# McElroy Bug Key Project

Like I don't have enough projects of my own, sometimes I inherit
the projects of others.

This was the case with my ![1936 McElroy telegraphy key][1].  It
was gifted to me by its former keeper who, I think, wanted to see
it revived and thought I had the interest and skills to do it.

Not wanting to look a gift horse in the mouth, I didn't really look
at what I'd been given until I got home later that day.

It's old, well used and it's been mostly stripped down to its
![component parts][2].

After a general inspection most of it seems to be there but, some
of it is in desperate need of replacement parts or repair.

[1]: https://morsemad.com/mc_files/mc063.jpg
[2]: gopher://gopher.icu/I/images/McElroy/P1010341-sml.JPG


## Things needing attention

The spring steel which joins the control linkage to the pendulum
has ![broken][3] and the broken part has been lost so, I don't know
it's original length. There is however still part of the steel
riveted inside one of the parts that I can check with a micrometer
for thickness, once I've removed the rivets...

The rivets appear to be aluminium and so should only require some
careful drilling of the rolled edge before I can drift them out.


The rubber feet have turned to a stone like consistency. Unfortunately
whoever removed them sheared one of the screws almost flush in the
![cast iron base][4]. This will probably be the most difficult thing
to rectify without doing any damage.

Suitably sized replacement feet are already on order. They are not
an exact match for the originals but will be close enough for me.


The various pivot points appear to be very worn and in some cases
the wall of hollowed screws have broken away. I'm not quite sure
yet how I will fix these, possibly by simply grinding down the
height; if there is sufficient material to allow it.

[3]: gopher://gopher.icu/I/images/McElroy/P1010340-sml.JPG
[4]: gopher://gopher.icu/I/images/McElroy/P1010342-sml.JPG


### Update 28/07/2026

Today I set to removing the rivets from the pendulum assembly.

I held the components with some fibre jaw covers in the vice, to
save the vice teeth marking them. Then I selected a drill which was
just a little bigger than the hollow inside the back of the rivet
and carefully ![drilled][5] a couple of mm in with a Dremel.

Once I was happy that I was close to the outer edge of the rivet, I
selected a punch that would go just inside the drilled hole and
proceeded to give it a few light blows with a small hammer to see
if they would budge. 

Looking at the underside I could see they were moving so, I proceeded
to drift them out. I did the same for ![both parts][6].

*Note:* Initially I had thought the rivets were aluminium, due to
their dull silver colour. However, upon starting to drill them, it
became obvious that they were in fact coated brass.

[5]: gopher://gopher.icu/I/images/McElroy/P1010350-sml.JPG
[6]: gopher://gopher.icu/I/images/McElroy/P1010351-sml.JPG


### Update 05/08/2026

The ![new feet][7] arrived and steadily I have day by day been
cleaning up components until all are clean and, where possible,
reassembled into their ![sub assemblies][8].

Plan A for removing the stuck piece of thread from the base failed.
There is not enough protruding to grab hold of it with some mole
grips. Plan B entails welding a new head (nut) onto the remains of
the protruding thread. For this I need my welder to be functioning
100% as I figure, I have just one shot at it.

My MIG welder has had a wire feed / jamming issue since I had it
and I never bothered to get to the bottom of it because it was good
enough. Unfortunately now, good enough is no-longer good enough as
I need to get a good weld straight off the bat.

Heat is usually a good way to remove stuck studs and nuts, but you
still need something to grip onto. My hope is that the heat from the
weld will do enough to loosen the stud and the welded on nut will
give me the purchase I need to unscrew it.

If I fail then, I will have to resort to drilling out the thread and
attempting to re-tap the hole. From bitter experience, this is not
a road I wish to travel...

[7]: gopher://gopher.icu/I/images/McElroy/P1010357-sml.JPG
[8]: gopher://gopher.icu/I/images/McElroy/P1010362-sml.JPG


### Update 08/08/2026

Plan B was a success at the second attempt. 
I gave it my best shot:
* Soaked the thread for several days with good quality penetrating
oil.
* Struck the head of the thread a few times with a light hammer to
try to jar it.
* Shielded the base around the screw with a piece of copper by
drilling a 5/32 hole in it, which was precisely the size of the
thread.
* Tried to get a good deep weld through the nut and down onto the
head of the thread below.

The first time I tried welding through the nut directly but, the
weld wasn't deep enough and it fairly easily broke off. I needed
to get the stud hot so at the second attempt, I put a hot tack of
weld directly onto the head of the thread. Then I placed a fresh
nut over it and proceeded to weld through the nut with plenty of
power.

I put a small adjustable spanner on the nut and slowly started to
turn.  This is the worst part as you don't know if your weld is
shearing or if the thread is actually starting to turn... After a
couple of rotations it was clear that I had it moving and I could
relax.

