Skip to main content
Home
  • Home
  • Browse All Issues
  • Model Aviation.com

The Engine Shop 2003/11

Author: Joe Wagner


Edition: Model Aviation - 2003/11
Page Numbers: 59,60,61

November 2003 59
CORRECTION: In the July column I
wrote that Brodak’s new .40 Control Line
(CL) engine displaced .426 cubic inch. I
was wrong about that. My faulty
information came from a data sheet packed
with the engine. It gave the displacement as
“7 cc”—which equals .426 cubic inch—and
gave the bore and stroke in metric
measurements too.
As Tom Hampshire (Belvidere, New
Jersey) quickly let me know, none of those
published dimensions was correct. Tom ran
the prototype test program for the Brodak
.40. Its actual bore (.831 inch) and stroke
(.722 inch) make the engine’s displacement
.392 cubic inch.
This apparently minor discrepancy
matters to fliers because AMA CL
regulations require a larger line size to be
used with engines larger than .40 cubic
inch. The Brodak .40 isn’t one of those!
In my most recent column I discussed
“breaking in” model engines on a test stand.
Several Radio Control (RC) fliers have
asked me why that is needed at all. “We
don’t have to break in power lawn mowers,
leaf blowers, or chain saws. What’s so
different about a model airplane engine?”
Plenty! Model engines run much faster,
and for their size they produce the most
power of any piston-type engine. That
means high pressures, temperatures, and
stresses. Also, RC engines must run
dependably through a wide range of speeds.
To the eye, the moving parts of a new
model engine appear quite smooth—even
polished—but a microscope will reveal tiny
serrations all over the surfaces. These are
tool marks: an inevitable result of the
machining or grinding processes used in
producing the part.
One reason for “running in” a model
engine is to allow the peaks of these
Joe Wagner
T h e E n g i n e S h o p
212 S. Pine Ave., Ozark AL 36360
Still being made after more than 50 years, the Fox .35’s steel cylinder sleeve needs rich
break-in running. (The NV assembly shown is a custom accessory.)
The O.S. .60 FP features ABC construction, needing a lean, hot break-in to expand its
brass sleeve and let the piston run freely at the top of its stroke.
A horrid example of what glow fuel can do to a brass clunk. This
took roughly three years—but it can begin in weeks.
Norvel .25 R/C, Brodak .40 use aluminum pistons in hardanodized
aluminum cylinders. They need long, careful break-in.
60 MODEL AVIATION
microscopic tool marks to slowly wear
away as the parts move in contact with each
other—or to “burnish” the tiny metal crests
down into the valleys.
Materials used in the engine make a
difference. If the cylinder bore is soft steel
(as in Cox and most Fox engines), a
burnishing effect during break-in is
The homemade glow igniter at left, using a D-size Ni-Cd cell, proved far more reliable
and longer-lasting than the typical Hot-Shot igniter on the right.
preferable to “wear-away.” That’s because
burnishing results in a much lower
coefficient of friction between the running
surfaces. Also, few metal particles are
eroded from the surfaces. (Those can cause
additional wear because of their abrasive
effect.)
Hardened cylinder bores—chromeplated,
hard-anodized, or case-hardened—
make burnishing difficult. The surface isn’t
ductile enough. Instead, the tool-mark
peaks must wear down during break-in
running. Fortunately ten-thousandths of an
inch or so of wear is usually sufficient. But
an even more important action takes place
when you break in your model engine:
dimensional stabilization.
When they are new, cylinders and
pistons can contain trapped stresses
produced by manufacturing processes.
During break-in, the heat generated by
combustion expands these parts and lets the
trapped stresses dissipate.
When the engine cools after running, its
piston and cylinder tend to return to their
original size—but not quite! The relaxed
stresses cause minute dimensional changes,
then those need additional “breaking in” to
achieve optimum running clearances.
That’s why it’s best—and fastest—to
break in a model engine via a series of
short runs, allowing a complete cooldown
after each. The repeated heating/cooling
cycles will then gradually bring the
moving parts into their ideal low-runningfriction
status.
November 2003 61
Most model engines with hardened
cylinder bores feature a “pinch” between
the piston and cylinder at the top of the
stroke to compensate for the expansion that
combustion heat produces.
In the chrome-plated brass cylinder
sleeves of Aluminum Brass Chrome (ABC)
engines, the combination of heat and high
pressure during the power stroke will
permanently slightly expand the upper part
of the sleeve. That frees up the pinch but
requires rather fast, hot operation during
break-in. Rich running is fine for breaking
in engines with soft steel cylinders—but it
won’t allow a new ABC engine to get hot
enough to do much good.
The Norvel engines, with their AAO
(Aluminum Aluminum Oxide) piston
cylinders, are another story! Chuck Morrow
(Anniston, Alabama) was surprised by the
unique break-in instructions that came with
