Ohhh, so that's why we bake the enamel one layer at a time!

That paint was harder than a bad day.

 I just now finished this.  It took me a lot longer than it should have.  The painting was routine.  Getting the paint off the lettering was not.  It turns out that baking enamel on as I do has the desired results - that stuff gets tough. It will not crack or chip, and it resists abrasion.

 I started with wet-sanding, but this went very slowly, even with nice, fresh/sharp, silicon carbide paper in a semi-course (200) grit.  Harrumph.

 Next I tried mounting it on the mill and running a little (sized for a Dremel) convolute abrasive point over it.  That seemed to move the paint around a bit.

 Finally, I decided to get aggressive; I put my largest fly cutter in there.  That did the trick.  Even so, it required multiple passes, and I had to touch up the 'ON', 'OFF', and the ®egistered Trademark symbol by hand-sanding with one of my paint-stirrer sanding sticks.

 The enamel finish was dulled in a couple of places by my carelessness during the first attempts to do it, and I haven't decided what to do about that.  First impulse is to mask off the shiny bits and just spray, but I worry the mask boundaries will look like dog yack.  Actually I think I will try wiping the dull spots lightly with acetone first.  We'll see.

Er, well, it turns out I own a buffer, with a selection of wheels and compounds.  I forgot. >_<  I'll mask off the lettering with aluminum tape and have a go with that, first.  We'll see, we'll see...

 The next step is wiring up the motor drive panel and its relays with the cable from the drive proper.  I'll need to cut back a bunch of the cable for service loop, and it's very tempting to insert a connector between electronic mounting panel and the housing with cables.  I like serviceability, even though I am  building this so it won't likely need serviceing again until the day it is scrapped for its iron. :(
(my ghost will haunt the person who does this, until I've driven them mad)

 I probably won't start that today because I have pain-brain.  I would probably cut the red wire instead of the blue wire and blow up the speeding bus. 😂 I need to think about the connector when I can think straight.  Oh shoot, I don't think I own a pair of crimper handles any more (the ones that accept different dies).  I'll have to look.  I think my regular ones (I had two) were in the tool kit that got stolen out of a rental property thirty years ago and I never replaced them because they are expensive.  Grump.

 I guess I'm competent enough to install the power cord, umbilical, and fuse holder into the box.  I might replace the plastic power cord with some nice SO cable. I have to give the original cable a hard look.  It doesn't move, so it's not as if it needs to be more flexible than the basic PVC, and the original power cord doesn't actually seem to have any age defects.  Perhaps I will let sleeping dogs lie.

 ~≈{👁}≈~

Does UV light speed polymerization of Boiled Linseed Oil?

 One camp says UV (such as in sunlight or from mercury vapor lamps - the shortwave stuff, not blacklights - hastens the curing of BLO, the other camp says no, only oxygen matters, and not even heat will speed up the process because it is not drying in the usual sense.

 We are Fooling Around with every intention of Finding Out:

100 watts (electrical) of short-wave UV

 I actually put cardboard around it because I wanted to work near it, and use the bench, and even the stray scattered light from these things will give you - and your corneas - a sunburn, with bonus increased risk of skin cancer thirty years later.  Stay away from the evil-looking death light. :D

The strongest peaks in the mercury vapor spectra are:

Wavelength (nm)  Name  Color 184.45 ultraviolet (UVC) - !! 253.7 ultraviolet (UVC) - !! 365.0 I-line ultraviolet (UVA) 404.7 H-line violet 435.8 G-line blue 546.1 green 578 yellow
The brightest peaks are the UV, blue, and green lines, which is why it looks like an eery sort of blue-green color to our eyes.
EDIT: early indications are that UV light may speed polymerization somewhat, it is not a dramatic effect, and not worth obtaining shortwave UV sources for this purpose.
~≈{👁}≈~

No. Gears.

  If I ever need gears for a project I'll either buy a finished gearbox somehow or cannibalize them from some other mechanism.  Fuque making gears! Why?

 To make real gears that will hold up for a while and won't grind the brass or cast iron or whatever into dust, you must first make an involute cutter for the size of gear teeth you want to cut.  That's not horribly complicated if you have a mill (I do), a "spindexer" (I don't), and... an involute cutter cutter.

