20th October 2014 - Door handles fabricated, doors installed

I made the internal door handles today. Using the ball turning attachment on the lathe I rounded both ends of two pieces of 8mm aluminium rod and one end of two pieces of 12mm aluminium rod. The 12mm rod was then cross-drilled and reamed to 8mm and drilled axially from the flat end with a 1/4" drill to match the latch spindle on the door. The 8mm rods were then pressed into the holes in the 12mm rod.
With the handles installed on the door spindles they were then cross-drilled with a 3/32" drill to take 12mm long 3/32" spring cotter pins (also sometimes called slotted spring tension pins). Then backing plates were "printed" to fit on the door and over the end of the 12mm rod. These were secured to the doors with double sided tape. The handles were then fitted and the cotter pins tapped in to hold them in place. Finally the doors were hung on the fuselage.
The second picture, taken through the passenger window, shows the pilot door shut. The handle lifts up to open to avoid it being opened accidentally.
Overall I'm quite pleased with the look of the cockpit and the way the doors fit into the framing.

29th October 2014 - Doors glazed

After superb weather over the last few days when I have been working in the garden, today it's raining so now the door paint is fully dry I completed glazing the doors. Like the rear windows, I'm just screwing and gluing the window to the frame. For the doors I'm using 5mm thick polycarbonate to add some structural stiffness. I've stuck some P-seal round the top, rear, and bottom of the door to seal it to the fuselage when closed. I've also applied P-seal to the fuselage along the front of the door jamb. Next job is to make some internal handles. As these will be used to pull the door in against the seal my original idea of printing them won't work so I'll get some aluminium rod and turn some up on the lathe - another item to add to the to-do list.

25th October 2014 - Rear windows and baggage compartment

It has been a superb autumn day today so I was able to get lots done working outside. The first picture shows a last view of the inside of the fuselage before the baggage compartment panelling is installed.
The first job was to cut out a pair of rear windows from 2mm thick polycarbonate sheet. These were drilled and countersunk to match the cut-out in the fuselage.
I don't want to use any sort of frame which just adds unnecessary weight so the windows are going to be bonded and screwed to the aluminium panels.
In order to make sure that the join stays flat I drilled some scrap pieces of aluminium angle to match the holes in the fuselage and then applied a bead of Dow Corning 738 adhesive/sealant to the inside of the frame. The window sheeting was then clecoed through the window frame and into the aluminium angles. I'm going to leave it like this for at least a week for the silicon sealant to fully set. Then the aluminium and clecoes can be removed and replaced with #6 stainless screws.
With the windows in their final location I could then re-install the baggage compartment panelling. The aircraft looks very different and much closer to finished once the inside of the fabric is hidden.

24th October 2014 - Rework and a few more "to-do" items

The first job today was to check out/fix the problems from the engine run.
A battery charger was attached to the alternator "B" lead and turned on. This confirmed that the ammeter sender was the wrong way round so it was then just a case of unbolting the two connections and reversing them. A repeat test with the charger confirmed this was now correct.
The CHT/EGT gauge was unscrewed from the panel and the connections of the two thermocouples reversed. Looking at the Westach instructions I can easily see why I got this wrong - the pinout in the instruction sheet doesn't really match my version of the instrument. The change was then tested successfully by using a heatgun on the two thermocouples.
The oil temperature sender was removed from the engine and the heatgun used directly on it - nothing. A 1K ohm rheostat was then connected in its place and the gauge responded as expected (low temperature=high resistance and visa-versa). Checking the sender with a multimeter showed it was open circuit so a new one has been ordered. This will take a week to arrive.
The intermittent fault with the fuel gauge was more of a problem. I had fitted a co-pilot PTT switch to the panel a few days ago. This required removing the KMD150 GPS and then reaching though the side of its rack to get to the back of the extreme left of the P2 side of the panel. I suspected that I must have disturbed a connection in doing this. The only way to find it was to remove all the instruments and the radio from the right side of the panel. In each case they had to come out through the side of the GPS rack. Once they were out I was able to establish that the problem was the positive voltage supply to the fuel gauge so this connection was re-made and the gauge started working again. Then everything had to be refitted. I just hope I haven't disturbed anything else! Next time the engine is run I will see if the vibration causes any more problems.
Once everything was back together I could get on with some new jobs. The lower cowl was test fit with the spinner and propeller in place. In no way would it go on because of the way the baffles fit under the lower edge of the airduct and into the spaces either side of it. I'll think about this some more but at the moment it looks like the propeller will have to come off any time I want to remove the lower cowl. This is less than convenient but certainly not a "show-stopper".
Then a hole was drilled in the cowl, just visible in the bottom right in the picture, to allow a fuel tester to be applied to the gascolator drain.
Pieces of baffle material were pop-riveted to the front of the airbox. These will fit over the internal extension to the port I made in the front of the lower cowl airscoop to roughly seal it and create a small ram-air effect.
Finally, pieces of heat resistant material were glued to the cowl using high temperature tank sealant where the exhaust pipes are closest. The edges were sealed with self-adhesive aluminium tape.
The lower cowl was then fitted in place and the propeller and spinner backplate re-fitted. The backplate only has 1/8" clearance from the cowl which confirms that without a different approach to the baffles it cannot be removed and refitted with the propeller in place.
I'll next run the engine once the new temperature sender arrives in about a week, Then I'll also look to adjust the vacuum regulator and further diagnose the low compression on cylinder 4. I've decided that the rear of the exhaust pipes are short enough not to require any further support so I've removed that item from the to-do list. New items have been added in red.

