Immersion Heaters and Potential Fire Risks

Immersion heater terminal box. © Eugene Brennan

So I was doing some gardening for the sister today and she mentioned a smell that was coming from her immersion. This immediately rang alarm bells in my engineering head, knowing what was likely to have happened—So I investigated.

Learn How to Solder: A Complete Guide for Beginners

Image by permission Weller Tools

A Basic Skill When Building Electronic Circuits

Soldering is a fundamental skill essential to learn when building electronic circuits or fixing circuit boards.

Missing Clip on My Quick-Release Skewer

Quick-release axle skewer from a bike. © eugene Brennan 

The cam lever fell out of the housing when I took my front wheel off to change a puncture, half-way between Ballymore Eustace and Harristown last Sunday. I hadn't noticed that the retaining clip had disappeared, possibly rusting off over the last 29 years. It's years since I changed a front puncture. I noticed too that I have five patches already on this front tube. 99% of my punctures are caused by whitethorn thorns. Thanks to the people out walking who offered me a lift. Anyway I had a discussion about forces in the axle with ChatGPT:

Eugene:

You know the skewers used on quick release wheels?

 

ChatGPT: 

Yeah, I know exactly the ones you mean—the quick-release skewers on bike wheels.

They’re that rod that runs through the hollow axle, with:

• a cam lever on one side (the bit you flip open/closed), and
• a nut on the other end (to adjust tightness)

When you flip the lever, it tightens via a cam mechanism, clamping the wheel securely into the dropouts—no tools needed.

A couple of subtle details people often miss:

• The lever doesn’t just “screw tight”—it should leave an imprint on your palm when closing if it’s properly tight.
• The springs on either side (little conical ones) help centre the skewer—narrow ends point inward.
• It’s not the thread that holds the wheel—it’s the cam force when the lever is closed.

Are you thinking about fixing one, replacing it, or did something odd happen with it?

Eugene: 

What type of retaining clip is used to hold on the cam lever? It must have corroded off mine. I haven't changed a front puncture for maybe 10 years, so I didn't notice it missing. It wasn't immediately obvious.

 

ChatGPT: 

Ahh, I know the bit you’re talking about—that tiny piece that quietly disappears over time.

On most quick-release skewers, the cam lever is held onto the skewer shaft with a retaining clip, usually one of these:

• a circlip (snap ring / C-clip) sitting in a groove on the shaft, or
• sometimes a peened-over end (the shaft is slightly mushroomed instead of using a removable clip)

If yours has “vanished,” it was almost certainly a circlip that corroded away or popped off years ago.

A couple of practical points

• Those clips are tiny and fiddly, and not always standardised between brands
• Replacing just the clip can be done if you can match the groove diameter, but…
• In reality, most people just replace the entire skewer—they’re cheap and far more reliable than bodging a new clip onto an old, possibly worn cam

My honest take

If it’s been missing for 10 years, I wouldn’t trust that skewer anyway. The cam faces and shaft could be worn or pitted.

A new skewer costs very little and removes any doubt—especially since that part is literally what keeps your wheel attached 😄

If you want, I can help you identify the exact type/size you need (front vs rear, road vs MTB, etc.).

Eugene: 

But there's a huge force on that cam, so unlikely for the lever to come out (remember limiting force = μRn?)

 

ChatGPT: 

You’re absolutely right to think in terms of friction holding things together—but there’s a subtle mismatch here between what force is doing what.

The clamping force you’re thinking of is captured by:

$F_{\text{friction}} = \mu R N$

That force is between the dropout faces and the hub axle. It’s what stops the wheel from slipping out of the frame.

---

But the clip does a completely different job

The little circlip isn’t resisting that big clamping force at all. It only:

• keeps the cam lever assembly from sliding off the skewer rod
• handles tiny axial forces (basically just gravity, vibration, and handling)

So even though the cam generates a large normal force $N$, that force is internal to the clamping system, not trying to eject the lever.

---

Why yours likely disappeared

Over time:

• corrosion weakens the circlip
• small vibrations + repeated opening/closing
• maybe one partial disengagement

…and eventually it just pops off or disintegrates.

---

A useful way to picture it

Think of it like:

• The cam = a powerful clamp
• The circlip = a tiny “end stop” so the clamp doesn’t fall apart when it’s open

When the lever is closed, everything is compressed together anyway—so the clip isn’t even doing much at that moment.

