Quarter Scale Merlin V-12

[mayhugh1]

Peter,
Yes, my effective groove depth was .050". I, too, saw small variations on the order of a few percent for the recommended compression depending on the particular reference. One reference I came across said to use no more than 40% because the o-ring material would take a permanent set with an eventual loss in sealing. So, I didn't feel like 5% or so was significant enough to be concerned with. I have some expensive high flow pond pumps that have static o-ring seals that are close to 35% compressed and those rings invariably take a set, leak, and have to be replaced during my yearly maintenance on the the pumps. And then, I also have some 7 year old underwater low voltage lights that are less than 10% compressed and feel like they would never be effective seals, but they have never leaked and have never been replaced. Although the references I checked said nothing about the particular o-ring material used, I would have thought it would have also mattered whether the o-ring was viton, buna, or silicone. -Terry

 

[Ken I]

The dynamic nature of an "O" ring is that it behaves like a fluid and resists any amount of applied pressure. Up until the "O" ring starts to extrude out of the clearance then it all goes South really quickly.

So even small amounts of compression are very effective - what is way more important are clearances and surface finishes.

You only need enough preload to seal against surface imperfections - after that physics takes over.

I have used "O" rings as piston seals for destructive pressure testing of steel tubes up to 16000 psi - the only failures were because of excessive clearance.

And while we are on the subject, using two "O" rings doesn't work - next to each other or even in two grooves - particularly in a dynamic environment - they can generate enough pressure between each other to bring about extrusion failure.

Regards,
Ken

 

[mayhugh1]

After finishing the machining on the liners I needed a short procrastination break before wading into the messy honing/lapping process. So, I thought I would take an initial stab at selecting the Merlin's starter motor. I began with a set of 'nice to have' specs which included 6 volt operation and a torque of 14 ft-lbs at 240 rpm available to the crankshaft. The 6 volt requirement was a 'nice to have' since I've been running my engines from the same small 6 volt gel cell whose voltage is compatible with the ignition modules, fuel pumps, and other accessories that I've typically added to my engines. The torque requirement came from cranking measurements that I made on my 18 cylinder radial during its build, and as a starting point I expected the Quarter Scale's requirement to be similar. The 240 rpm cranking spec might be ambitious and maybe even a little over-kill for a twelve cylinder engine, but I needed to start somewhere. A quick horsepower calculation using these 'nice to have' requirements gives (14 ft-lbs)x(240 rpm)/5252 = .64 hp or, equivalently, 477 watts. Assuming a dc brush motor efficiency of 70%, such a motor would draw (477 watts)/(6 volts)/.7 = 114 amps from my 6V battery.

With space only for an under-two inch diameter motor, these power requirements seemed a bit excessive, and so I began again with a set of minimum 'must-have' specs. I reduced the torque requirement to 10 ft-lbs in order to adjust for the Merlin's twelve cylinders rather than continuing to assume the requirements of the eighteen cylinder engine. I considered 60 rpm to be an absolute minimum cranking speed but continued to hope for 120 rpm. I then decided I'd be willing to make the changes necessary to run the Merlin from a 12 Volt battery or else use a separate battery for the starter. With the same 70% efficiency assumption and my new 'must have' requirements, the current draw from the 12 V battery fell to about 10 amps.

Since there is an internal wheel case gear reduction of five between the starter and the crankshaft, my dc motor requirements became 2 ft-lbs torque at 300 rpm (600 more desirable) measured at the shaft of a 12 volt dc motor. The diameter of the motor needed to be no more than 1-3/4", but its length could be as long as 4 or 5 inches.

I initially limited my search to dc brush motors because I knew they were significantly cheaper, and there seemed to be more readily available application information for using them in scratch designs. In addition, a first glance showed I was going to have to become familiar with a whole new world of jargon if I was going to navigate the world of R/C brushless motors. Finding a good brush candidate first will give me something against which to compare brushless alternatives so I can determine whether their increased performance will be worth the additional cost and hassle of controlling them.

After hours of online comparisons, I focused on an inexpensive ($9) Nichibo 775-8511FDAS dc brush motor whose specs seemed to be representative of the high output motors available in a packaging envelope I could live with. One has to be careful when comparing these little motors based on their marketing specs because the high torque, low current, and high efficiencies that are often advertised may not be available simultaneously. I've included a set of performance curves for this particular motor that I also found online. In fact, the availability of these very useful curves was one of the reasons for selecting this particular motor.

The curves for this 14.4 volt motor were generated for operation on 12 volts, and so the 12 volt axis on the left hand side of the graph may be confusing since it has little meaning. If the motor were to be operated at a voltage other than 12 volts, then the curves shown would morph into a new set of curves for the new operating voltage.