his new Norvel .061 RC engine. They call
for removing the glow plug, coating the
cylinder/piston with castor or car-engine
oil, then letting that oil film “cure” for 24
hours with the piston at the bottom of its
stroke.
After that, the Norvel instructions want
you to put a propeller on the shaft and
rapidly turn the engine over backward at
least 100 times. (As Chuck learned, rotating
the new engine’s propeller
counterclockwise in the usual way causes
the propeller nut to loosen.) Continue this
backward rotation until the piston feels
reasonably free in the cylinder, then you
can reinstall the glow plug and can begin
running the engine.
The instructions for my similarly
constructed Norvel .15 and .25 RC engines
aren’t as elaborate as that. However, I
learned that the factory-recommended
break-in of 15 minutes wasn’t enough; it
took more than an hour of total run time
before either engine would idle dependably.
That is because of the one-piece
cylinder design. It makes for highly
effective cooling while the engine runs,
which is good, but it reduces the bore’s
expansion. Also, the stiffening effect of
the fins (acting like barrel hoops) restricts
the cylinder bore from expanding much.
That means it takes a great deal of break-in
running before the Norvels reach peak
performance. After that, they’re superb.
In a recent column I wrote about catalytic
reaction between brass and glow fuel. An
Atlanta, Georgia, reader (who asked me
not to use his name) mailed me a horribly
corroded brass clunk and wrote:
“You were right about brass clunks!
This one came out of the tank in my 3-
year-old Seniorita. I haven’t had time to
fly that as much as I want, and between
flights the airplane’s been hanging nose-up
on my garage wall. Yeah, I empty the tank
each time I fly. But I guess enough fuel
stays in there to cause this kind of mess.”
I’ve seen other examples of brass
corrosion, although none were quite that
bad. In one of my RC tanks the fuel-outlet
tube broke from internal corrosion, and
I’ve heard from other fliers who have
suffered similar problems. That’s why I
have quit using brass fuel tubing.
Rigid plastic fuel tubing for
automobiles was available in “model
sizes” awhile ago. I’ve used that and I
liked it, but I haven’t been able to find it
lately at any of the auto-parts dealers
around here.
Therefore, I checked with one of my
favorite mail-order “model hardware”
suppliers: Small Parts, Inc. (www.small
parts.com). It carries stainless-steel tubing:
1⁄8-inch-diameter, annealed 304 alloy. It’s
not cheap, at roughly $4-$5 a foot
depending on how much you buy, but it
will last forever in any kind of model fuel,
and it won’t catalyze methanol into rustinducing
acetic acid the way brass can.
(Small Parts Inc. is the place to find
almost anything in the way of model-size
hardware, such as screws, genuine music
wire, tools, and tubing. A comprehensive
catalog is free for the asking.)
Have you had trouble starting your engine
with a Hot-Shot type of glow-plug igniter?
If so, the most likely reason is that the
battery wasn’t getting the plug hot enough.
Here’s why.
This kind of glow igniter is usually
powered by a C-size Ni-Cd cell. The
nominal capacity of that runs at
approximately 1500 milliampere-hours
(mAh); that is, 1.5 ampere-hours worth of
power. Glow plugs normally need roughly
3 amperes for optimum heat. So a 1.5-
ampere-hour battery ought to light the plug
nicely for a half-hour, right? That’s plenty
of time to start an engine!
Theoretically, yes. But it’s not that
simple! First, Ni-Cd batteries are rated at a
lower current drain than 1 ampere. At a
much higher drain, their power-delivery
ability is substantially lower.
Second, Ni-Cd cell ratings are for new,
properly charged batteries. Glow igniters
seldom meet those criteria—especially
“properly charged.” It takes only 14 hours
for the typical plug-into-the-wall battery
charger to recharge a fully depleted Hot-
Shot-type cell. After that, excessive
overcharging can cause irreversible
chemical changes that reduce the Ni-Cd’s
capacity.
And most modelers (including me) tend
to forget how long they have left their Hot-
Shot igniters “on charge.” Sometimes that
can be for days.
The result of these “detrimental
factors” can be a glow igniter that is
capable of adequately heating the plug
element for only a few minutes. That
element may still get hot enough to let the
engine fire occasionally—but not quite hot
enough for a fast and dependable start. My
solution is to replace the Hot-Shot with an
industrial-type, D-size Ni-Cd cell. (At
www.megabatteries.com you can buy a
5000 mAh Ni-Cd D cell for approximately
$7.)
I soldered one end of a lamp cord to the
cell terminals and the other end to a plug
connector. After doing that, I lashed the
lamp cord tightly to the cell with a few
turns of Dacron cord, then I dipped the cell
in a container of red “toolhandle
compound.” It’s fuelproof and an excellent
electrical insulator.
I use the wall charger that came with
my Hot-Shot to recharge the D cell. That
cell’s capacity is high enough to eliminate
overcharging worries. (Mine is almost 10
years old and still works nicely.) MA