---

One caution

Even if it feels secure now, without that clip:

• the lever can drift or misalign when open
• you risk losing parts when removing the wheel

So your physics instinct is solid—but it applies to wheel retention, not lever retention.

---

If you want, we can actually estimate the clamping force from a typical cam geometry—that’s a fun little mechanics exercise.

Eugene: 

Yes, but there's an axial load force on the cam that pushes on the cam housing. And that's huge, because you're supposed to wrap your fingers around the lever when squeezing upwards. I probably put 10 kg or maybe 20 kg force or more on the lever with my fingers, and that's magnified by the mechanical advantage. Lever is about 8cm long, so the load can be resolved to midway, say 4 cm. The cam acts at about 5mm, so that's a mechanical advantage of 8 Let's call it 6. So there could be up to 100 kg on the cam, pushing on the axle. a radial force would have to overcome that for the lever to come loose.

 

ChatGPT: 

You’re thinking about it the right way mechanically—but the key is which direction that big force actually acts in.

Your estimate is pretty reasonable:

• ~10–20 kg hand force at ~40 mm
• cam acting at ~5–6 mm
  → mechanical advantage ~6–8
  → axial clamp force easily in the ballpark of 600–1200 N (≈60–120 kgf)

That checks out.

---

The subtle bit: where that force goes

That large force is almost entirely:

• axial along the skewer, and
• compressing the cam faces and dropouts together

It’s not trying to push the lever off the end of the shaft.

When the lever is closed:

• the cam is wedged hard against its mating surface
• everything is in compression
• the system is actually self-locking (that’s the cam design doing its job)

So in that state, the circlip is basically doing nothing at all.

---

When the clip actually matters

The circlip only earns its keep when the lever is:

• open, or
• somewhere mid-travel

In that position:

• there’s little to no clamp force
• the cam isn’t wedged
• the lever assembly can slide axially along the shaft

And the forces there are tiny:

• vibration
• gravity
• a bit of handling

That’s why a flimsy little circlip is enough.

---

Why your “it shouldn’t come off” intuition still makes sense

You’re right in a practical sense:

• If the skewer is closed, the lever is very unlikely to just fall off
• The geometry + friction + compression all help keep it together

But:

• once you open it, there’s nothing preventing the lever from sliding off the end
• over years, that’s when parts go missing

---

One more nuance (you’ll like this)

Cam quick-releases are designed so that:

• the cam passes slightly over-centre
• which means the reaction force actually helps keep it closed, not open

So even that big force you calculated is stabilising, not ejecting anything.

---

Bottom line

• Your force estimate → solid
• Your conclusion → slightly misapplied direction-wise
• The circlip → only a retainer for the loose state, not a load-bearing component

---

If yours has survived 10 years without the clip, that’s actually kind of impressive 😄
…but it also means you’ve probably just been lucky not to lose the lever while the wheel was off.

Want to go one level deeper and sketch the force directions? It’s a nice little statics problem.

Eugene: 

Now possibly radial shock forces could loosen it?

 

ChatGPT: 

Yeah—that’s the right line of thinking now 👍

Radial shocks (bumps, vibration, wheel chatter) can definitely contribute to the lever working loose—but not in the way your earlier force model was pointing.

---

What radial shocks actually do

They don’t overcome that big axial clamp force directly. Instead they:

• create small oscillating forces and micro-movements in the mechanism
• introduce slight lateral and axial play when the lever isn’t perfectly loaded
• cause fretting + wear + tiny shifts over time

Think of it less like a single big force and more like:

thousands of tiny nudges gradually walking the part outward

---

When this matters most

Radial shock becomes relevant when:

• the lever is open (no clamp force → nothing “locking” it)
• the cam surfaces are worn or dry
• there’s clearance in the fit between parts
• corrosion has already weakened things

In that state, vibration can:

• rotate the nut slightly
• shift the cam position
• and crucially… let the lever creep along the shaft

---

Why the clip still matters

Even with shocks:

• there’s still no strong continuous axial force pushing the lever off
• but shocks can give it just enough intermittent movement to eventually:
  • reach the end of the shaft
  • and slip off if there’s no circlip to stop it

---

Analogy (closest real-world feel)

It’s like a bolt that doesn’t unscrew from one big force—but from:

• vibration
• tiny cyclic loads
• gradual loss of friction

Same idea here—just linear instead of rotational.