23rd October 2014 - First engine run

The replacement coil for the left magneto arrived today. My local aircraft engineering company is also certified for maintenance and release of magnetos, so under their supervision the new coil was fitted and I learned how to re-assemble the magneto and time the rotor arm correctly. The magneto was then put on their test bed and a healthy spark demonstrated.
Back home the propeller was used to turn the engine until it was at top-dead-centre on cylinder one. A TDC position is marked on the flywheel but will correspond to either cylinder one or cylinder two. As my engine has conventional tappets it was easy to check which by removing cylinder one's rocker cover and ensuring that both valves were closed. The crank was then rotated until 25 degrees before TDC which is the standard firing position for this engine. The magneto was also then positioned in the firing position for cylinder one using a special pin though a hole in the top of the magneto to locate a matching hole in the rotor arm drive gear. The magneto was then inserted into its position in the accessory housing at the back of the engine.
Then the flywheel was repositioned in the 25 degree before TDC position. When timing an impulse coupled magneto it is essential that the engine is moved past the point at which the impulse coupler trips and then rotated back to the 25 degree position. Using the magneto timing device  I made a few days ago, the magneto was then turned until the points were just opening and the magneto clamped into position.
As a final check the plugs for the left magneto were removed (cylinder 1 & 3 top, cylinder 2 & 4 bottom) and the engine rotated with the propeller. A spark duly appeared at the correct plug and at the correct time allowing for the delay of the impulse coupling (this is a mechanical device that delays the spark when engine revs are low to allow for easier starting).
Then it was time to try and start the engine. First job was to secure the aircraft to ensure it couldn't run away. To do this it was chocked but also chained to the back of my car! My neighbour, Daryl, kept watch to ensure that no-one came close to the propeller. The elevator trim was moved to hold full up elevator to ensure the tail wouldn't lift.
I've fitted a conventional pump primer to the engine which feeds all 4 cylinders; 4 strokes were used. Then with the master on and then left impulse-coupled magneto on the starter button was pushed. A few coughs and splutters ensued but the engine didn't want to run. Then Daryl saw some liquid under the aircraft. This proved to be water and further inspection found that the airbox was full of it! Obviously the recent rain had somehow ended up in the airbox and the poor engine was trying to start on a fuel/air/water mixture. The airbox should have had a small hole in its lowest point - it has now! To solve the problem temporarily the airbox was disconnected and left hanging down below the carburettor. The spark plugs were removed and dried. The engine was turned over on the starter without the plugs in to ensure there was no water in the cylinders and then the plugs refitted.
Finally it was time to try again and this time the engine started perfectly. It ran for a few seconds on the prime before running out of fuel, presumably because the float chamber in the carburettor had been completely empty. However, after re-priming the engine started and with the right magneto also on began to run smoothly.
Then it was time to do some preliminary checks:

