Carburetors

The original 12mm Keihin PA carburetor on the NC50 (shown on right), really was problematic for me. The pressed in main jet really limits any easy tuning. The carb didn't respond well to temperature and density changes, so much so that I zip tied a small screwdriver to the NC50 so I can readjust it on days that it didn't want to idle well.

I moved onto the Dellorto SHA 15.15, middle. It does have replaceable main jets and has been much better. The idle air/fuel mixture, you have to adjust by replacing slides, which is unique to my experience. Treatland, http://treatland.tv, has an adapter that will let you bolt this carb directly to the NC50 stock intake. This makes it a very easy swap. The large 60mm air horn on the carb limits the choices in air filters, but again, Treatland has some options and I am using a Mallosi air filter from their site.

I have purchased a Mikuni 20mm VM round slide carb and a Weak Ends intake to use with it. Again, both from Treatland. My suspicion is that I will like this carb since it is similar to many of the Motocross VM carbs that I worked on in the 80s and 90s. The Mikuni has a 43mm air horn.

Right now, I running about 10,000 rpm at 38.6 mph. If we use the Jennings formulas for carb sizing, and the K-Star 44mm cylinder I am using, safe carb bore size comes out 19.6mm.

Starting Port Molds

Starting to make some molds of the ports on the stock cylinder, so that I can take measurements.

Seal off the inside of the cylinder. This is a stock 1978 NC50 cylinder.

I am going to have to do this in two steps, since I have to set the cylinder at an angle to fill the exhaust port. The transfer ports I can lay the cylinder flat. Grease the inside of the port:

Microwave the molding compound. In this case, I am using Composimold:

Pour into the exhaust port:

Fill it up:

On the left is the Composimold exhaust port from the stock NC50 cylinder. I get Composimold from Amazon, http://smile.amazon.com/ComposiMold-FC40-Reusable-Material-40-Ounce/dp/B0096BSQ2S/ref=sr_1_1?ie=UTF8&qid=1438548890&sr=8-1&keywords=composimold. I used to use another molding material, that was a reusable molding compound available in different stiffnesses. It was like a silicon rubber, but turned to liquid in the microwave and reverted back to rubber state quickly. An example of this is the ZXI 1100 cylinder head mold on the right. I still have some, but I am using it sparingly since every time you use it, you introduce some contamination. I don't like the Composimold that much, but it is readily available.

Now the transfer ports.

From the port mapping, I get a rear transfer port width of 18.54 mm as a cylinder segment width. Converting to chord length, this is 17.88 mm. Measurement of the mold, I get 18.02 mm.

From the port mapping, I get the magneto side transfer port width of 19.20 mm as a cylinder segment width. Converting to chord length, this is 18.47 mm. Measurement of the mold, I get 16.51 mm.

From the port mapping, I get the clutch side transfer port width of 19.24 mm as a cylinder segment width. Converting to chord length, this is 18.50 mm. Measurement of the mold, I get 16.67 mm.

From the port mapping, I get an exhaust port width of 29.42 mm as a cylinder segment width. Converting to chord length, this is 26.83 mm. Measurement of the mold, I get 22.02 mm.

Starting Port Map

Exhaust Port Open

I get the exhaust port opening at 29.7mm from top dead center, this would put the angle at 113 ATDC. http://hondaspree.net/wiki/index.php5?title=Cylinder_Port_Timing gives the export port open at 65 degrees BBDC which would be 115 degrees ATDC. The shop manual also gives the exhaust open as 65 degrees BBDC. This seems pretty late for two stroke motors I am used to. Also note that the exhaust port on the cylinder I mapped wasn't very symmetrical.

The other problem you can see both from the port map and the above picture is that at Bottom Dead Center, the ports are not fully open in this picture of a stock cylinder. This is the original motor from my NC50.