---

So your revised model is basically spot-on

• Big clamp force → not the issue
• Shock/vibration → slow migration mechanism
• Missing clip → nothing to stop the final exit

---

If you’re half tempted to keep running it as-is… I’d still swap the skewer. It’s one of those parts where failure is very “all at once,” and usually at the worst possible moment.

what is the piece called that the cam lever fits into?

That part is usually called the cam body (or sometimes the cam housing).

On a typical quick-release skewer you’ve got:

• the skewer rod (the long shaft)
• the cam lever (the flip handle)
• the cam body / housing ← the bit you’re asking about
• the adjusting nut on the other side

---

What it actually does

The cam lever pivots against this cam body, which has:

• a curved cam surface
• a seat for the lever to press into

When you close the lever, it rolls/wedges against that surface and generates the clamping force.

---

Other names you might see

Depending on brand or context, it can also be called:

• cam assembly (when grouped with the lever)
• lever seat (less common)
• cam mechanism

But “cam body” is the clearest, most widely understood term.

---

If you’re trying to source parts individually, that’s where it gets annoying—most manufacturers don’t sell the cam body separately. They expect you to replace the whole skewer.

Are you trying to rebuild it piece-by-piece, or just figuring out what went missing?

Making a New Part For My Bicycle Tail Light

© Eugene Brennan

If possible I always repair rather than replace. I've always done this before it became a trendy green thing, simply because it saves money and I've never had lots of money to spare. I remember my father fixing a hole in the silencer on the car, with a section of tin plate, cut from a Cow & Gate baby formula tin, and fixing it to the silencer with coat hanger wire. So I didn't lick it up off the ground, as the saying goes. I've got slagged off by Millennials and successive demographic cohorts, for doing such things, and asked why I don't "just but a new one". Anyway, today I was fixing the tail light for my bike.

Printer Ink Experiment

Printout using Quink fountain pen ink. © Eugene Brennan

I'm fed up buying inkjet cartridges for my Lexmark printer, which usually dry up and stop working. I was speaking to a technician about this and he said that the advice from manufacturers is that inkjet printers really need to be "exercised" every week or two by printing a couple of pages to prevent this happening. I predicted this in the 90s when I bought my first inkjet printer, having owned a fountain pen and knowing that the nib occasionally dried up. The print nozzles in inkjet printer heads are only a few tens of microns in diameter, so it's not unexpected.

57 Modes and Nothing Wrong

Bicycle rear light. © Eugene Brennan

Mystery light.

I've bashed it almost to the point of cracking the lens, put it in the freezer for half an hour while turned on and left a strong magnet in its vicinity and it still won't turn off. Yet when I'm cycling, after 100 m or so, it manages to turn itself off. And that keeps happening. The only thing I can think of is that deceleration in a certain direction is bending a spring in a clicky button. Or maybe it's condensation on the PCB. Or possibly constant vibration rather than large shocks.

Anyway, I hate the way they don't put proper switches on these things instead of buttons for changing modes that inevitably cause them to switch off on bad roads. It's difficult to get a decent backlight anywhere. I've tried eBay, Amazon and AliExpress.

 

Leaky Schrader Valve

Hand truck tube. © Eugene Brennan

So being a member of 60 Facebook groups and squandering a lot of time nattering about all sorts of things, I try to get at least one useful thing done per day. Today was damp, so dealing with weeds in the driveway was out of the question. One of the items on my ever-growing-longer "to do" list I hadn't got around to was fixing the wheel on my hand truck which always seemed to be soft when I went to use the truck. I thought it might be a puncture caused by running over thorns (what usually happens with the bike). However doing a water test showed the Schrader valve was leaky. There's a special tool for taking the innards out of the valve casing, but I usually just use a needle-nose pliers. These valves have a conical part that sits into a conical seat, the cone being pressed against the seat by a spring. I blew out the valve with compressed air, gave the red ring that seals against the valve body a wipe and put the bits back together. This time there were no leaks whatsoever after the water test.

Removing the Valve innards. © Eugene Brennan

C

Check valve. © Eugene Brennan

© Eugene Brennan

Nice New Wheels Delivered!