Author: Joe Wagner


Edition: Model Aviation - 2003/11
Page Numbers: 59,60,61

November 2003 59
CORRECTION: In the July column I
wrote that Brodak’s new .40 Control Line
(CL) engine displaced .426 cubic inch. I
was wrong about that. My faulty
information came from a data sheet packed
with the engine. It gave the displacement as
“7 cc”—which equals .426 cubic inch—and
gave the bore and stroke in metric
measurements too.
As Tom Hampshire (Belvidere, New
Jersey) quickly let me know, none of those
published dimensions was correct. Tom ran
the prototype test program for the Brodak
.40. Its actual bore (.831 inch) and stroke
(.722 inch) make the engine’s displacement
.392 cubic inch.
This apparently minor discrepancy
matters to fliers because AMA CL
regulations require a larger line size to be
used with engines larger than .40 cubic
inch. The Brodak .40 isn’t one of those!
In my most recent column I discussed
“breaking in” model engines on a test stand.
Several Radio Control (RC) fliers have
asked me why that is needed at all. “We
don’t have to break in power lawn mowers,
leaf blowers, or chain saws. What’s so
different about a model airplane engine?”
Plenty! Model engines run much faster,
and for their size they produce the most
power of any piston-type engine. That
means high pressures, temperatures, and
stresses. Also, RC engines must run
dependably through a wide range of speeds.
To the eye, the moving parts of a new
model engine appear quite smooth—even
polished—but a microscope will reveal tiny
serrations all over the surfaces. These are
tool marks: an inevitable result of the
machining or grinding processes used in
producing the part.
One reason for “running in” a model
engine is to allow the peaks of these
Joe Wagner
T h e E n g i n e S h o p
212 S. Pine Ave., Ozark AL 36360
Still being made after more than 50 years, the Fox .35’s steel cylinder sleeve needs rich
break-in running. (The NV assembly shown is a custom accessory.)
The O.S. .60 FP features ABC construction, needing a lean, hot break-in to expand its
brass sleeve and let the piston run freely at the top of its stroke.
A horrid example of what glow fuel can do to a brass clunk. This
took roughly three years—but it can begin in weeks.
Norvel .25 R/C, Brodak .40 use aluminum pistons in hardanodized
aluminum cylinders. They need long, careful break-in.
60 MODEL AVIATION
microscopic tool marks to slowly wear
away as the parts move in contact with each
other—or to “burnish” the tiny metal crests
down into the valleys.
Materials used in the engine make a
difference. If the cylinder bore is soft steel
(as in Cox and most Fox engines), a
burnishing effect during break-in is
The homemade glow igniter at left, using a D-size Ni-Cd cell, proved far more reliable
and longer-lasting than the typical Hot-Shot igniter on the right.
preferable to “wear-away.” That’s because
burnishing results in a much lower
coefficient of friction between the running
surfaces. Also, few metal particles are
eroded from the surfaces. (Those can cause
additional wear because of their abrasive
effect.)
Hardened cylinder bores—chromeplated,
hard-anodized, or case-hardened—
make burnishing difficult. The surface isn’t
ductile enough. Instead, the tool-mark
peaks must wear down during break-in
running. Fortunately ten-thousandths of an
inch or so of wear is usually sufficient. But
an even more important action takes place
when you break in your model engine:
dimensional stabilization.
When they are new, cylinders and
pistons can contain trapped stresses
produced by manufacturing processes.
During break-in, the heat generated by
combustion expands these parts and lets the
trapped stresses dissipate.
When the engine cools after running, its
piston and cylinder tend to return to their
original size—but not quite! The relaxed
stresses cause minute dimensional changes,
then those need additional “breaking in” to
achieve optimum running clearances.
That’s why it’s best—and fastest—to
break in a model engine via a series of
short runs, allowing a complete cooldown
after each. The repeated heating/cooling
cycles will then gradually bring the
moving parts into their ideal low-runningfriction
status.
November 2003 61
Most model engines with hardened
cylinder bores feature a “pinch” between
the piston and cylinder at the top of the
stroke to compensate for the expansion that
combustion heat produces.
In the chrome-plated brass cylinder
sleeves of Aluminum Brass Chrome (ABC)
engines, the combination of heat and high
pressure during the power stroke will
permanently slightly expand the upper part
of the sleeve. That frees up the pinch but
requires rather fast, hot operation during
break-in. Rich running is fine for breaking
in engines with soft steel cylinders—but it
won’t allow a new ABC engine to get hot
enough to do much good.
The Norvel engines, with their AAO
(Aluminum Aluminum Oxide) piston
cylinders, are another story! Chuck Morrow
(Anniston, Alabama) was surprised by the
unique break-in instructions that came with
his new Norvel .061 RC engine. They call
for removing the glow plug, coating the
