of effort needed to remove them and restore that land to use: that is the byproduct. But that job would be counted in whatever nation’s BLS numbers.

So how do we get to those sunlit uplands? Single payer healthcare would be one way, since access to healthcare in the USA  is linked to jobs and is often one of the differences between full and part time work. When we hear any politician claiming to be a champion of small business, consider their stance on single payer vs corporate healthcare, and how easy it is for someone to get paying work if their employer doesn’t have to weigh the cost of adding them against the overhead of benefits. If you don’t support a plumber or carpenter taking on some help, even part-time, to better serve their customers, you are not a champion of small business.

Consider also that labor statistics don’t included unpaid or volunteer labor. How many Meals on Wheels volunteer cooks and drivers are out every day providing necessary service but uncounted because it isn’t paid? What about school volunteers, filling the gaps that classroom teachers can’t cover with reading or math support for those who need it? We have to remember, the recipients of Meals on Wheels were once members of the labor economy: their jobs were listed in those statistics, just the schoolchildren of today will be doing work that will be included in reports yet to come. Why is the work being performed on their behalf not counted?

The economy is bigger than the remunerative labor market. The root of the word economics — eco — stems from the Greek oikos  which referred to “three related but distinct concepts: the family, the family’s property, and the house.” Ecology and economics both stem from that root, the idea of a system of mutual exchange or interdependence. To exclude unpaid labor is akin to excluding the need for sunlight in a garden because it can’t be purchased.

We continue to learn more about the ecological systems that surround us and Adam Smith’s 1776 book An Inquiry into the Nature and Causes of the Wealth of Nations was just that…an inquiry, an exploration of what he saw as natural phenomena that he wanted to understand and explain (Smith’s book has a similarity to the Bible as being often quoted but rarely read in full or understood). Not a week goes by without some new revelation about an unseen link or process in nature, most often because humans have disrupted it and are now seeing the consequence, whether it’s beavers managing flood control better than the Army Corps of Engineers to flowers evolving to self-pollinate as their insect and bird partners are declining in population.

Maybe the idea of sunlight as the unpurchasable but necessary component to life as we know it is apt. We all do necessary work that goes unpaid and perhaps is impossible to value in any market. We feed each other to fortify each other for the work that needs to be done. Sometimes presence is what’s being given, being there without being busy. And where that is the work, it should be counted. Does this mean that caregivers should draw a salary? I wouldn’t argue against it but maybe the work we pay to have done isn’t worth paying for. It’s always been interesting me that some jobs get wages while others get compensation…I think the work that requires compensation is work that we know is not worth doing (see Bullshit Jobs above), and we admit that we are compensating people for the waste of their precious time spent doing it.

 

 

 

This kid gets it

“When houses are a million dollar plus and an older couple will likely outbid us anyway, we’re gonna relinquish any lingering delusions about homeownership.”

This kind of talk might make holiday get togethers a little tense. Today’s young adults (and some older ones) don’t have a lot to look forward to.

Meanwhile, with a looming recession, multiple ongoing wars, and climate change at a catastrophic tipping point, others voiced doubt over whether the future is even worth investing in.

It’s as if someone in the family already spent the rent and food money on their expensive travel, the very thing young adults are put on blast for.

we could elect a houseplant to D5 and get more for our tax dollars

Seattle’s worst city council member never ceases to disappoint. This is the same person who described the removal of homeless people from an abandoned home as “humane relocation.” That’s the kind of phrasing used to describe the rehoming of a food-habituated bear or nuisance family of raccoons. Or perhaps how indigenous peoples were sent to industrial schools or reservations…you’d think someone who puts their indigenous heritage front and center would be more aware. But then you wouldn’t be thinking of the same person. So what is she up to this week?

She seems to think this local business that is open one night a week and sits on 8800 square feet of prime developable land is worthy of our praise. A roadhouse smaller than most houses in the area, open one night a week…how much tax revenue that that generate for the city, that we should be so grateful? What acts play there that we should appreciate its status as Seattle’s last roadhouse?

The fact of the matter is it sits on 1/5 of an acre of land on Lake City Way, a busy arterial/stroad/busway, where housing or mixed-used development would be welcome. As you can see, the tavern is a teardown, valued at $1000, the lowest value you’ll see. And the land is valued at $1.1 million, up from $71,000 at the end of last century and  $528,000 10 years ago.