Transfer Port Open

The transfer port opening measurements I get are 34.06, 33.94 and 34.02. Using 34mm as a rough average, this gives a transfer port opening of 130 degrees ATDC. http://hondaspree.net/wiki/index.php5?title=Cylinder_Port_Timing give the transfer port opening (scavenge) at 47 degreess BBDC or 133 degrees ATDC.  The shop manual also lists 47 degrees BBDC as the scavenge open time.

Non-symetrical small ports that open late in the stroke don't make for much power or RPM.

Measuring the crankcase volume at 290cc at TDC.

Measuring the stock head at TDC, I get 8cc. Calculating the compression ratio:

Stock bore - 40mm

Stock stroke - 39.6mm

Stock displacement - 49.76cc

Static Compression Ratio - 5.65:1

Full Stroke Compression Ratio - (49.76+8)/8 = 7.2:1

The shop manual gives the compression ratio as 6.5:1 for the 1978 NC50 (mine). Any way you look at it, the compression ratio is anemic.

First look at the K-Star 44mm bore cylinder, the ports don't look that different in timing. With the rear transfer port bottom at 46.95mm, that is 10mm further down than the stroke. To really use this cylinder, I probably need to cut 10mm off the top of the cylinder and get or make a cylinder base plate spacer.

Exhaust Port Open
28.74mm gives an exhaust port open at 109 degrees. My gut says I would like to see something more like 89 degrees.

Trying to find the correct hub for the two speed

This is the stock rear wheel:

Pythonista

In the 80's and 90's, I used to write my two stroke programs in TI Calculator Basic. Mainly I didn't want to risk a $3000 laptop in the garage around heavy things, gasoline, oil, etc. So I would write it in TI Calculator Basic and risk a $100 calculator to the environment of the garage.

Between JetSki's and Georgia's rules on mopeds, I have renewed interest in two stroke motors, I have been resurrecting some of the programs I wrote back in the 80's and 90's. I have found that if I write / re-write them in Python, they will run on Linux, Mac, Windows and with Pythonista on my iPad and iPhone.

http://omz-software.com/pythonista/

I am finding I like Pythonista a good bit and I am liking the reduced decoration requirements of Python (Perl is historically my preferred scripting language).

An example of how I am using it, is my hp.py script, which I have up on GitHub:

https://gist.github.com/edt11x/52c69a6448f7a379ad19

Oil Injection Block Off

I bought from Ebay a Oil Injection Block Off for a Honda DIO. It's close but not quite right, but close enough that I will buy a couple more and modify them.

The grommet that seals the oil injection block off is right about where the hole for the outside bearing oil goes. I have taken off all the metal below grommet and will use a cut off tool to cut a new location for the grommet.

There's the oil hole which provides oil to the back side of the crank bearing that I am worried about.

Two Speed Motor

Happy that these later motors have case bolts rather than Phillip head screws:

Motor seems like it hasn't been apart, bolts look fairly clean, old spark plug doesn't show anything scary.

Insides:

Downside is the motor appears to have very little compression.

Output shafts are different lengths. Need to get a rear wheel from the two speed honda express.

SIP Horsepower Dyno

Program I am playing with attempts to estimate a HP curve by listening to the RPM and having you input gear ratios and weights:

http://www.siphorsepowerdyno.de/siphorsepowerdyno_en/home.html

http://forum.xda-developers.com/showthread.php?t=1821522

One of my runs:

Leak Down Test on Original Motor

The rule I learned, was that if you sealed the intake and exhaust, a 2 stroke motor should hold 5 lbs for 5 minutes.

The original motor doesn't fair too well on the leak down test. I guess I shouldn't expect anything from 35 year old seals.

Hand pump to generate the 5 psi:

Leak down gauge to measure the result:

Into the cylinder:

Test tube rubber stopper clamped in the exhaust port:

And and intake part that has been sealed with plumbers Goop. I just filled the whole turn down:

I also did a straight cylinder compression leak down test with the piston at top dead center, to test the original rings. It lost pressure quickly, so much, I didn't bother to record the data. 

Using the wind up starter, I was able to get up to 100 psi on a compression gauge.

