Grind Your Welds With Pride, If That’s The Way You Do It

Grind Your Welds With Pride, If That’s The Way You Do It
By Jenny List

To grind or not to grind? What a question! It all depends on what you’re really trying to show, and in the case of welded joints, I often want to prove the integrity of the weld.

My ground-back piece of welded tube. Eagle-eyed readers will spot that the grinding reveals a weld that isn't perfect.
My ground-back piece of welded tube. Eagle-eyed readers will spot that the grinding reveals a weld that isn’t perfect.

Recently, I wrote a piece in which I talked about my cheap inverter welder and others like it. As part of it I did a lower-current weld on a piece of thin tube and before snapping a picture of the weld I ground it back flat. It turns out that some people prefer to see a picture of the weld bead instead — the neatness of the external appearance of the weld — to allow judgment on its quality. Oddly I believe the exact opposite, that the quality of my weld can only be judged by a closer look inside it, and it’s this point I’d like to explore.

What Lies Beneath

A weld is a balance of several different factors to use an arc to melt a steel filler into the gap between two pieces of steel, shielded from the oxidation by some kind of non-reactive cover. In the case of my stick welder the cover is the molten and gaseous phases of the flux covering the welding rod as they are heated by the arc, while in the case of a MIG welder it is whatever shield gas the operator is using. As the operator I have control of a few factors such as the current setting on the welder, the size of the arc I draw, and the speed at which I move it across the metal. These affect the temperature and thickness of the molten metal and the integrity of the shield, which will in turn affect the metallurgy of the resulting weld. It may be brittle, it may have inclusions of flux, it may have internal tensions that could cause cracks, or it may not have proper adhesion to the pieces of metal it’s trying to join.

I had a lot of fun creating these awful welds for this piece. Mostly they are the result of too long an arc length.
I had a lot of fun creating these awful welds for this piece. Mostly they are the result of too long an arc length.

Among Hackaday’s readership will be a few elite-level metalworkers whose welds are works of art. I may sometimes approach that level of skill with soldering, but when it comes to welding I am not one of them. I suspect I share this with many readers, while I have a basic handle on the job I am still pretty inept. I’ll get the welding current wrong, I’ll sometimes get my arc length wrong, and I’ll move across the work at the wrong speed. Despite that, I will often manage to produce welds with a creditable outward appearance once I’ve knocked off the flux, but which may still conceal fundamental flaws beneath them.

Many people who have received professional instruction on welding will have had their teachers devote countless hours in the pursuit of drawing a perfect bead. It’s an entirely appropriate skill to acquire because by repetition it teaches those skills of maintaining arc length and cover integrity that are the key to producing consistently good welds, but it shouldn’t lead to the conclusion that it is the bead that’s important rather than those arc skills. An inept welder can produce a surprisingly good-looking bead that conceals a less pretty weld, I should know this because I have done it.

Do You Have Ten Thousand Hours?

My welds will probably never achieve as neat a result as this automated TIG welder.
My welds will probably never achieve as neat a result as this automated TIG welder. Ytrottier, [CC BY-SA 3.0]

There’s a saying that to master a skill you need to have ten thousand hours experience of it, which puts me in an odd position with respect to welding. I grew up around a working blacksmith’s forge, but with a dad whose livelihood depended on his welders. My welding time when I was younger was thus surprisingly sparse and only when supervised by my dad, which means though having spent a lot of time around welding I have never achieved anything like those ten thousand hours. What he did teach me though was about all the things that could go wrong with a weld, how the metal could burn and how oxides and flux could be integral to an otherwise decent-looking weld produced in my inept hands. He thus taught me to take a look at my welds and grind them back a little, because the angle grinder is a harsh tutor that spares no feelings as it exposes the shortcomings of your work.

I thus make full use of the angle grinder when I weld, though I don’t always grind my welds as far back as I did with that 1mm tube. The sight of good metal exposed under that bead is the confirmation that the weld is a good one, and if I see any anomalies it’s my cue to take it back all the way and try again. If you take a look again at the tube you’ll see just such a flaw in the join, if it wasn’t simply a test to see whether such a thin piece of stock was possible with that welder I’d be doing it again.

I think professional welders would universally consider this welding gear a toy — as an Easy-bake oven is to a professional baker. So for you pros out there, let’s see those pristine beads, for those weekend welders, by all means, grind them down and show us what they’re really made of.

Man welding header image: Jorge Barrios [Public domain].

October 22, 2019 at 07:01PM
via Blog – Hackaday


Creepy Halloween Doll Might Make You Betsy Wetsy

Creepy Halloween Doll Might Make You Betsy Wetsy
By Kristina Panos

If you want to terrify your neighborhood this Halloween, you might go for the old standbys like skeletons or zombies. But you don’t have to go gory to find glory. Consider the talking doll. Those things are creepy enough already, right? Well, [cabuu] says no, the doll should be animated with servos and have remote control. She should still be able to talk, just not when you expect her to.