 Ohhhh, okay, and to make that involute cutter cutter, you just need a heat treat oven and a surface grinder...

 See what I mean?!?  Fark a bunch of making gears! 

 ~≈{👁}≈~

mill power table feed restoration cont'd

shouldn't it be the same
color as the column?

 This is the subject of today's blather; it's the motor drive electronics and housing for the mill's power table feed. -->

 It's defunct, demised, past-on, ceased to be, it's expired and gone to meet its maker.

 These controls were built - at least the electronics were - by a contract manufacturing shop, not Bridgeport themselves.  I'm not sure who made the cast aluminum housing.

 The original electronics weren't well-designed even for the time.  And, as usual, insufficient thought was given to heat rejection. The stock unit appears to have just one bad $10 SCR and it's still not worth my time to repair.

Old n Busted

<-- This is the stock drive electronics.  It wasn't terrible, but it wasn't good.  I guess they assumed every shop would have air conditioning and wouldn't have welders.

 I am replacing this with a modern motor drive (from Minarik) which is half the size and needs a quarter as much heat rejection because it uses newer devices with lower forward voltage drop.  I heat-sunk it to the back panel anyway, with silpad because that's who I am, and my shit doesn't break.  That's literally one of my mottos; don't build things that will break.

 

The New Hotness

 The outside of the housing is kinda grim.  Apparently whoever coated the housing, be they Bridgeport or bridge-dweller, did not know about acid etching primers for aluminum.  And so the paint kinda - fell off - over time, because it wasn't bonded to the aluminum, it was bonded to the (relatively loose) oxide layer.

 The good news is that this has been a solved problem for decades.  I used Rustoleum's "automotive" (I think they mean, for engine blocks) acid-etch primer for aluminum, stainless and mumble.  Then a few coats of conventional industrial enamel.

I used a rattlecan color called "Smoke Gray" because the 'Machinery Gray' (Rustoleum) that Bridgeport used only comes in gallon cans and I am not planning to repaint the whole mill.

notice where the round corners suddenly turn square?

 It feels good knowing you're doing a better job of something than the factory did.  My guess is that Bridgeport had little to do with the production of these, and here's why: the paint on the mill column is .045in thick!  I know this because former users hit the column with work that came loose and knocked chunks of the coating off.  The thickness is impressive.  (that's what he said)  I could measure the depth of the hole in the paint to the surface of the cast iron with a pair of calipers, that canyon's almost deep enough to have an echo. ;)

 I will put on several top coats, baking each in turn, and then I will use a sanding block with 200 grit waterproof paper with water to wet-sand the paint from the logo and from the "on" and "off" labels so they are bare metal.  That's the stock look and I think it's sharp.

 One weird thing I noticed is that the casting mold was modified from its original form, at the bottom rear of the exterior, where the round corners of the housing turn squarish right before meeting the rear.  That feature is not present on the top, and the mold marks there are less smooth than the rest of the casting.  This creates a little bump-out on the bottom rear of the housing for the drive mounting panel, which apparently turned out to be wider than originally planned.  I can easily imagine how that went down.

 And even tho it wasn't originally, and doesn't need to be now, this box will be nearly oil-tight (coolant spray tight by implication) when assembled.  The panel has a foam neoprene seal rather than an o-ring which is the only reason I say, "nearly".  The original cable glands appear to be re-usable, and will get some sealant on their threads... this time.  The new switch and pilot lamp will be oil-tight models, with a silicone rubber boot on the switch.

 Because that's how I roll.

More as it happens.

 ~≈{👁}≈~

mill power table feed (restoration)

 Ugh.

 After decades of owning a mill with a non-functional power table feed, I have finally embarked on the adventure that getting it working again will entail.

 First piece, because it looked easier, has been the motor drive.  I'm dumping the old, dead 1970s tech in favor of a modern drive with the same capabilities, which is half the size and generates one quarter of the heat as the old one.