 

And this valuable contribution to District 5 remits $5,690.58 in property taxes every 6 months. About 1% of the assessed value over the course of a year. Seems like a great deal for someone. Not Seattle or anyone who lives there though I will give the owner full marks for using the business address for his tax mailings. Not Bellevue or Glendale, CA…

District 5 has Lake City Way, Aurora Ave, Northgate, and Holman Road/15th Ave NW, all inhospitable stroads/car sewers, as well as half a dozen strip clubs. Not that I care about how people make a living but one of them a block away from where I live, on half an acre of land, has an adjacent parcel  zoned for multi-family development. To no one’s surprise, it remains undeveloped. In the same block is a disused car wash, closed for about 2 years now and taking on a new life as a graffiti canvas.

Just up the block is an installation of a couple hundred eco-blocks, placed there after some local vigilantes attacked an RV/homeless encampment. I didn’t see that attempt at “humane relocation” mentioned in her newsletter.

Further up the street we have a disused car lot, an idled truck sales office, and entirely too much low density development for such high value land. Further south, more disused land and low-density development on a high volume roadway. But at a 1% tax — no pressure to increase density or redevelop that land — everything is fine, as far as our councilor is concerned. I asked. She thinks it’s great.

Hayekian property rights outweigh human/civil rights, it seems, in this progressive city. Certainly in District 5.

can we find a better word than “gentrification” to describe this?

Business is booming at the Roadrunner Grab’n’Go deli outside Joshua Tree national park. Demand for sandwiches, mezze boxes and local vegan cheese has remained high through the summer, even as temperatures soar in a desert landscape that now attracts more than three million visitors each year.

But the shop’s co-owner, Merilee Kuchon, has a problem. Her employees, many of whom grew up here, are struggling to afford to stay. Over the past year, she’s lost at least a dozen staff, driven out by local rental prices that have soared during the pandemic. Now, she’s worried about hiring enough employees to keep the shop going when even more tourists return in the fall.

Gentrification is technically correct…

but this isn’t just a haphazard evolution of some bohemian district becoming cool. It’s more calculated than that…some of the land in the area was bought in the event someone else — like this sandwich shop owner or some backcountry outfitters — put in the work to make the location valuable.

I don’t buy the “inevitability of gentrification” unless we want to accept the inevitability of being hit by a bus if we cross against the lights. It’s avoidable, not inevitable. Better land use policy, managed by land use taxes and hard boundaries, would prevent both the destruction of beautiful places like this and the evisceration of working people’s lives by speculators. It’s not inevitable, unless we accept it to be so. I don’t.

Reading DeLong’s “Slouching toward Utopia,” I realize I am firmly on Team Polanyi and now understand why I have never bought into Team Hayek. Property rights don’t trump human/civil rights or the Maslovian requirements for a good life. The pendulum has some way to swing before we get there, though.

on perverse incentives

Walking the picket line today, some of my colleagues were simultaneously griping about their property taxes and rejoicing that they would be able to retire with a nice cushion when the time comes (a reasonable house in the $300-500k range 1-15 years go now commands a high 6 or even 7 figure valuation).

I wonder how many of them understand that the valuation they are seeing is not from their loving cared for home or the carefully curated plantings they have invested in but the land itself, the coordinates they own exclusive rights to. I further wonder how they would handle it if their property taxes/property value dropped. Again, the value is in the land: their house is the same house today as it was yesterday or last year. But the value today is much higher than it was a year or five years or ten years ago. And none of that has to do with any expense or effort on their part.

I doubt there are any plans to upzone or increase the density of any land near them, like the commercial land along the arterials that is often occupied by single story strip malls or single family homes. But if there were and they were promised a lower tax bill that didn’t require a lower valuation, I wonder if they would accept it? What if commercial property, either multifamily residential or pure commercial was taxed at a different rate to force more dense development and what if the resulting increase in tax revenue meant that residential property — non-remunerative property — could now be taxed at a lower rate, less than the 1% that every property owner pays now?

Right now, we have the perverse incentive of families investing in their shelter as their retirement fund, paying higher and higher taxes on it, in the hopes of a payback. A low risk bet, to be sure: they’ll get their money but what are they forgoing on both a higher purchase price/higher mortgage payments and the rising taxes that are both driven by the finite nature of land, the artificial scarcity of limiting development, enforcing parking minimums, keeping building heights down, etc.?