Forget pushing on her stomach, ’cause Baby’s got a Wemos D1 mini and her own Blynk app now.  A set of sliders in the app control a micro servo that animates her eyes, and another servo that twists her head from side to side. Her head doesn’t go all the way ’round, but that’s probably for the best. There are preset fright modes [cabuu] can set and forget until she springs to life via motion sensor.

We particularly like the bracket [cabuu] designed and printed that joins the eyeballs with the servo, along with his clever use of printed mate brackets to hold the servos in place within the head. If you think you can stomach it, there’s a demo video after the break. Stay tuned for total doll dissection after that as [cabuu] builds and inserts the terrifying tidbits.

We love hacks that combine innocence with insanity. Have you ever seen Thomas the Tank Engine singing Rick Astley?

October 22, 2019 at 05:30PM
via Blog – Hackaday

Developing Guidelines for Sustainable Spaceflight

Developing Guidelines for Sustainable Spaceflight
By Tom Nardi

In the early days of spaceflight, when only the governments of the United States and the Soviet Union had the ability to put an object into orbit, even the most fanciful of futurists would have had a hard time believing that commercial entities would one day be launching sixty satellites at a time. What once seemed like an infinite expanse above our heads is now starting to look quite a bit smaller, and it’s only going to get more crowded as time goes on. SpaceX is gearing up to launch nearly 12,000 individual satellites for their Starlink network by the mid-2020s, and that’s just one of the “mega constellations” currently in the works.

Just some of the objects in orbit around the Earth

It might seem like overcrowding of Earth orbit is a concern for the distant future, but one needs only look at recent events to see the first hints of trouble. On September 2nd, the European Space Agency announced that one of its research spacecraft had to perform an evasive maneuver due to a higher than acceptable risk of colliding with one of the first-generation Starlink satellites. Just two weeks later, Bigelow Aerospace were informed by the United States Air Force that there was a 1 in 20 chance that a defunct Russian Cosmos 1300 satellite would strike their Genesis II space station prototype.

A collision between two satellites in orbit is almost certain to be catastrophic, ending with both spacecraft either completely destroyed or severely damaged. But in the worst case, the relative velocity between the vehicles can be so great that the impact generates thousands of individual fragments. The resulting cloud of shrapnel can circle the Earth for years or even decades, threatening to tear apart any spacecraft unlucky enough to pass by.

Fortunately avoiding these collisions shouldn’t be difficult, assuming everyone can get on the same page before it’s too late. The recently formed Space Safety Coalition (SSC) is made up of more than twenty aerospace companies that realize the importance of taking proactive steps to ensure humanity retains the unfettered access to outer space by establishing some common “Rules of the Road” for future spacecraft.

No More Secrets

The first directive suggested by the SSC in their document “Best Practices for the Sustainability of Space
Operations” is simultaneously the easiest and hardest to accomplish: establishing an open dialog between spacecraft owners and operators about what their respective craft are actually capable of. The aerospace industry is notoriously secretive, but the SSC argues that when it comes to safety, companies are going to have to divulge more about their hardware than they might traditionally be comfortable with.

According to the document, it’s critical that all operators are made aware of each craft’s maneuvering capabilities and intended “Flight Plan” should evasive action be required. This will allow rapidly determining which vehicle has the right-of-way in the event of a close pass. Just as smaller boats are expected to steer clear of larger and less maneuverable ships, the more agile spacecraft should logically be the one to initiate evasive maneuvers if all other factors are considered equal.

Perhaps predicting pushback from aerospace firms that are reluctant to detail the capabilities of their proprietary hardware and software, the SSC clarifies that the exchange of this information must be done in good faith and performed in such a way that it respects the intellectual property rights of all parties. In other words, don’t expect this information to be listed on Wikipedia anytime soon; while it’s important for spacefaring agencies and operators to know what each other’s craft are capable of, it’s not necessarily something the public needs to be privy to.

Voting With Your Wallet

The second directive is aimed at those purchasing launch services from commercial providers, and states that customers should stipulate in their contract that the booster which carries their spacecraft to orbit must make every effort to responsibly dispose of itself after successful completion of the mission. In other words, once the upper stage of the rocket has deposited your satellite into the intended orbit, it then must adjust its own course to either burn up in the atmosphere or continue on to deep space.

SpaceX concept for a reusable upper stage

Specifically, the SSC says that launch vehicles delivering spacecraft to low Earth orbit should demonstrate at least a 90% success rate for deorbiting themselves post-mission. In the case of spacecraft heading towards the far higher geosynchronous orbits, then the launch vehicle should be able to propel itself into a deep space orbit that will ensure it doesn’t return for a minimum of 100 years.