 The new motor drive requires two industrial control relays, which I thought I had ordered, but Mouser canceled my order while still showing stock. Their customer service rep seemed clueless and unhelpful.  Mouser is kicked out of the clubhouse.  So are Grainger and Newark as their over-enthusiastic security plugin for the web server is falsing on my browser's plugins and won't let me give them money.  Are they hiding some vulnerability?  Better not risk it!

 There are many other competing suppliers.  Even eBay.

So that rankles because even if I wire it all together including the relay sockets, it's no good without the relays.

 There is other work to be done on this box; the pilot lamp and power switch needed replacement, I had the switch on hand, but not a good lamp.  Those are cheap, I'll get one ordered soon. The box itself needs to get sandblasted (done), acid-etch primed (primer not arrived yet), top coated, then the raised letter parts carefully hand-sanded.

 After that box is done comes the more arduous table feed itself.  This consists of a motor, a gearbox, and an electrical box with switches inside and a pushbutton (for fast-traverse) and a speed control knob, like this -->

 The shaft from the operating handle passes through the oil-filled gearbox, into the electrical box (through o-ring seals) in order to actuate the limit switches that live down there.

 I said, the oil-filled gearbox which is located on top of the electrical box.  And apparently they did not have Viton O-rings back then, so guess what always happens eventually?  The oil leaves the gear box and fills the electrical housing.  It was a screwball, in-house, ad hoc design.  A good cleaning and replacing the failed o-rings with modern Viton (fluoroelastomer) rings should solve the problem for a lifetime or three.

 And of course, the wire ends need to be trimmed back and replaced, and the electrical components, which are full of oil.

 And I don't think I'll be able to resist sandblasting and coating the power feed assy either.

 I think I have posted a few pictures of this stuff already.  If not, I'll post some soon.

 ~≈{👁}≈~

minor update on EDC flashlight purchase Fenix LD-22 v2.0

 My flashlight arrived, I noticed some things, and unsurprisingly, I have opinions.

Except for some niggling details, it's housing is more-or-less identical than the LD-20 it replaces*.

Niggling Detail #1: it's no longer possible to stand it on its tail cap.

Niggling Detail #2: I forgot to get my name engraved on it for free, and there is no longer a flat place for that on this housing, you have to settle for it being wrapped around a cylindrical part of the light.  Fun fact: laser engraving does not harm Type III HA - it penetrates the outer HA coating, bleaches the dye out of the dye layer, but leaves the coating above it intact.  That is assuming it's done with CO2 and not a fiber laser.  I don't know for sure about fibers on anodizing.

Niggling Detail #3: it no longer has enough non-round stuff sticking out to prevent the light from rolling off of flat surfaces.  It is virtually round everywhere, and the few small places where it is not, make no difference.

 That's it, that's all my gripes.

 It came with a Li-ion "AA" size battery pack already installed.  I did not know this because who reads fine print, so I ordered what is now an extra one.  WTH, they are only $13.  In this economy, that's not bad.  It came with all the accessories which are typical - a lanyard, strong belt or hat clip, charging cable for the battery.

FWIW, Fenix's manufacturing is in China or Taiwai, but they seem to be controlling prices well.

 Their offices are in Colorado and their product warehouse is in Monticello, Arkansas - 161 Market St, Monticello, AR 71655 - that's where stuff ships from.

 Their showroom is in Littleton, CO.  I've never been there, I don't know how long that's been there.
I think I might drive down there some time just for kicks.

 The new light emits a maximum of 800 lumens in Turbo Mode.  My old LD-20 in Turbo Mode was good for 160 lumens.  Even 800 is not enough for a light bulb to light a room, but for a "mini" size light, it seems staggering at night.  I'm a "more light is always better" kinda guy - usually - so I am  content.  I don't need my EDC light to be so bright it sets things on fire.  Perhaps a camping light should do that. ;)

 ~≈{👁}≈~ 

____
* I need to dig the old light out of the "raygun parts" box to compare them side by side.

If I'd known we were going to jam, I'd have brought my kazoo

 Gothsigh, grumble, and gripe: Operations stalled.  Kinda literally.

 I haven't messed with the printer for a few days because I am avoidant - like, it's a personality disorder, man.

