Michael (Micksa) Slade said:
Does this mean the tips you buy come with temp sensors in them, at least
for the 941?
YES.
One thing where opinion seems to be divided is on the subject of thermal
response (I presume that's how quickly the tip temp recovers after losing
heat to a joint?). I've read that having an iron with a smaller tip that
loses temp is better for SMT because its low heat capacity is less likely
to wreck tiny components from uneven heating or whatever. Is there any
truth to this?
Mick.
There is lots of subjective argument about soldering irons, and everybody
has their favourite iron. Some people get quite passionate about it!
Ultimately though thermal engineering analysis will tell you what works
best. I will ignore smt devices in particular at first.
Your basic soldering iron consists of a heating element followed by a
thermal conductor (the tip). The heater dissipates some power P (the iron
power rating). This causes the tip to heat up. Heat then flows from the tip
to the surroundings, through the surface of the tip. Heat loss is
proportional to temperature, so as the tip gets hotter, it loses more heat
to the surroundings. Eventually a steady-state temperature is reached,
whereby the heat lost to the surroundings is equal to the power dissipated
in the heater. (the actual heat flow calculations are pretty nasty but
do-able).
When the iron is used to solder a device, the tip is applied to the device,
solder is applied etc. When this occurs, because the device and the solder
are at fairly low temperatures (compared to the iron - around 300 deg C) and
have mass (lumps of metal really) heat flows from the tip of the iron
(wherever its touching) into the device and solder. This heatflow through
the tip will cause a temperature drop across the tip. Long, super-skinny
tips are TERRIBLE, I'm sure most people have experienced them sticking to
moderate-sized components/pads etc.
cheap crap soldering irons have no temperature control at all, and instead
rely on the temperature drop across the long tips (40mm or more). Slightly
less crappy soldering irons have a "temperature control" thats usually like
a light dimmer in the input, ie allows you to set the steady-state
temperature by varying heater voltage and thus power P.
Try an experiment: get your soldering iron, and try to solder a 1/4W
resistor leg onto a 4"x4" square of copper-clad pcb. Try on the edges
first - in the middle of one edge, then at a corner. Then try in the centre
of the pcb - where is where you are likely to find your iron tip sticking to
the copper, and not melting the solder. The problem here is the large
temperature drop across the tip, as it tries to heat up the big copper plane
(my metcal will solder a resistor to 1mm Cu sheeting using a tip I also use
for smt work
.
A monstrously fat tip is a much better conductor of heat (less T drop from
element) and also stores a sizeable amount of heat in its mass, further
adding to the ability to supply heat quickly. Problem is, that sort of tip
can do a lot of damage (ie heat stuff up real fast)
In a temperature-controlled Iron, the temperature sensor notices the drop in
temperature, and winds up the heater power P, to try and keep the tip
temperature constant. If the temperature sensor is a long way from the tip
(my Ersa soldering station tips are about 25mm long, and 2-3mm in diameter,
compared to the 5mm long, 5mm diameter conical tip on the Metcal. The Ersa
wont solder onto 1mm Cu sheet without a tip thats almost 1cm in diameter -
its 3mm chisel tip sticks to the Cu.
Some types of temperature controlled soldering irons (eg weller, metcal) use
fancy materials/properties (eg curie temperature) to precisely regulate
temperature. These can be really expensive - tips for my Metcal are
$hundreds.....
In addition to having a thermal "resistance" (relating temperature drop to
heat flow and conductor dimensions, analagous to current flow), metals also
have thermal "capacitance" - they will absorb heat at a certain rate,
causing their temperature to rise (specific heat capacity, J/kg/K)
So think about applying a heat source to one end of a rod of metal, which is
at a uniform, lower temperature. Initially the "hot" end of the rod slowly
heats up. as it does, the next bit of the road slowly heats up.....each bit
taking time. This is of course what you observe when you stick a poker in a
fire - heat travels up the poker slowly, and depending on the length of the
poker, might not make it to the end - heat is also being lost through the
surfaces.
the themal model of a tip is therefore a current source (power loss) in
parallel with a thermal capacitance, followed by a series thermal resistor.
When tips get beefier, R goes down but C goes up, so although the
temperature drop is lessened, the slow RC response time can somewhat (or
even completely) negate the advantage of bigger size.
For fast thermal response, you therefore want very little thermal
capacitance (not much material) and very low thermal resistance (low
length/area of ratio of tip, ie short fat tip)
Soldering SMT components is a different matter altogether. Any component
sees this when soldered: Pad temperature rises during soldering. Heat flows
up into lead, and along lead (usually small leads, so temperature drop is
high) before entering device, and heating the device up. As everything heats
up, it expands, due the the Coefficient of Thermal Expansion (CTE) of the
various materials involved. Unfortunately all devices are made from a
variety of different materials, with differing CTE's so some bits expand
more than others.
Leaded devices can tolerate quite large changes in dimension, as the leads
can flex. many smt devices (R, especially C etc) have metallised endcaps on
a ceramic substrate - no leads. Therefore no temperature drop across leads.
Therefore smt device endcaps get stinking hot, ie maximum expansion. But
there are no leads to flex to accommodate the dimensional changes, so the
endcaps crack - perhaps immediate failure, perhaps later on.
This "thermal shock" problem of course occurs in leaded parts too, just to a
lesser extent. A wave-soldering machine has a preheater specifically to
reduce thermal shock. smt soldering machines are precision dynamic
temperature controllers, and have a pre-programmed temp profile that slowly
ramps the temp up to say 100C or so, lets it sit for a while, then cranks up
to 200C or so, and back down - this is in most smt component manufacturers
data, say
www.avx.com
High-voltage smt caps in particular are susceptible to this - micro-cracks
develop due to overheating, which (IIRC) the electrostatic forces inside the
cap can make propagate, leading to eventual failure - I have personally
witnessed this with a flyback smps. Looking at the charred remains, we found
a 15nF 1000V X7R cap had burst, causing the problem. Two of these were in
series across a 600V bus (which could go up to 950V), so the caps should
NEVER have failed. The prototype was hand-assembled with a soldering iron.
Ideally, use an smt placement machine & reflow soldering machine. This is a
bit too rich for most hobbyists! Hot-air rework stations are the next best
thing - companies which re-work smt R's, C's with soldering irons are just
asking for field failures - and are not too expensive. If you have to do it
by hand, use a good temprature controlled iron, and keep the temperature
low. IC's (eg SOIC) are usually ok, as are any other leaded smt part. Its
just the metallised pads you have to watch. A heat gun (hair dryer) can be
useful as a pre-heater, too.
Or, be prepared to check bits after you solder them.....
cheers
Terry
