Broken screw removal.

[Rupunzell]

Took apart the nose on the x1/20 to fix corrosion in the battery box area. The side nose trim pieces had to come off to gain access to that area of the nose. Both M6 screws snapped off with a modest tug of a 10mm wrench even after being soaked in penetrating oil for a few days. Both screws had to be removed without damaging this part as trim parts for the x1/20 is difficult to get. Adding to this problem, the part is made of plastic (could be bakelite or similar brittle plastic).

Typical method would be to set the part up in the BP mill, center the broken screw, flatten the rough end with an end mill so the center drill has a chance to drill on center, progressively drill out the broken screw with a left hand drill until it spins out or the remains of the broken screw hollowed out enough to be un-wound like a spring.

Initial try to set the part up in the mill using an ER32 collet same diameter of the threaded boss,
not gonna work due to the odd shape of this part. Part being plastic would not enjoy being clamped
in a milling vise. Padding does not work as it will apply pressure in a way that would collapse the part.

Made a drill guide on the lathe for a# 29 drill to start.

Drilled sort of on center, which is most of what could be hoped for.

Left hand drills used to progressively enlarge the hole in the broken screw. Last left hand drill
size is# 8 or 0.199" which is about the minor diameter for M6 x 1.00 thread.

The hole was slightlyoff center and the residual bits of the broken screw still corroded in place.
Goal is to thin the wall of the broken screw enough to collapse it for removal. almost happened, but not there.

Ended up installing a Helicoil to fix the broken screw. Difficult to do due to the steel insert moulded
into the part can easily bust off. Folks who designed this part was wise enough to make these
inserts robust enough to get this done without serious damage.

The second broken off screw was subjected to much the same procedure. Difference being part
of the residual broken off screw got snagged by the second to final size left hand drill. This was
enough to allow digging out most of the residual broken screw bits and a M6 tap applied to
clean up the threads. Original threads preserved on this one.

Machine tool and related to do this fix:

Threaded parts are high precision items that are often taken for granted and if they bust off, the
difficulty to repair can be more than most would know or suspect. The greatest of care should
be applied to prevent busting off threaded fasters. This is why re-using threaded fasteners usually
never happens and what is used to put it all back together is new of equal or better. Big fan of
Anti-sieze or Loctite of the proper variety as needed.


Bernice

 

[kmead]

There are few things I fear more than a broken fastener when doing a repair. Fixing the seemingly minor fastener issue takes longer than all the repairs that were intended to be done.

Nice tutorial.

 

[Rupunzell]

Guess where the tools and skill required to do this came from... fixing LOTs of busted threaded fasteners .

Having done SO much of this over the years, this is one of many reasons why re-using threaded fasteners is mostly prohibited. It is why threaded fasters are purchased in quantity and replaced in quantity.

Most wrencher folks do not appreciate the precision involved with threaded parts, while folks who do serious machine work ALL do on a intimate level as they are often making threaded parts that must meet a specification.

Having said this, one of the eye rollers are folks who do the strip and re-plating thing with original fasteners as part of an "original" restoration. This is a potential problem waiting to happen. Most get away with it due to the margins designed into a threaded fastener joint.


Bernice

There are few things I fear more than a broken fastener when doing a repair. Fixing the seemingly minor fastener issue takes longer than all the repairs that were intended to be done.

Nice tutorial.

 

[motoTrooper]

Mills can be nice for stuff but often when the bit begins to wander it can be helpful to cant the workpiece by hand quickly to guide it back to center. I had to do the same process to the same pieces on mine. Fun!

 

[Janis]

These are helpfull to get drill into center.

 

[Dan Sarandrea (Phila)]

Your average laptop is held together by 16-24 very miniaturized versions of the molded-in thread insert that we see in Bernice's post. The difference is that in a laptop, failure is always from material fatigue in the plastic boss surrounding the molded-in insert, as opposed to weather-induced corrosion of the two threaded metal components. The fastener bosses that most often break are the ones that bear the stress of the lid hinges.

Depending upon the availability of replacement parts (often good availability and inexpensive), it's often cheaper and faster to just order up a new or good used replacement lid, base, etc. When none can be found, I have ground off the remnants of the broken boss, glued a new boss made from stock plastic rod into place, and then drilled the new boss for a new insert. The hard part is matching the replacement plastic to the existing plastic so the glue can bond correctly.

In your case Bernice, if broken screw extraction went wrong, it looks like there is plenty of meat on the plastic boss into which the threaded inserts have been molded, perhaps the existing insert could have been removed and another inserted. McMaster seems to have a good selection of stainless inserts here: https://www.mcmaster.com/inserts-for-plastics/

 

[Rupunzell]

Typical lap top and such are made of a flexi ABS type plastic. The side trim pieces on the Lancia is made of a hard plastic that can easily crack and shatter. If that mounting boss cracked or shattered during this repair, that repair would be FAR more complex and difficult to possible scrapping this part due to damage incurred.

Bernice

 

[Cratecruncher]

The molded in bosses in our electronics are often designed with no internal threads. The screw taps it's own thread during assembly creating a bunch of hoop stress. I've found as the plastic shrinks and becomes embrittled over time those bosses split and shatter. I collect old HP calculators and they are notorious for this problem. I've had some success using fiberglass fibers spooled around the posts and wicked with thin cyanoacrylate for reinforcement. The fiberglass strands disappear and the reinforcement is hardly noticeable. I think it might be scaled up for automotive use using epoxy in place of the ca if there were a worry about damage to the plastic during repairs.