cylinder/piston with castor or car-engine
oil, then letting that oil film “cure” for 24
hours with the piston at the bottom of its
stroke.
After that, the Norvel instructions want
you to put a propeller on the shaft and
rapidly turn the engine over backward at
least 100 times. (As Chuck learned, rotating
the new engine’s propeller
counterclockwise in the usual way causes
the propeller nut to loosen.) Continue this
backward rotation until the piston feels
reasonably free in the cylinder, then you
can reinstall the glow plug and can begin
running the engine.
The instructions for my similarly
constructed Norvel .15 and .25 RC engines
aren’t as elaborate as that. However, I
learned that the factory-recommended
break-in of 15 minutes wasn’t enough; it
took more than an hour of total run time
before either engine would idle dependably.
That is because of the one-piece
cylinder design. It makes for highly
effective cooling while the engine runs,
which is good, but it reduces the bore’s
expansion. Also, the stiffening effect of
the fins (acting like barrel hoops) restricts
the cylinder bore from expanding much.
That means it takes a great deal of break-in
running before the Norvels reach peak
performance. After that, they’re superb.
In a recent column I wrote about catalytic
reaction between brass and glow fuel. An
Atlanta, Georgia, reader (who asked me
not to use his name) mailed me a horribly
corroded brass clunk and wrote:
“You were right about brass clunks!
This one came out of the tank in my 3-
year-old Seniorita. I haven’t had time to
fly that as much as I want, and between
flights the airplane’s been hanging nose-up
on my garage wall. Yeah, I empty the tank
each time I fly. But I guess enough fuel
stays in there to cause this kind of mess.”
I’ve seen other examples of brass
corrosion, although none were quite that
bad. In one of my RC tanks the fuel-outlet
tube broke from internal corrosion, and
I’ve heard from other fliers who have
suffered similar problems. That’s why I
have quit using brass fuel tubing.
Rigid plastic fuel tubing for
automobiles was available in “model
sizes” awhile ago. I’ve used that and I
liked it, but I haven’t been able to find it
lately at any of the auto-parts dealers
around here.
Therefore, I checked with one of my
favorite mail-order “model hardware”
suppliers: Small Parts, Inc. (www.small
parts.com). It carries stainless-steel tubing:
1⁄8-inch-diameter, annealed 304 alloy. It’s
not cheap, at roughly $4-$5 a foot
depending on how much you buy, but it
will last forever in any kind of model fuel,
and it won’t catalyze methanol into rustinducing
acetic acid the way brass can.
(Small Parts Inc. is the place to find
almost anything in the way of model-size
hardware, such as screws, genuine music
wire, tools, and tubing. A comprehensive
catalog is free for the asking.)
Have you had trouble starting your engine
with a Hot-Shot type of glow-plug igniter?
If so, the most likely reason is that the
battery wasn’t getting the plug hot enough.
Here’s why.
This kind of glow igniter is usually
powered by a C-size Ni-Cd cell. The
nominal capacity of that runs at
approximately 1500 milliampere-hours
(mAh); that is, 1.5 ampere-hours worth of
power. Glow plugs normally need roughly
3 amperes for optimum heat. So a 1.5-
ampere-hour battery ought to light the plug
nicely for a half-hour, right? That’s plenty
of time to start an engine!
Theoretically, yes. But it’s not that
simple! First, Ni-Cd batteries are rated at a
lower current drain than 1 ampere. At a
much higher drain, their power-delivery
ability is substantially lower.
Second, Ni-Cd cell ratings are for new,
properly charged batteries. Glow igniters
seldom meet those criteria—especially
“properly charged.” It takes only 14 hours
for the typical plug-into-the-wall battery
charger to recharge a fully depleted Hot-
Shot-type cell. After that, excessive
overcharging can cause irreversible
chemical changes that reduce the Ni-Cd’s
capacity.
And most modelers (including me) tend
to forget how long they have left their Hot-
Shot igniters “on charge.” Sometimes that
can be for days.
The result of these “detrimental
factors” can be a glow igniter that is
capable of adequately heating the plug
element for only a few minutes. That
element may still get hot enough to let the
engine fire occasionally—but not quite hot
enough for a fast and dependable start. My
solution is to replace the Hot-Shot with an
industrial-type, D-size Ni-Cd cell. (At
www.megabatteries.com you can buy a
5000 mAh Ni-Cd D cell for approximately
$7.)
I soldered one end of a lamp cord to the
cell terminals and the other end to a plug
connector. After doing that, I lashed the
lamp cord tightly to the cell with a few
turns of Dacron cord, then I dipped the cell
in a container of red “toolhandle
compound.” It’s fuelproof and an excellent
electrical insulator.
I use the wall charger that came with
my Hot-Shot to recharge the D cell. That
cell’s capacity is high enough to eliminate
overcharging worries. (Mine is almost 10
years old and still works nicely.) MA