Sealed bearing wheels. © Eugene Brennan

It took me forever last week to source these. The engineering aspect and excuse to post about them is that they have sealed ball bearings, so somewhat better than the cup and cone bearings used in traditional mountain bike wheels. The latter are a pain in the arse to maintain and water gets into the ones on the back wheel when hosing the sprockets after they get covered with muck. So they have to be dismantled, cleaned and degreased (see photo below). Several local bike shops hadn't heard of these wheels or said they didn't exist. Anyway 360cycles.ie will be getting a good review because they were really helpful and gave lots of updates on the progress of sourcing them for me.

Sealed bearing image public domain, courtesy 4volvos.

Components of a rear cup and cone bearing. © Eugene Brennan

Maps for Council Workers

Image © Google Maps

Apparently when we report road or other issues on the KCC portal and specify the location of the problem, repair crews just get a description, but not the map pinpointing where they have to fix them (there's an option to tag the position on a map when making a report on the portal). I had a long discussion with a repair crew a month or so ago and got a great insight into how they feel about this and other things. If someone reporting a problem doesn't clearly specify road numbers and distances from landmarks, obviously it can be difficult for a repair crew to find the problem. Saying that it's "just down the road from Johnny's house" or similar isn't very helpful. A point on a map would be more logical (assuming the person who made the report zoomed in sufficiently to accurately place a tag). Reading a map is hardly rocket science and I'm sure KCC can train workers to do so.

 

Maybe someone can let me know whether the tarmac is broken up and gone from around this drain opposite LaTouche Cottage in Brannockstown so I can report it? I make mental notes of these things when I'm out and about on the bike, but I've got a head like a sieve and forget where I've seen the defects. I don't want to make a false report. A cyclist could easily lose control if they hit this. It was ok last August when the Street View imagery was taken, but as far as I recall, the tarmac fill around the drain has now disintegrated.

Attacking the Brasses With Acid

Shiny fittings on the newly-painted front door, the result of hours of polishing. © Eugene Brennan

Painting is something I tend to procrastinate on and life is too short for cleaning brasses regularly. I do a good job of cleaning the hardware on the front door every two or three years when I paint it, but then I only polish the brasses a couple of times before I lose interest and have more pressing chores to attend to. After three years of non-cleaning, a thick layer of oxide builds up on the fittings, which takes several hours to remove with wire wool and Brasso. (Brass is an alloy of copper and zinc, so the tarnish is zinc oxide and copper oxide).

Gear Grinding

Lawn mower gearbox. © Eugene Brennan

Using diamond burrs to square the teeth of the freewheeling gears from the lawn mower gearbox. They were rounded with wear and slipping. A bit like being a dentist (They use a similar tool, but it's pneumatically or hydraulically driven rather than battery powered. My dentist showed me all the bits they use) Almost finished now. Crude, but worth a try to see if it'll work. The alternative is €55 for a new one. Still have to do a MacGyver job on the driveshaft bearing which has worn and come apart. I've got lots of ball bearings, but none the right size.

Two Repair Jobs Today: Replacement Wheel Bearings and New Bushing for the Secateurs

Wheel bearing replacement and new bushing for secateurs. © Eugene Brennan

I still have to put the mower gearbox back together after doing some dental work on the gears' teeth. Not sure whether it'll work, but it was worth a try. I got replacement bearings for the wheels yesterday from Amazon. One of the bearings on the back axle had corroded after five years from turning the mower on its side to hose it and remove caked-on grass. Water getting into the bearing rusted it away totally. Anyway I'm going to have to make some sort of protective shroud that'll slip onto the axle and act like an umbrella when hosing. The old bearing, or rather what was left of it, was really difficult to remove from its housing or "sheath" as it's called in the parts description. Bearings are usually put into housings using a bearing press and are a tight fit with a small tolerance. But this one had also rusted into place and so was stubborn and refused to come loose. Because the back of the sheath was tapered, I couldn't tap the back of the bearing to push it out. So I had to grind a narrow groove with a diamond burr, cracked the outer race (the thing the balls run around in) at this point with a chisel and then was able to use the corner of the chisel to get enough purchase to bend and snap a chunk off the race. That reduced the force on the sheath allowing me to prise the rest of the bearing race out. The new bearing was pressed into place in the jaws of a vice, taking care to do it evenly.

 

Worn secateurs

 

The pivot hole in one arm of the secateurs was worn, so the jaws didn't move parallel to each other when closing, affecting cutting of thicker branches. Rather than throw it out, I made a bushing ring from scrap tube of a suitable diameter which reduced the hole size to what it was originally, so the pivot bolt returned to being straight.