The ![extraction device][9] looks ugly, but it worked like a charm.

[9]: gopher://gopher.icu/I/images/McElroy/P1010364-sml.JPG


## Reassembly

Now that the stuck thread issue is out of the way and the base has
been cleaned. I have started the process of ![reassembly][10] and
ordered some of the replacement fixings.

This should leave me with just the pendulum assembly issues:
* Dah pivot bar being excessively worn.
* Replacement 0.017" spring steel strip.
* Replacement truss head, nickel plated 1/16 brass rivets x4.
* Rivet clincher, to close the above.
* Drilling the centres of the replacement grub screws.

[10]: gopher://gopher.icu/I/images/McElroy/P1010366-sml.JPG


### Update 11/08/2026

With my attention now fully on the pendulum mechanism, I started
to make enquiries to identify the length of the broken spring
steel. 

I contacted a fellow amateur radio operator who was listed in the
above pdf document as having a McElroy key of the same model. He
very kindly took the pendulum from his own key and took several
measurements for me along with some high resolution photographs.

The dimensions he provided confirmed my height (3/8") and thickness
(0.016" - 0.017") measurements and also gave me the missing
![length][11] measurement (12mm between parts / 29mm total) which
I did not have.

After reading a few articles on how best to cut spring steel, which
has a high carbon content and is quite hard, there was a listing
of items commonly made from spring steel. This listing included
bulldog clips...

My brain started it's analytical process "I have bulldog clips, I
wonder what their thickness is?". So, I went hunting the house for
bulldog clips with micrometer in hand. The first one I found was
0.015" just a hair too thin, but promising. The second one was just
a hair thicker at 0.016", that was good enough for me.

Armed with the information that cutting was probably best done with
a cutting wheel on a dremel or hand grinder. I got the dremel, a
small vice and a face shield from the garage.

Fifteen minutes later I had a piece of ![spring steel][12] which
was over size but, right for the job once filed to the correct
dimensions. An hour later, after gently whittling away at the
dimensions with a file, I was ![within 0.5mm][13].

[11]: gopher://gopher.icu/I/images/McElroy/P1010376-sml.JPG
[12]: gopher://gopher.icu/I/images/McElroy/pendulum-assy-sml.jpg
[13]: gopher://gopher.icu/I/images/McElroy/P1010381-sml.JPG


### Update 14/08/2026

Often, taking the time to think something through before getting
the tools out is wise.

After trying in vain to dress what was left of the pivot that
actuates the 'dah', it basically crumbled away when I touched it
with a file, I knew I would need to replace the pivot bar.

I took a day or so to think about it and determined it was probably
pressed in, so it would be best to press it out. Rather than take to
it with a hammer and punch, possibly causing damage to the brass
by attempting to hold it.

The plan came to me. I had a piece of square aluminum in the garage
that has been used as a drift, it's about 30mm square by 120mm long.
I would drill a hole fractionally wider, and deeper, than the bar
I was trying to remove into the aluminum and use this as the base
of the press.  The aluminum, being less hard than the brass, should
avoid any marking but offer enough support to push against to remove
the steel pivot bar.

There was about 3mm of bar protruding on the top side so I used a
4" bench vice and opened it up to take the aluminum and the component
and began to tighten the jaws of the vice to act as a press against
the protruding pointy head of the pivot bar. 

As I started to bare down on it, I felt it ease. It had begun to
move, so I continued on until the vice jaw was within about 1mm of
the brass surface.

I didn't want to get any closer than that so I backed off the vice,
took the aluminum block placed it on the floor with the hole
facing up with the component sat on the top. At that, I went looking
for a suitably sized punch and a small hammer.

A couple of light blows with a small hammer was all it took to drive
it the rest of the way ![out][14].

Having put a micrometer on there it appears to be 1/8" steel bar
which should be reasonably easy to source. Not sure how I might
put the knurled ridges in it though...

[14]: gopher://gopher.icu/I/images/McElroy/P1010394-sml.JPG


### Update 15/08/2026

The knurling thing had been playing on my mind. How could I do it
without using machine tooling? Then, [I had an idea][15].

[15]: gopher://gopher.icu/0/phlog/Radio/Ghetto-knurling.md


### Update 20/08/2026

The fixings I ordered (2x grub screws, 6x stainless machine screws
and 8x machine nuts) arrived ahead of schedule and ![fit perfectly][16].

I've fitted two of the three feet using two of the six machine screws.
The hole of the third foot, the one which had the thread stuck in, has
something stuck in the thread part way. This is probably what lead to 
the original fixing getting stuck and breaking off. Not wanting a repeat
performance at some point, for either myself or whoever should own it
in the future, I have ordered an 8-32 UNC plug tap.