Author: Joe Wagner


Edition: Model Aviation - 2003/11
Page Numbers: 59,60,61

November 2003 59
CORRECTION: In the July column I
wrote that Brodak’s new .40 Control Line
(CL) engine displaced .426 cubic inch. I
was wrong about that. My faulty
information came from a data sheet packed
with the engine. It gave the displacement as
“7 cc”—which equals .426 cubic inch—and
gave the bore and stroke in metric
measurements too.
As Tom Hampshire (Belvidere, New
Jersey) quickly let me know, none of those
published dimensions was correct. Tom ran
the prototype test program for the Brodak
.40. Its actual bore (.831 inch) and stroke
(.722 inch) make the engine’s displacement
.392 cubic inch.
This apparently minor discrepancy
matters to fliers because AMA CL
regulations require a larger line size to be
used with engines larger than .40 cubic
inch. The Brodak .40 isn’t one of those!
In my most recent column I discussed
“breaking in” model engines on a test stand.
Several Radio Control (RC) fliers have
asked me why that is needed at all. “We
don’t have to break in power lawn mowers,
leaf blowers, or chain saws. What’s so
different about a model airplane engine?”
Plenty! Model engines run much faster,
and for their size they produce the most
power of any piston-type engine. That
means high pressures, temperatures, and
stresses. Also, RC engines must run
dependably through a wide range of speeds.
To the eye, the moving parts of a new
model engine appear quite smooth—even
polished—but a microscope will reveal tiny
serrations all over the surfaces. These are
tool marks: an inevitable result of the
machining or grinding processes used in
producing the part.
One reason for “running in” a model
engine is to allow the peaks of these
Joe Wagner
T h e E n g i n e S h o p
212 S. Pine Ave., Ozark AL 36360
Still being made after more than 50 years, the Fox .35’s steel cylinder sleeve needs rich
break-in running. (The NV assembly shown is a custom accessory.)
The O.S. .60 FP features ABC construction, needing a lean, hot break-in to expand its
brass sleeve and let the piston run freely at the top of its stroke.
A horrid example of what glow fuel can do to a brass clunk. This
took roughly three years—but it can begin in weeks.
Norvel .25 R/C, Brodak .40 use aluminum pistons in hardanodized
aluminum cylinders. They need long, careful break-in.
60 MODEL AVIATION
microscopic tool marks to slowly wear
away as the parts move in contact with each
other—or to “burnish” the tiny metal crests
down into the valleys.
Materials used in the engine make a
difference. If the cylinder bore is soft steel
(as in Cox and most Fox engines), a
burnishing effect during break-in is
The homemade glow igniter at left, using a D-size Ni-Cd cell, proved far more reliable
and longer-lasting than the typical Hot-Shot igniter on the right.
preferable to “wear-away.” That’s because
burnishing results in a much lower
coefficient of friction between the running
surfaces. Also, few metal particles are
eroded from the surfaces. (Those can cause
additional wear because of their abrasive
effect.)
Hardened cylinder bores—chromeplated,
hard-anodized, or case-hardened—
make burnishing difficult. The surface isn’t
ductile enough. Instead, the tool-mark
peaks must wear down during break-in
running. Fortunately ten-thousandths of an
inch or so of wear is usually sufficient. But
an even more important action takes place
when you break in your model engine:
dimensional stabilization.
When they are new, cylinders and