The perverse incentive of paying too much for artificially scarce land to pay into a retirement fund while keeping housing prices out of reach seems like something we should address. Conflating housing and land as housing is part of the problem, mostly due to the single family home as the only option for buyers, especially if they have a family. But taxing commercial land at the same rate as residential land is also part of it. Why should we tax land used purely as residential land at the same rate as the land under a store or restaurant? Are commercial and personal auto registrations taxed at the same rate? Is electricity sold to homeowners at the same rate as businesses?

If a parcel of commercial land commands $1 million per acre per year, as we have seen offered, what is that land worth? At 1% that’s $100 million. But the assessment didn’t reflect that. The developer said it was worth that much to rent, and buyers set the price in every market, as sellers eventually discover.

wait, you’re paying the people who are gutting your community?

I think this is backwards…

On Aug. 1, a new pilot program launched in North Tahoe that pays homeowners up to $24,000 to rent long-term to locals.

The goal is to “unlock” homes that are sitting vacant in Lake Tahoe and alleviate some of the mounting pressure on local workers who are in need of places to live that they can afford on their wages.

So you are going to pay homeowners a bonus, on top of the unearned wealth they already enjoy, to rent to people whose wages are too low to allow them to pay market rate rent?

In North Lake Tahoe, 65% of single-family homes are second homes and short-term rentals, according to a 2021 Mountain Housing Council study. Many of those houses sit empty for the vast majority of the year. Meanwhile, at any given time, about 200 to 300 households who work in the Tahoe region are searching for housing,

Rather than reward rentiers for tying up valuable land and housing, the rents on which will not be spent in Lake Tahoe, why not tax them for the value of their land, the value that all the local people who can’t afford to live there created? Those property owners didn’t create that wealth. They may never have even have set foot in town or might have inherited that property. Meanwhile, the local economy is built on the backs of people who can’t actually enjoy the place where they live.

Is inflation the right term here?

Inflation, as in a weakening dollar? Or hyper-inflated land values?

The original Kidd Valley restaurant, which opened in 1976, will be permanently closing, according to a Facebook post from Kidd Valley.

Located on Northeast 55th Street near Ravenna Park and originally opened in 1976, the 800-square-foot restaurant needed to be remodeled to bring it up to code for the Americans with Disabilities Act.

According to Kidd Valley, because of inflation, it became too expensive to remodel or build a new store.

Hmm, I wonder…what’s the value of the land under that burger joint?

Well, now. The burger joint itself has been a $1,000 teardown since the year I moved to Seattle. But look at the value of the land since then…

About 1/10 of an acre valued at one and a half million dollars. Why is it worth so much? And why did it rise so quickly?

  • It’s right near the University of Washington, within blocks of a very densely populated student quarter
  • It’s on an arterial with frequent bus service
  • It’s also near a very successful shopping destination

All this as Seattle’s population increased from around 600,000 to 750,000…about 25% more people with a steep rise in land values/prices.

Which of those items did the owners of Kidd Valley build? At best you could argue their taxes helped fund the transit service. But the rest of it? All built by the community around it. Sure, they own the value of the business, the loyalty to the brand, the reasons why people go there. But the land value? Not so much. A vacant lot would command a high price at that location.

In a way, they are right: if they were to tear down and rebuild that 1960 building, it would be assessed at far more than the $1000 it’s valued at now. So they would be paying that onerous 1% or so on the $1.5 million in land with perhaps another $1 million on the new building. Maybe that doesn’t pencil out for a fast food restaurant, no matter how good the location is. And that’s fine. Let the land be repurposed. We’re not talking about cutting down the redwoods here: this is land that has already been developed multiple times.

So they are sitting on land valued at $15 million per acre. One could extrapolate that value along that corridor for a few blocks in either direction: how would one assess a land value tax there? Probably more than $13,364.34 per year, I think (and paid to an address outside Seattle, so Seattle doesn’t even get the paltry tax anyway).

If you turned the property tax into a ground rent/and value tax — maybe raise it by a factor of ten as the value of the location increases — the decision to move on would have happened 10 or 15 years ago, freeing up that land to find it’s highest and best use. Perhaps as a multistory mixed-use building with a Kidd Valley at the street level and 3 stories of residential above it. Look just down the street and see the dense development across from the University Village mall. Assess the value of the land at it’s most remunerative use, manage the improvements at a lower rate with zoning and land use guidelines to let developers maximize that value, and stand back.