 What I've been avoiding, until today (until some mysterious switch in my head got flipped the other way) is a filament jam in the extruder.  These were unheard-of with PLA, despite a really aggressive (non stock) tensioner and aftermarket... rims (had to say it) I mean extruder wheels.  I could make it quite tight to overcome resistance in the filament feed chain... like ferinstance when the reel is inside a dryer with stupid filament ports, outside the printer tent.  PLA tolerates a lot of squeezing and knurling.

 I switched over to some new TPU replacement (maybe, we'll see) which is a high flexibility PLA.  If you can see it coming now, why couldn't I see it coming then?

 Yup: I forgot to untension the extruder so while trying to load filament, the too-tight drive rolls squished the PLA-flex into funny shapes all around and near the bourdon tube and drive wheel, everywhere but into the tube.  100% my bad.  A right little mess.

 Unfortunately, with the printer inside the tent, it is a neck-craning pain in the, er, neck, to do anything to the extruder - even just adjusting the tensioner, never mind unjamming plastic bits from a tight mechanism and after five minutes of poking and proding with tweezers, dental probes, and a No. 11 X-Acto blade all to no avail, I walked away for a few days.

 There's always other stuff around this dump our majestic mansion that needs doing, after all.  Like all the things we should have done the first ten years we lived here, when we had money.  Cough, but I digress. 

 I really do not want to move the printer, I especially do not want to remove it from the tent unless I simply can't get this done any other way.

 Now it is time to dig into the extruder again (hopefully I can avoid disassembly) then see whether this flex PLA stuff can even be shoved through a bourdon tube at all, or if it requires a direct drive extruder.

 And where did my filament cleaners go? 😬

 I got the extruder wheel unjammed, still need to remove the bowdens full of filament and empty them.  Around then my leg started hurting again, so I'm not going downstairs to fool with the printer until it feels better.

 I'm also not working on shop projects, both for that reason and because my paint won't arrive until tomorrow, and because fucking Mouser canceled my order for bizarre reasons their abysmal customer service either could not or would not explain.  Meanwhile, they still show them in stock, but they won't sell me any.  I am so done with Mouser.  And why is an $11 relay now $30?!?

 ~≈{👁}≈~

light it up forever

 You like flashlights?  Let me tell you about Fenix Lighting.  This is not a promotional thing, Fenix doesdn't know I exist.  I am not a shill.  I am not for sale.

 I got into the brand during the old Candlepower Forums days, when someone was verifying flashlight brands' claimed output lumens vs. actual, measured in a real integrating sphere.  Twenty-five years or so ago, I was sick of my Mini Mag™Lite penlight's pitiful output so I was shopping for a replacement.  LEDs had come along, but Mag Instrument™ were behaving like idiots.  They completely biffed their original LED offerings.  Apparently that plastic focusing cam was their one and only innovation, and it seemed then they weren't interested in any further innovation.

So I was listening to online forums for other people's experiences, and the name Fenix kept coming up. 

 Fenix was the ONLY outfit out of more than twenty tested, that did not exaggerate their output.  Ok, cool as far as that goes.

 Their construction and materials are top-notch.  Type 4 Hard Anodize has kept my old light looking near new even now.  Hardly any scratches, yet I've used it constantly.  The lens is anti-scratch coated, toughened glass, not plastic.  My old light has not got a single scratch in the lens, and I carried it every day for twenty plus years.  I beat the crap out of it.  I used the end of the light as a blugeon (not on a person).  The bezel showed more wear and tear than the lens.

 There is some choice as to switch location, depending on model.  That hard-coated, toughened-glass lens won't scratch without a concerted effort and they are nearly unbreakable.

 I ended up with a EDC "penlight" and a bike head light.  They are both adequately bright when turned up, and like any other light, they have several light levels.  Both have seen heavy use and have held up well.  I like how they work.  The bike light has a warning light on its indicator which lets you know a recharge is required, before it will no longer run on high.  When I've neglected it, and gone for a ride when it needs a charge, I can usually finish my ride even with the light on full/high. Good design.

 I've had them for more than twenty years now, and had occasion to call in for help once with the flashlight.  Their customer service was excellent.  Other uses have reported similar.