pistons can contain trapped stresses
produced by manufacturing processes.
During break-in, the heat generated by
combustion expands these parts and lets the
trapped stresses dissipate.
When the engine cools after running, its
piston and cylinder tend to return to their
original size—but not quite! The relaxed
stresses cause minute dimensional changes,
then those need additional “breaking in” to
achieve optimum running clearances.
That’s why it’s best—and fastest—to
break in a model engine via a series of
short runs, allowing a complete cooldown
after each. The repeated heating/cooling
cycles will then gradually bring the
moving parts into their ideal low-runningfriction
status.
November 2003 61
Most model engines with hardened
cylinder bores feature a “pinch” between
the piston and cylinder at the top of the
stroke to compensate for the expansion that
combustion heat produces.
In the chrome-plated brass cylinder
sleeves of Aluminum Brass Chrome (ABC)
engines, the combination of heat and high
pressure during the power stroke will
permanently slightly expand the upper part
of the sleeve. That frees up the pinch but
requires rather fast, hot operation during
break-in. Rich running is fine for breaking
in engines with soft steel cylinders—but it
won’t allow a new ABC engine to get hot
enough to do much good.
The Norvel engines, with their AAO
(Aluminum Aluminum Oxide) piston
cylinders, are another story! Chuck Morrow
(Anniston, Alabama) was surprised by the
unique break-in instructions that came with
his new Norvel .061 RC engine. They call
for removing the glow plug, coating the
cylinder/piston with castor or car-engine
oil, then letting that oil film “cure” for 24
hours with the piston at the bottom of its
stroke.
After that, the Norvel instructions want
you to put a propeller on the shaft and
rapidly turn the engine over backward at
least 100 times. (As Chuck learned, rotating
the new engine’s propeller
counterclockwise in the usual way causes
the propeller nut to loosen.) Continue this
backward rotation until the piston feels
reasonably free in the cylinder, then you
can reinstall the glow plug and can begin
running the engine.
The instructions for my similarly
constructed Norvel .15 and .25 RC engines
aren’t as elaborate as that. However, I
learned that the factory-recommended
break-in of 15 minutes wasn’t enough; it
took more than an hour of total run time
before either engine would idle dependably.
That is because of the one-piece
cylinder design. It makes for highly
effective cooling while the engine runs,
which is good, but it reduces the bore’s
expansion. Also, the stiffening effect of
the fins (acting like barrel hoops) restricts
the cylinder bore from expanding much.
That means it takes a great deal of break-in
running before the Norvels reach peak
performance. After that, they’re superb.
In a recent column I wrote about catalytic
reaction between brass and glow fuel. An
Atlanta, Georgia, reader (who asked me
not to use his name) mailed me a horribly
corroded brass clunk and wrote:
“You were right about brass clunks!
This one came out of the tank in my 3-
year-old Seniorita. I haven’t had time to
fly that as much as I want, and between
flights the airplane’s been hanging nose-up
on my garage wall. Yeah, I empty the tank
each time I fly. But I guess enough fuel
stays in there to cause this kind of mess.”
I’ve seen other examples of brass
corrosion, although none were quite that
bad. In one of my RC tanks the fuel-outlet
tube broke from internal corrosion, and