By asking operators to take post-mission disposal of launch vehicles into account when booking passage for their spacecraft, the SSC is looking to effectively incentivize the development of more robust boosters. If customers won’t use your rocket because it can’t safely deorbit itself after payload separation, you’ll suddenly have a very compelling reason to research and implement this capability. Taking that idea a step farther, it might even spur the development of reusable upper stages, a concept SpaceX has been flirting with for years.

Safety by Design

The rest of the document, indeed the bulk of it, has to do with best practices for those designing and manufacturing spacecraft. These recommendations include both active and passive steps that can be taken to give a spacecraft the best chance of survival in an increasingly crowded airspace, but also ways to make sure the vehicle doesn’t inadvertently become a piece of space trash itself.

It seems like it should go without saying, but one of the first recommendations is that spacecraft be thoroughly tested on the ground to prevent a Dead-On-Arrival (DOA) condition. But the reality is that testing can be expensive and time-consuming, and the new breed of low-cost miniature satellites are often designed and assembled on an extremely tight budget. You don’t have to look long to find examples of CubeSats that failed as soon as they were released from their carrier rockets.

The SSC also recommends that satellites be deployed into a lower, less trafficked, orbit until a self-diagnostic test confirms they are fully functional. This means it will take more time and propellant to get a satellite into its intended orbit, but the advantage is that if the craft ends up DOA, it’s less likely to be a hazard to navigation. In a low orbit, even a non-functional spacecraft will dip back into the atmosphere and burn up in a relatively short period of time. SpaceX utilized this approach for the first Starlink launch, and in the end found that three of the sixty satellites never passed their self-check and failed to move on from their initial “parking” orbit.

Finally, the SSC agrees with the European Space Agency that future satellites should utilize “Design for Demise” or D4D. These are design principles which ensure that even a non-functional satellite will deorbit itself and completely burn up in the atmosphere. Techniques include spacecraft shapes and orientations that maximize atmospheric drag, and the use of thermal reactive adhesives which will release as the vehicle heats up during reentry.

Actions Speak Louder than Words

While the Space Safety Coalition’s list of best practices for designing and operating spacecraft in Earth orbit are reasonable and in many cases logical, the group has no authority to actually enforce them. As such, any company that decides to follow these rules does so of their own volition. Smallsat launch providers Virgin Orbit and Rocket Lab have already decided to endorse the document, as have some companies which operate fleets of satellites, such as SES and Iridium.

But while it’s encouraging to see early support from industry players, there are a number of very high profile names currently missing from the list of endorsees. First and foremost is SpaceX, which is not only the world’s premiere commercial launch services provider, but in the future will operate the largest fleet of individual spacecraft ever assembled. Also absent are major aerospace companies such as Boeing, Northrop Grumman, and Lockheed Martin, firms that are responsible for designing and building a vast array of civilian and military satellites.

If the industry has any hope of getting a handle on critical elements of close-quarters spaceflight like collision avoidance and spacecraft disposal, it’s going to take more than a handful of companies signing their names to a non-binding agreement. Larger companies will need to be persuaded to align themselves with the ideals of the Space Safety Coalition soon, as the ever increasing pace of orbital launches means that we might not have a lot of time left before things reach a tipping point.

October 22, 2019 at 04:00PM
via Blog – Hackaday

Robot Joints Go Modular With This Actuator Project

Robot Joints Go Modular With This Actuator Project
By Donald Papp

[John Lauer] has been hard at work re-thinking robot arms. His project to create modular, open source actuators that can be connected to one another to form an arm is inspiring, and boasts an impressively low parts cost as well. The actuators are each self-contained, with an ESP32 and a design that takes advantage of the form factors of inexpensive modules and parts from vendors like Aliexpress.

Flex spline in action, for reducing backlash

Each module has 3D printed gears (with an anti-backlash flex spline), an RGB LED for feedback, integrated homing, active cooling, a slip ring made from copper tape, and a touch sensor dial on the back for jogging and training input. The result is a low backlash, low cost actuator that keeps external wiring to an absolute minimum.

Originally inspired by a design named WE-R2.4, [John] has added his own twist in numerous ways, which are best summarized in the video embedded below. That video is number three in a series, and covers the most interesting developments and design changes while giving an excellent overview of the parts and operation (the video for part one is a basic overview and part two shows the prototyping process, during which [John] 3D printed the structural parts and gears and mills out a custom PCB.)

Is your interest piqued? If so, head over to the GitHub repository for all the details, including Fusion 360 models, PCB design files, and a complete bill of materials with online sources. While not all robotic actuators require motors or gears, one thing that is undeniable is that the availability of 3D printing, PCB fabrication, and access to electronic components has been a tremendous boon to projects like this.