• the cylinder head temperature was registering off the scale but the exhaust gas temperature gauge was not moving - looks like the thermocouples are wired the wrong way round. 
• the bus voltage looked good but the amp meter was showing a large discharge - I had a 50:50 chance of getting the current sensor round the correct way and had guessed wrong.
• The oil pressure was good but the oil temperature seemed pegged. Either the vernatherm hasn't opened or there is a fault in the temperature gauge or sender- another thing to check
• The fuel gauge had stopped working - presumably a bad connection somewhere
• The vacuum gauge was reading a bit low but the gyros were definitely spinning up

With these basic checks completed the engine was stopped and an external inspection carried out:

• No oil leaks
• Nothing coming loose
• High cylinder head temperature confirmed as a wiring error
• Good compressions on cylinders 1, 2, and 3 but not on 4
• Crankcase still cool so oil temperature not yet confirmed as an issue

Given that there was nothing untoward the engine was restarted and run again for ten minutes. This time the engine was run up to 2000 rpm to allow for proper mag-checks which were well within limits. A very quick burst at full power saw a static RPM of 2250. Time to stop annoying the neighbours and pack up for the day.

18th October 2014 - Magneto Woes and Timing

Yesterday I finalised all the jobs on the check-list needed to attempt a first engine start. However, things did not spring into life as hoped.
The engine starts on  the left magneto only which has the impulse coupler but nothing happened. Fault finding showed that the secondary field winding in the magneto coil was open circuit and no spark was forthcoming. I will liaise with my engineer Monday to see if the coil can be replaced or a serviceable magneto will need to be procured.
When the new magneto is refitted it needs to be timed to the engine so that the points open at 25 degrees before top dead centre.
To do this a circuit is required that can tell if the points are open or not measuring from the P-lead connection. The problem is that when closed the resistance to ground is zero, when they are open it will be the primary coil resistance which is roughly 0.7ohm.
Commercial magneto timing testers are widely available but expensive and based on very old technology so I decided to build my own.
As with the commercial offerings, the key to the design is to feed the P-lead with an alternating current. The inductance in the primary will then act differently to the short to earth and the difference can be measured.
The circuit uses a picaxe microchip to generate a 2KHz square wave which is applied to the P-lead through a 330 ohm resistor. The voltage on the P-lead is then rectified using a schottky diode, converted to DC using a simple single pole filter and applied to an analogue input on the picaxe. This continuously measures the voltage and if it is above a threshold level (6% of the microprocessor supply voltage) lights an LED. The circuit is powered by a single 3V coin cell and works perfectly. Total cost approximately $5!

16th October 2014 - Voltage Regulator installed

Although you can buy voltage regulators fairly cheaply, particularly second hand, I decided to build my own. There is a single chip solution available, the L9911 from ST Microelectronics. This is constructed with an easily mountable package like a large transistor and the pins are on a 0.1" pitch so it is easy to mount on standard stripboard (Reinforce the tracks with solder as currents up to 3 amps will be flowing). The tab on the package is earthed so this can be fastened directly to the aircraft chassis via the aluminium project box I have chosen. Some thermal conductive paste was used to ensure that any heat generated would be dissipated to the firewall. The regulator needs a light or LED connected between the "ignition" pin and the regulator supply which it senses and this turns it on. I used a panel mount LED with a 1K ohm resistor as a current limiter.
Unlike most aircraft regulators it needs a phase input from the alternator which you take from between any one of the pairs of diodes. Some alternators have a connection for this. Mine didn't but it was easy to add an extra wire to the alternator. The regulator uses this to sense that the alternator is turning and switch the field in sync with the rotation. This helps to reduce any chance of radio interference. The regulator also only applies a small current to the field if the alternator is not turning unlike most aircraft regulators which apply full current any time the voltage is below the regulator set-point. The LED illuminates if the alternator is not turning and goes out when the regulator senses the phase input.
Other than that you only need to connect the +ve supply and the field output as per any other regulator. The large capacitor is used to smooth the regulator +ve supply which it uses to compare with the set-point and control the field current. Total cost of the completed regulator, including the box, was less than $20.
Today, I also installed the oil temperature sender and replaced the internal battery in the KMD150 GPS. This is a bit of a fiddle as it requires the unit to be stripped down. However, the new battery should last about 8 years. There are good instructions on the web how to do this if you google "KMD150 battery replacement".
So that is another three items ticked off the check list:-)