**Note:** Plug taps are usually the final tap in a set of 3 and
they will go down close to the bottom of a blind hole.

At the moment I am still undecided on the grub screws. I have tried
to keep the restoration looking original by using slot head fixings
but, the grub screws have hexagon heads. What I am considering doing
is using them upside down and using a dremel to carefully cut a
screwdriver slot into the other end so, they look closer to the
originals...

With every step forward I am more motivated to press on and so have now
ordered: 
* 2.2mm carbide drill bit (for drilling rivet holes in the spring steel)
* 2mm rivet clincher (to secure the new semi-hollow rivets)
* 1/8" silver steel rod (for the now 'dit' pivot)

The rivets I am being sent are a little over size and in plain brass
finish.  Plain brass is better as I will need to make some adjustments,
heads will need to be reduced in height and the shanks reduced in
thickness by a hair.

Once the rivets are of the correct dimensions then, I will look at
ordering some materials for nickel electroplating. Yes, I'm serious and
no, it's not that difficult.

[16]: gopher://gopher.icu/I/images/McElroy/P1010399-sml.JPG


### Update 21/08/2026

The tap arrived faster than anticipated. Unfortunately, despite
being advertised as a 'bottoming tap', it had a large point on the
end. To cut a long story short, no pun intended, I ground down the
point using an angle grinder.

I put some vaseline on the tap before I used it so that the swarf
/ debris would stick to it and come out along with the tap.

With the hole now cleared I fitted the remaining foot.

While I was touching up some damaged paint with a permanent marker. I
noticed it was a little glossy, unlike the matt crackle finish.

This gave me an idea as the original name/serial tag had a similar
finish which had all come off with time. I decided to try a bit of
free hand to see how it looked on the name/serial tag. It was fairly
convincing at a squint so, I took it upstairs under my lighted
magnifying glass and continued for half hour or so. I ran over into
the letters in a couple of places but, it was fairly easily rectified
by scraping it back out with a needle.

When I looked at it without the magnifying glass it looked
![fairly original][17] and really improved the ![look of the key][18].

[17]: gopher://gopher.icu/I/images/McElroy/P1010402-sml.JPG
[18]: gopher://gopher.icu/I/images/McElroy/P1010401-sml.JPG


### Update 27/08/2026

The ![1/8 silver steel rod][19] arrived this morning. 100mm so,
there should be enough for a second or third try should things go
pear-shaped.

I've been thinking how best to tackle this as, it needs to be done in
stages due to the knurling. 

The first operation needs to be the knurling. If I grind the point
before the knurl then, I have no flat surface to support the bar
to do the knurling as, it's very close to the end of the bar. I
also need a length of bar to turn to make the knurl, which I won't
have if I cut it to length. Total length of the completed part is
less than 15mm...

After the knurling, I will cut the bar to length and attempt to put
it into a drill chuck, to spin it so that I can use a file or stone
to achieve the points on each end.

[19]: gopher://gopher.icu/I/images/McElroy/P1010409-sml.JPG


### Update 28/08/2026

The silver steel appears to be a lot harder than the nail I knurled.
I've broken some of the teeth off one of my files trying to do the
knurling and had three goes at it. 

I'm not happy with the result but I've cut off a length with the
best knurl. 

I'm putting the ![points on the bar][20] by clamping the work piece
into the chuck of my eletric drill, which has a trigger locking
button, and applying a flat file to the end of the spinning rod to
shape it.

Some time later, after several sessions of turning down the pointy
ends in the drill, I had the ![new pivot in place][21].

The brass rivets also arrived in the post and appear to be a good
match for the originals, all be it in plain brass.

[20]: gopher://gopher.icu/I/images/McElroy/P1010413-sml.JPG
[21]: gopher://gopher.icu/I/images/McElroy/P1010414-sml.JPG


### Update 29/08/2026

The rivet clincher arrived in the post this morning however, I'm
still waiting on the 2.2mm cobalt drill which should have arrived
on Tuesday...

I've ![re-fitted the bar with the new pivot][22]. This appears to
be working as it should although, I am using the hexagon recesses
in the grub screw heads as the pivot base. Essentially, they are
fitted inverted.

I'm still unsure whether to attempt to cut slots in what should be
the base, and use the hexagon head as the pivot, or to drill the
correct end and leave the hexagon heads. Another alternative is to
drill the base and slot the head through the hexagon...

The 'dit' pivot at least now works and you can send morse on it
like a straight key, if you turn the key on its left side.

Jobs left to do:
* Drill the spring steel strip for the pendulum.
* Nickel coat the plain brass rivets.
* Assemble the pendulum and rivet in the spring.
* Make a decision on the grub screws (slot, drill or both...).
* Re-install pendulum and make final adjustments.
* Make a cable to connect it to a rig.

[22]: gopher://gopher.icu/I/images/McElroy/P1010417-sml.JPG