 I've used both heavily for a long time, and it's now time to buy a new EDC "pen" light, because the electronics module finally gave up the ghost.  I broke a piece of unimportant glamour plastic off the bike light, but it continues to work fine in all weather conditions, hot, cold, and/or wet.

 To replace the penlight, I picked out another "old fashioned" (long and slender, when the recent trend has been toward short and fatter because 18650s exist) unit that is basically an updated version of my old light.  I like long slender penlights because they fit nicely in my Nite-Ize EDC belt pouch. (Nite-Ize - also recommended product)

 The new one runs off of either your own AAs (disposable or rechargeable) or Fenix's Li-Ion AA-shaped battery pack.  No charger needed, the battery pack has a USB-C connector, and there is a status LED shining through a hole in the positive button contact which changes from red to blue when the pack is charged.  The fact that they fit the entire charging circuit and a connector into a AA battery pack kinda impresses me.  It's not the flying cars so much as the trivial advances in tech that get me. :)

 OBTW, for cylindrical cells of all types smaller than C, I also recommend the NiteCore Intellicharger i4, which will automatically identify the cell chemistry, and will charge any rechargeable battery which will fit into it:

Li-ion/IMR/LiFePO4: 10340, 10350, 10440, 10500, 12340, 12500, 12650, 13450, 13500, 13650, 14350, 14430, 14500, 14650, 16500, **16340(CR123)**, 16650, 17350, 17500, 17650, 17670, 18350, 18490, 18500, **18650**, 18700, 20700, 22500, 22650, 25500, 26500, 26650

Ni-MH(NiCd): AA, AAA, AAAA, C, D

Q: OK, that's all well and good, they make nice lights, I get it, but so what? Surely other makes are good as well?

A: Yes but: 25% Military Discount and free shipping.  And other lights lie about their light output.  Fenix themselves do not crow about this, it's just something the flashlight fetishists noticed.

LD22 V2.0 light: $65.95
ARB-L14 battery: $12.95
military discount: -19.72 (25%!)
shipping: FREE (on any order over $25)
total: $59.18

 They use either GOV-X or ID.ME to verify your vet status.  Recommended company, recommended products.  They are also very popular, so I would not buy their stuff via Amazon or other resellers because those are very likely to be counterfeits -- just buy direct from the Fenix web site.

 ~≈{👁}≈~

ABL - All 'Bout Leveling (3DP)

 This came up in the Creality Discord a while back, I've been meaning to mention it. problem: "I can't get my bed leveling system to work. I am using a Prusa style sensor with an Ender 3V2. (etc etc) " The Enders have glass beds. The Prusa sensors are inductive, which cannot detect a glass bed. Inductive sensors only interact with metals.
An inductive sensor will operate on an Ender, but you would need to change to a metal bed (ugh, why), or better, change to a capacitive sensor, which will detect any surface. It turns out that a mechanical switch like a microswitch, or an E/M sensor attached to a mechanical switch, has the best accuracy and repeatability of every sensor that multiple people have tested. Amazingly, in this at least, multiple testers' results all agreed. Except for the one fellow who built a horizontal test jig for stuff that works vertically against gravity (WTF OOPS) and sooprahz, sooprahz, sooprahz! All his results for the mechanical sensors were worse than everybody else's... >_< But anyway. IOW, if you have a BL Touch or equivalent, you're good, you can probably stop thinking about bed leveling upgrades. ;)

Just as an aside, I have heard from some folks that were clearly unclear about how ABL works. This is understandable. It can be non-intuitive.

So, let's take my Ender 3V2 for example. It has a glass bed, but I have a mechanical 'BL Touch' type leveling sensor. Not only do we install a physical sensor on the hot end carriage, we must also install a firmware update to the little microSD card that runs the printer (in my case) because we are adding functionality to the machine which it did not have when it left the factory. How the install is done, the firmware used, etc, can vary from printer to printer, and it's off topic to try to explain several different ways, so I won't try.

Once the tech is installed, there is either a new menu or a menu item, in your printer's GUI. You use this to "scan" the bed at a selected number of points. Then the firmware moves about the bed, measuring the height of each point. This creates a new "mesh" which is simply a digital map of the bed's humps and valleys. You could just use this information to manually level the bed, and to a degree, I do this, just to get the bed to within spitting distance of level.