The original motor was running, I was using it. On a good day I would get 28 mph. If I reverse this into RPM, it would be 7234.

Notes on the Stock Motor

1978 HONDA NC50'78 EXPRESS

NC50

A 49 cc scooter which was manufactured from 1977 until 1983, it featured an air-cooled 2-stroke engine. The versions from 1977 until 1980 had a manual choke, points ignition, and a spring starter. Versions from 1980 onwards included a kick start, capacitor discharge ignition, and automatic choke as well as an improved compression ratio. Most versions include an automatic one speed transmission with a centrifugal clutch, double suspension, and oil injection. Two speed automatic transmissions were introduced towards the end of their production run.

• ENGINE
  • Engine: 49cc 2-stroke single cylinder
  • Bore 40mm
  • Stroke 39.6mm
  • Compression Ratio 6.5:1
  • The engine was lubricated by oil injection.
  • Spark plug: BP4HS
  • Plug gap: .028
  • The ignition was controlled by breaker points.
  • It also had a manual choke.
  • The starting system used a wind-up spring.
  • Battery: 6N4-2A-6 (see also 6N4-2A-8)
• DRIVETRAIN
  • Transmission: only one speed
• CHASSIS / SUSPENSION / BRAKES
  • N/A
• DIMENSION
  • N/A
• OTHER
  • Colors:
    • Parakeet Yellow
    • Pal Green
    • Mars Orange
  • It had a two-tone black and white solo seat.
  • The downtube featured a logo change with the name "EXPRESS"
  • Serial number began NC50-2000013.

Rear Wheel Diameter, approximate 18.5”

Rear Wheel Circumference 58.2” (pi * diameter = pi * 18.5)

I measure a little over 58.25

The stock tires are 2.25 - 14. The 14 gives us the rim diameter, 14 inches. The 2.25, the way I understand it is that this gives an imaginary circle sitting on the rim, representing the tire:

And we know the tires are not ideal circles. In the case of the stock tires, 2.25 - 14, we would expect the outside diameter of the tire to be 14 inches + 2 x 2.25 = 18.5 inches. Then the circumference would be 18.5 inches x pi = 58.12 inches. In the ballpark of what we measured.

Gear Ratio 14.220/1

(from http://endless-sphere.com/forums/viewtopic.php?f=10&t=59952 )

Service manual also gives the Gear Ratio as 14.220 : 1

MPH = (RPM * WheelCirc/GearRatio * ft / 12inches * mile / 5280 ft * 60 mins/hour)

38.7 MPH = 10000rpm * 18.5 * pi  ÷ 14.220 ÷ 12 ÷ 5280 * 60

Gear Ratio that SIP wants is the Gear Ratio (14.220) divided by the tire circumference in meters (18.5” is 0.4699) which is 30.2617

Old engine 100psi of compression

Seals leak down within a minute

Some additional information from https://sites.google.com/site/vpcrabtree/home/honda-nc50-tuning

Measured Parameter	Value
Bore	40
Stroke	39.6
Con Rod Length	80mm (est)
Number of Transfer ports	2 plus 1 boost
Main transfer Height	6mm
Main transfer chord width	17mm
Boost port height	6mm
Boost port chord width	17mm
Exhaust port Height	9mm
Exhaust port chord width	19mm
Reeds Cage	2 petals, 10 by 20mm windows
Carb dia	12mm
Exhaust Type	Plain Muffler
Combustion chamber volume	7cc
Squish Clearance	~1.5mm
Compression Ratio	~6.6

Ultrasonic Analyzer

I use FFT iPhone programs to find my peak RPM. I have a new find that is much better than the ones I was using, Ultrasonic Analyzer. Right now I am turning about 9660 rpm, 60 * 161Hz.

Jet down to 74

Jetting down to a 74 main on the Dellorto SHA 15.15 carb. I've been very cautious about approaching a good main jet size, starting very rich.

Finished Oil Sight Plug

I made it a 0.024" interference fit and coated the joint with a little bit of epoxy, so it shouldn't leak like the original sight.