I’ve heard from other fliers who have
suffered similar problems. That’s why I
have quit using brass fuel tubing.
Rigid plastic fuel tubing for
automobiles was available in “model
sizes” awhile ago. I’ve used that and I
liked it, but I haven’t been able to find it
lately at any of the auto-parts dealers
around here.
Therefore, I checked with one of my
favorite mail-order “model hardware”
suppliers: Small Parts, Inc. (www.small
parts.com). It carries stainless-steel tubing:
1⁄8-inch-diameter, annealed 304 alloy. It’s
not cheap, at roughly $4-$5 a foot
depending on how much you buy, but it
will last forever in any kind of model fuel,
and it won’t catalyze methanol into rustinducing
acetic acid the way brass can.
(Small Parts Inc. is the place to find
almost anything in the way of model-size
hardware, such as screws, genuine music
wire, tools, and tubing. A comprehensive
catalog is free for the asking.)
Have you had trouble starting your engine
with a Hot-Shot type of glow-plug igniter?
If so, the most likely reason is that the
battery wasn’t getting the plug hot enough.
Here’s why.
This kind of glow igniter is usually
powered by a C-size Ni-Cd cell. The
nominal capacity of that runs at
approximately 1500 milliampere-hours
(mAh); that is, 1.5 ampere-hours worth of
power. Glow plugs normally need roughly
3 amperes for optimum heat. So a 1.5-
ampere-hour battery ought to light the plug
nicely for a half-hour, right? That’s plenty
of time to start an engine!
Theoretically, yes. But it’s not that
simple! First, Ni-Cd batteries are rated at a
lower current drain than 1 ampere. At a
much higher drain, their power-delivery
ability is substantially lower.
Second, Ni-Cd cell ratings are for new,
properly charged batteries. Glow igniters
seldom meet those criteria—especially
“properly charged.” It takes only 14 hours
for the typical plug-into-the-wall battery
charger to recharge a fully depleted Hot-
Shot-type cell. After that, excessive
overcharging can cause irreversible
chemical changes that reduce the Ni-Cd’s
capacity.
And most modelers (including me) tend
to forget how long they have left their Hot-
Shot igniters “on charge.” Sometimes that
can be for days.
The result of these “detrimental
factors” can be a glow igniter that is
capable of adequately heating the plug
element for only a few minutes. That
element may still get hot enough to let the
engine fire occasionally—but not quite hot
enough for a fast and dependable start. My
solution is to replace the Hot-Shot with an
industrial-type, D-size Ni-Cd cell. (At
www.megabatteries.com you can buy a
5000 mAh Ni-Cd D cell for approximately
$7.)
I soldered one end of a lamp cord to the
cell terminals and the other end to a plug
connector. After doing that, I lashed the
lamp cord tightly to the cell with a few
turns of Dacron cord, then I dipped the cell
in a container of red “toolhandle
compound.” It’s fuelproof and an excellent
electrical insulator.
I use the wall charger that came with
my Hot-Shot to recharge the D cell. That
cell’s capacity is high enough to eliminate
overcharging worries. (Mine is almost 10
years old and still works nicely.) MA

ama call to action logo
Join Now

Model Aviation Live
Watch Now

Privacy policy   |   Terms of use

Model Aviation is a monthly publication for the Academy of Model Aeronautics.
© 1936-2025 Academy of Model Aeronautics. All rights reserved. 5161 E. Memorial Dr. Muncie IN 47302.   Tel: (800) 435-9262; Fax: (765) 289-4248

Park Pilot LogoAMA Logo