October 22, 2019 at 01:00PM
via Blog – Hackaday

A Spectrum Analyzer for the Smart Response XE

A Spectrum Analyzer for the Smart Response XE
By Tom Nardi

Remember the Girl Tech IM-me? It was a hot-pink clearance rack toy that suddenly became one of the hottest commodities in the hacking world when it was discovered they could be used for all sorts of radio frequency shenanigans. Now they go for triple digits on eBay, if you can even find one. Well, we’re probably about to see the same thing happen to the Smart Response XE.

Thanks to the work of a hacker named [ea], this cheap educational gadget is finally starting to live up to the potential we saw in it back when a teardown revealed it was powered by an Arduino-compatible ATmega128RF chip. With a big screen, a decent QWERTY keyboard, and integrated wireless hardware, it seemed obvious that the Smart Response XE was poised to be the next must-have repurposed piece of kit.

Though as it turns out, [ea] isn’t using the device’s built-in wireless hardware. Step one in this exceptionally well documented and photographed project is to tack a CC1101 transceiver module to the SPI pins on the ATmega128RF. Then with the appropriate firmware loaded up, that nice big screen will show you what’s happening on the 300 MHz, 400 Mhz and 900 MHz bands.

But the fun doesn’t stop there. With the CC1101-modified Smart Response XE, there’s a whole new world of radio hacks you can pull off. As a proof of concept, [ea] has also included a POCSAG pager decoder. Granted the RTL-SDR has already made pulling pager messages out of the air pretty easy, but there’s something to be said for being able to do it on something so small and unassuming.

If you can’t tell, we’re exceptionally interested in seeing what the community can do with the Smart Response XE. At the time of this writing, the going rate on eBay for a good condition unit looks to be about $10 USD, plus the $3 or so for the CC1101 module. But the prices went through the roof when we first posted about it, so get them cheap while you still can.

[Thanks to bburky for the tip.]

October 22, 2019 at 10:00AM
via Blog – Hackaday

Vintage Transistor Radio Gets Internet Transplant

Vintage Transistor Radio Gets Internet Transplant
By Lewin Day

The invention of the transistor revolutionized radio, allowing receivers to be made far more compact and portable than ever before. In the middle of the 20th century, the devices exploded in popularity, and pocket transistor radios took the market by storm. [MisterM] had fond memories of such times, and when he found a 1970s Flirt radio at a car boot sale, it led to a cute little build.

The radio was stripped of its original hardware, with [MisterM] preferring internet radio to the terrestrial variety. In its place, a Raspberry Pi Zero was installed. This was fitted with a cavalcade of off-the-shelf modules to make it fit for pumping out the tunes. A Speaker PHAT was used for audio, while an Adafruit Micro Lipo board handled battery charging and a Pimoroni Lipo Shim served as the power supply. All this was bundled up inside the original casing.

The radio’s controls are a neat hack. The original volume and tuning dials were removed, sliced up, and glued onto two lever microswitches. This allows them to act as buttons instead. A new power switch was installed behind the original, and the Speaker PHAT’s LEDs were placed behind the tuning dial to act as a rudimentary display.

It’s a tidy build that repurposes an attractive vintage artefact into a device of the internet age, while remaining externally the same. [MisterM] has form in this space, with his vintage radio console build and video doorbell being particular highlights. Video after the break.

October 22, 2019 at 07:00AM
via Blog – Hackaday

An Algorithm For De-Biasing AI Systems

An Algorithm For De-Biasing AI Systems
By Sharon Lin

A fundamental truth about AI systems is that training the system with biased data creates biased results. This can be especially dangerous when the systems are being used to predict crime or select sentences for criminals, since they can hinge on unrelated traits such as race or gender to make determinations.

A group of researchers from the Massachusetts Institute of Technology (MIT) CSAIL is working on a solution to “de-bias” data by resampling it to be more balanced. The paper published by PhD students [Alexander Amini] and [Ava Soleimany] describes an algorithm that can learn a specific task – such as facial recognition – as well as the structure of the training data, which allows it to identify and minimize any hidden biases.

Testing showed that the algorithm minimized “categorical bias” by over 60% compared against other widely cited facial detection models, all while maintaining the same precision of detection. This figure was maintained when the team evaluated a facial-image dataset from the Algorithmic Justice League, a spin-off group from the MIT Media Lab.

The team says that their algorithm would be particularly relevant for large datasets that can’t easily be vetted by a human, and can potentially rectify algorithms used in security, law enforcement, and other domains beyond facial detection.

October 22, 2019 at 04:00AM
via Blog – Hackaday