But that is not where the beauty of ABL lies.

Because the mesh is also used by the firmware while printing; it moves the head (or bed, depending on printer - in the Z axis anyhow) up and down as it prints, to follow the real contours of the bed as it is, leveled or not! This is dynamic virtual leveling, it is bad-ass, and you want it.

Even if your bed is so messed up it will never be level, ABL will make that problem go away. Scanning for a new mesh if fully automated, just click on a menu item, it takes only a few minutes, and you should only need to do this when you have moved or bumped your printer, cleaned the bed vigorously, etc. You won't be doing it all the time.

This action is faded out as the print continues and for most prints, it can be faded out completely after 10 or 20 layers. That's usually a simple setting in the printer's control panel, under the ABL menu item. That's how the obviously-distorted first few layers transition to normal, straight, printing. The first few layers may be distorted slightly (in the Z axis) but because they follow the bed's real contours perfectly, first layer adhesion is vastly improved, and that's really what this is all about. Keep spaghetti on the dinner table where it belongs!

So, a bed could be really messed up, to the extent that you couldn't even print on it without ABL, yet with ABL, it will print, and it will print just fine. There are demos of this on YouTube. My own Ender 3V2 has a leveling screw at each corner, but the ABL mesh reveals that no matter how much manual leveling I try, in an effort to get the four corners and center to more-or-less agree, and it's not quite possible.

ABL permits you to have a perfect virtual bed, even if the real bed is not perfect.

Out of all the upgrades and tweaks I've made to my humble bed-slinger, ABL has been the best bargain. 

 ~≈{👁}≈~

If you need to make a complex part, BUT you're in a submarine 500 feet down, you definitely need one of these machines

 I'm not suggesting anyone try to buy one of these, I just think it's an absolutely fascinating machine.

https://www.lathes.co.uk/hommel/
(the site needs white space and he's wordy - too much detail core-dumped at once IMNSHO; it would benefit from a more modular / chapterized approach.)

An extraordinary machine, built to very high standards for accuracy and precision.  By using the base and column to mount combinations of a headstock, tailstock, compound slide rest, T-slotted boring table, rotary table (and numerous other accessories) the assembly could be used as a plain-turning or screwcutting 3.5 inch (90 mm) centre height by 14 inches (360 mm) between-centres lathe, high-speed sensitive drill, vertical and horizontal miller, jig borer, shaper, gear cutter, grinder, engraver, slotter, metal saw or surface table.

A Smithy can only *wish* it had a UWM in its lineage -- these things are special and unique with far greater versatility, precision, and accuracy, vs. a Smithy.  I'd be willing to bet it could be set up to be a drill / milling cutter sharpener.  Just need the right wheel.

The UWG1 (alone, not counting all accessories!) is small enough to fit on top of your shop's beer fridge.  The footprint is smaller than a Smithy.

The UWG2 is the same machine but larger footprint/capacity.  All accessories and tools swap between the two sizes of machine.

 These machines were designed to be an "Eierlegende Wollmilchsäue" or, literally translated, "egg-laying wool-milk-pig" - a machine that could do everything.
(everything that can be done in a tight space, that is - they went onto warships and submarines, even airplanes and ground mobile repair units.)

They are not perfect - warts and goblins are detailed at the page linked above.

There is a German language site with more information, but it's insecure and has eight or nine open vulnerabilities so I can't recommend it. :(

The machines themselves are breathtakingly expensive when they appear on the market at all, and there are a LOT of accessories in the complete tooling set.  Most machines are not sold with everything present.  A single base plate part is on Ebay for $700.  Complete machines can be found more often in Europe, where they sell for $3500 - $5000.  You can buy  a brand new full sized Chinese knee mill with DRO, vise, oiler, everything turnkey, for that price.  You can have it shipped from China to your door for another grand or less.

 So paying that amount for a weird little machine that dates from before WW-2 might be a stretch for most people I know.  Definitely a niche market. :)

And changing it from a mill to a lathe and so on is fiddly and time consuming.

 ~≈{👁}≈~