Lawn Mower Woes, Safety Glasses and Spare Parts From Vehicles

Lawn mower float bowl. © Eugene Brennan

I usually wear safety glasses when cutting the grass bank outside the house. I don't in the garden, but outside there can frequently be accumulations of stuff since the last cut: rocks and stones and sometimes parts that come off vehicles. I've hit the handle of a bucket on one occasion, which flew out from under the front of the mower and became airborne. Today I found this pin. Not sure what it is, but it seems to have broken off a longer piece and has a wear mark as though something was rubbing it. It could easily have struck someone walking along the footpath, either from me hitting it with the blade or when it came off a vehicle. Anyway, I do a recce before cutting to check for such items and also dog poo. Hitting something hard like this usually puts a big dent in the blade. 

 

Mower Repair Frustration

 

I starting cutting grass this morning. Then I got delayed on a long phone call. I was going to paint the cappings of my gate pier this afternoon, but ended up attempting to resume grass cutting instead. But of course the mower wouldn't start. I predicted a fuel flow problem, took the float bowl off (It functions pretty much the same as a toilet cistern) and just as expected, dirty fuel with some black grime particles ran out of the bowl. This stuff comes from the filling station, because my Jerry can is sealed and dust doesn't go into it. I guess the tanks in filling stations aren't exactly pristine and corrode over time or coatings on the inside wear off. Or maybe the grime is in the fuel on delivery. Anyway, fixing something outdoors isn't a good idea, because parts usually get lost in grass or gravel, but I decided to do it in situ, rather than dragging the mower back to the workshop. I thought I had lost the sealing washer from the bowl on the driveway because I had heard something falling. I wasn't sure what I was looking for and I couldn't remember what the washer looked like, so I had to put a post on a lawn mower repair forum to enquire. It turned out it wasn't lost, but stuck to the underside of the bowl. After all that, I put everything back together and the mower started on first pull. Then I started it again outside the wall and it decided to backfire and snapped the starter rope. So that'll have to be fixed tomorrow. These farcical scenarios are regular occurrences, or a variation on the theme is "There's a hole in the bucket dear Liza" versions, where 5 minute jobs turn into day-long projects.

Float bowl. © Eugene Brennan

Carburetor jet. © Eugene Brennan

Float bowl float. © Eugene Brennan

Broken starter rope. © Eugene Brennan

 

Lucky I didn't hit this with the blade! © Eugene Brennan

More Boiler Woes

Cleaning the filter on the oil tank. © Eugene Brennan

I let the oil run out a couple of weeks ago while waiting for price to drop. Got a fill of oil and thought I'd just have to bleed the boiler, but no air and oil escaped from the bleed vent when unscrewed. Checked the filter at the tank and it's not blocked. So fine sludge has probably got through the screen mesh of the filter and perhaps blocked the pre-filter in the boiler. So hopefully the oil line won't have to be dug up, because it's under concrete, and the boiler service people can just clear the line with compressed air or flush it. Anyway it's more expense. Moral of the story, check your oil regularly. I always do, but just let it down a bit far while waiting for price to drop.

Bleeding air on the boiler. © Eugene Brennan

Mouse Surgery

Circuit board from a computer mouse. © Eugene Brennan.

I've been struggling with the left button on my cordless optical mouse for the last six weeks or so. The momentary microswitch for the left button has worn out (they're good for something like 100,000 operations) and a single click is frequently doing a double click, which is really frustrating (known as "switch bounce"). I replaced the switch today with one harvested from a mouse I bought from Dealextreme and which only lasted a week before breaking down. I have a pack of five microswitches on order from AliExpress, but the double-click symptoms were getting worse and I couldn't wait any longer to fix the problem. 

 

Eugene Brennan

To get an idea of the size of the switches, they're 6 mm square as shown in the photo below.

 

© Eugene Brennan

This is the second time I've replaced the switch on this particular mouse due to it wearing out from use. It's a Microsoft Wireless Notebook Optical Mouse 4000 that's very ergonomic and a "Goldilocks" product: not too big or small, comfortable in the palm and the scroll wheel has a really smooth movement. The product is no longer manufactured and the replacement Microsoft mouse hasn't had good reviews. Also computer mice can't normally be "test driven" in stores because they're usually blister-packed, hence my decision to keep fixing this one.

 

Related reading: My guide to cleaning the scroll wheel on a mouse.