 

[Rupunzell]

Famous and very well known problem in the hp 8640 signal generator, plastic gears that crack. The plastic gears have a metal hub. Plastic continues to shrink with the passage of time. Sharp-rough edges on the moulded plastic gears are built in stress risers causing the plastic gear to crack.. The vast majority of them have done this

Same happens on Tektronix knobs. They can be repaired by removing the plastic knob from the aluminum hub then turning down the aluminum hub to the current diameter of the shrunken plastic knob, then apply epoxy to put it all back together..

Speaking of hp calculators... few of these have been with me for years.. some of the buttons have
"softened" from being pressed SO many times.

Bernice

I've found as the plastic shrinks and becomes embrittled over time those bosses split and shatter. I collect old HP calculators and they are notorious for this problem.

 

[dllubin]

The molded in bosses in our electronics are often designed with no internal threads. The screw taps it's own thread during assembly creating a bunch of hoop stress. I've found as the plastic shrinks and becomes embrittled over time those bosses split and shatter. I collect old HP calculators and they are notorious for this problem. I've had some success using fiberglass fibers spooled around the posts and wicked with thin cyanoacrylate for reinforcement. The fiberglass strands disappear and the reinforcement is hardly noticeable. I think it might be scaled up for automotive use using epoxy in place of the ca if there were a worry about damage to the plastic during repairs.

I never thought of my self as a collector of HP calculators but I just opened my desk drawer and found four of them including HP41C serial number 67 from the first prototype run. They all are still in one piece but I'll make a note of that problem.

I had the problem with the inserts holding the hinges of my 17" HP laptop pulling out. I solved the problem by making thin metal plates to go over the inserts to spread the load and then using a whole lot of JB Weld to hold it together. The root cause may also have had something to do with the spring force on the hinges being way too tight. It was like trying to open a bear trap! I was able to get the tension down, and the fix has lasted for many years.

 

[dllubin]

Famous and very well known problem in the hp 8640 signal generator, plastic gears that crack. The plastic gears have a metal hub. Plastic continues to shrink with the passage of time. Sharp-rough edges on the moulded plastic gears are built in stress risers causing the plastic gear to crack.. The vast majority of them have done this

Same happens on Tektronix knobs. They can be repaired by removing the plastic knob from the aluminum hub then turning down the aluminum hub to the current diameter of the shrunken plastic knob, then apply epoxy to put it all back together..

Speaking of hp calculators... few of these have been with me for years.. some of the buttons have
"softened" from being pressed SO many times.
View attachment 33844


Bernice

At one point, there was somebody manufacturing replacement gears for the 8640. They may have been in Europe.

 

[Rupunzell]

Brass repro gears for the hp 8640(B) are made in India.. They appear on eBay and else where. Have an 8640B with precisely this problem but no real incentive to fix it as the hp 8662A has taken it's place for now. The plastic gear fiasco was the design folks act to do "cost reduction" as requested by management. Have one of the militarized hp 8640's that was used as an environmental test unit at hp. It's been tested LOTs, but works fine, no problem with gears. Possible the Mil version does not have plastic gears, not gone inside to check.

Trivia, hp wanted to replace the hp 608 that was designed by Art Fong long ago. Turns out that Art Fong design using fire bottles had very low phase noise due to the LC oscillators and fire bottles. This is how the hp 8640 ended with with a cavity oscillator then divided down with ECL logic then filtered to get reasonable sine wave distortion. It was not until the hp 8662A & 8663A that RF generators from hp achieved low phase noise performance. Oh, other 8640 problem area, output amplifier. There is a story to this, but it is WAY off topic here.


Bernice

At one point, there was somebody manufacturing replacement gears for the 8640. They may have been in Europe.

 

[Rupunzell]

Speaking of hp calculators, the mechanical engineer who designed the buttons, how the buttons functioned and feel, packaging and all lives in Los Altos. Also tinkers with ...cars.
He told me some great stories about designing the hp calculators and Bill Hewlett's reaction to them a more.


Bernice

 

[dllubin]

Interesting. I happened to be living in Los Altos back when I was involved with the calculator programs. The HP35 keyboard was superior to the competition in that it had a great feel with tactile feedback and used a double injection molding process to insure the symbols could not be rubbed off. I once got a tour of the molding operation in Singapore. They were also using the double molding technology for their computer terminal keys. That was back when a dumb terminal could cost as much as a car so the manufacturing cost pressures were not too severe. The PC business changed all that.

 

[Rupunzell]

hp 35 calculator got it's name from 35 keys via Bill Hewlett. Those calculator buttons roll downward at an angle. This was designed based on finger ergonomics with the calculator sitting on a desk with the buttons being presses. The click is feed back to the user's finger to facilitate the button being pressed. This is common to ALL hp designed momentary front panel switches. If you've ever use a item of hp test gear with hp designed momentary switches, this is why.

The double shot molded keys used in hp's first calculators were made in the basement of 1501 Page Mill Road. For any who've visited "hp on the hill" the basement was were mechanical happened. At that time hp purchased an injection moulding machine from Germany to make keys, that machine did not fit the loading doors, it was delivered and installed via an opening in the roof. There was a recall for early hp-35 calculators due to an "error"

Pocket

Bernice

 

[dllubin]

You could always tell you were getting close to the basement from the smell of machine oil.

 

[Cratecruncher]

I never had a problem with buttons on any of my calculators. The card readers seem to all get gummy wheel syndrome but that's an easy fix. More difficult are the shattered screw posts because the clamping force of the screws keeps the boards in contact with each other. A big problem with the 48's is deteriorating solder joints. So easy to fix but the case can't be opened without damaging the calculator. For this and other reasons I don't use any of my HP's much anymore but I've downloaded several emulators onto my android phones. I love it when some one asks if my phone has a calculator. "What is THAT?" they ask. Let me show you RPN....