Re: [CAD_CAM_EDM_DRO] A bit off topic - resistance welding

Ian Wright wrote:

> Does anyone have any idea whether it would be possible to make a
> simple
> resistance welding set-up for tiny jobs? Basically what I am wanting
> is a
> way of spot welding bits of watch or small clock mainspring material
> together - spring steel up to about 15 thou thick with welds probably
> about
> 1.64" diameter.

Are you serious about having the weld spot be 1.64" diameter? Or, do
you mean several
weld spots across an area of 1.64"? The full 1.64" welds are bigger
than what they use
to hold cars together with! Those welds are 3/8 to 1/2" diameter on
thick frame members.

I think there may be some ratio of how large a weld spot is practical
to the metal thickness. One of the problems is how to make a perfect
contact with the entire area. If the contact is imperfect (guaranteed),

then the current will flow through the good contact points, and the
concentrated current will likely burn holes in the metal. Thicker metal

allows the current to 'expand' from the contact point into the bulk of
the material. For .015 maximum, there won't be much spreading.
My experience with this is the connections between Ni-Cd cells, which
is usually done with stainless tape about .010" thick. It is always
done
with several small spots. I know from experimenting with this stuff
that
it is VERY easy to burn holes through that material.

Could you possibly mean 1/64 of an inch? That would make much more
sense.

> I was thinking it might be possible using just a big power
> supply and capacitor like in EDM and discharge this through some kind
> of
> 'clip' but I'm not sure how to control the discharge (SCR, Triac?)
> Also I'm
> not sure about the voltage/current needed - a 24 volt, 10 Amp
> transformer
> with a bridge rectifier and 10,000uf capacitor seems to give an
> impressive
> spark but would this be enough to weld my bits of steel or is it only
> good
> at screwdriver blades? ;o)

A small transformer or power supply is fine. A variable regulated one
allows
setting the voltage to accomodate the material. Several smaller
capacitors
with switches to connect the desired combination would give you the best

controllability. 10,000 to 50,000 uF sounds like a good range, from my
experiments. I used heavy copper wire, #10 or #12 solid copper, for the

electrodes. I just jammed them down at the desired location, which
didn't
work real well. A mechanical holder-force unit to align the two
electrodes
parallel to each other and apply a known force would help a great deal.
A large SCR could be used to dump the selected capacitors into the
work when everything is ready. I'll try some ascii art :


_________________________________________\
| | + | /
___|__ | o |_ PB SW
| PS | ___ o | weld
------ M ___ __| electrodes
| | | ___
| | \ ___ T
| | / |
| | R\ |
| | / \ | /|
_______________|____\|/ |__________________\
|\ | /
| \|

Here, M is the main energy storage cap, R is a 100 Ohm
resistor, and T is a capacitor of about 1 - 5 uF to deliver
a trigger pulse to the SCR when the pushbutton switch is
pressed. A resistor can be put between the power supply and
the main cap to prevent overloading of the power supply
every time it charges the main cap. If you go much above
18 Volts, it might be a good idea to put a 22 to 100 Ohm
resistor in series with the trigger cap, T.

I rebuilt something like this that was used to 'flash
volatilize' a sample in a tiny tungsten 'boat'. The boat
would flash red hot for about 20 mS and then cool down
so fast you weren't even sure you saw the red. It ate
knife switches every two weeks. I finally put an SCR
in it with the circuit above, and it worked until its'
retirement.

Jon