It seems there is no standard for units of torque among the manufacturers of these small motors, and the kg-cm used by Nichibo is a little unusual. My 2 ft-lb. torque requirement is roughly equivalent to 30 kg-cm. To start the design process, I entered the horizontal torque axis at 1 kg-cm and then moved up to find its intersection with the operating point curves. Since I'll have to machine a 30:1 gear reducer to convert the 1kg-cm torque into my required 30 kg-cm, I must then calculate its impact on the motor's speed. The rpm curve is intersected at 17,500 rpm, and so my gear reducer will drop this down to 580 rpm which is (hooray!) greater than my minimum required 300 rpm and nearly equal to my more desirable 600rpm. The amperage curve intercept is at about 22 amps, and this will be the actual current draw from the battery. The power delivered by the battery will be 12V x 22A = 264 watts. The intercept with the output power curve is at 180 watts, and this will be the actual mechanical power that the motor will be putting out. The efficiency will be (180W/(264W)= 68% which agrees closely with the intercept of the efficiency curve.

I can reduce the current draw by reducing the required output power of the motor. And so for a lower output operating point I started the process again using the 0.5 kg-cm coordinate which is equivalent to .036 ft-lbs torque. In order to turn this into 2 ft-lbs, I will now need a gear reducer of 55. The new rpm curve intercept is 20,000, and the gear reducer will drop this to 360 rpm which is only slightly above my minimum requirement of 300 rpm. What I've gained from this bit of compromise is that the current draw has now been reduced to 13 amps, and the power required from the battery has been reduced to (12V)x (13A)= 156 watts. The new mechanical output power interpolated from the new intercept is 105 watts, and the calculated efficiency is now 105/156 = 67% which again compares favorably with the graph's efficiency intercept.

And so, a representative dc brush motor that appears to meet my requirements is a $9 motor that will provide 156-180 watts of output power. It will require a mechanical gear reduction of between 30:1 and 55:1 as well as a 12 V battery that can supply between 264 watts and 156 watts. The motor's diameter is 1.75" and its length is 3.8". Its shaft diameter is 5/32", and it has a milled flat.

I ordered a couple of these motors for some hands-on experience. If I don't end up using one of them for the starter, I'll at least have something to use for the supercharger tests I plan to run. The next step in the motor selection will be to find a comparable brushless motor so I can make cost, size, and performance comparisons. - Terry

 

[Naiveambition]

I anxiously await every post on this motor from the very beginning of your build, and I remember you had mentioned starter issues early on.
At the time I was doing some brainstorming and thought of a small starter for a push mower .
I had a John Deere 6 hp, with electric start that had a small starter and thought maybe this could be a solution. I believe it was a js30. It could possibly fit your parameters with a quick solution as being small and strong , with the added benefit of being made for these applications .
Apologies for the lateness of my post, but I often am mesmerized by your scientific approach to your solutions, and quite simply forgot.

Mike

 

[bigrigbri]

Why not make a scale sized rotary vane motor running from a Co2 bulb or high pressure air which will give plenty of torque especially geared down.
Brilliant work so far sir.

 

[Ken I]

Terry,
I've got quite a bit of experience in cranking out dubiously large amounts of power from small motors - for slotcar racing (a hobby) and for robotic applications (my usual line of work) where power to weight is an issue.

You can only get so much torque out of a saturated armature so the only way to get more is to spin the beans out of it and gear it down as I am sure you are well aware.

I posted on the site an article on getting more out of your A.C. motor (a lot more) which is germane.

www.homemodelenginemachinist.com/showthread.php?t=25236

Also from my automotive experience - starter motors are not 100% duty rated - they burn out if you run them too long but normally longer than it takes to crank the battery flat. So you can overload a motor for brief periods given the relatively light time-duty cycle of a starter.
With a motor that runs close to saturation this won't help but a motor that runs at 20% saturation can be loaded to 500% for brief intervals.
The "brushless" servo motors we use in our robots are built like this they will handle 200% load for 20 minutes and 500% for 5 seconds.

Anhoo - the reason for my response is that you will need to gear it down and those ganged planetary gear drive sets used in cordless drills and pneumatic tools can be cannibalised to form the basis for your reduction gearbox.

A lot of starter motors now do just that as opposed to the direct drive types on older cars.

And as Naiveambition pointed out you can go scrounging for an actual starter built for the purpose from something or the other - none spring to mind but perhaps some other members might make suggestions.

Another thought for a compact gearbox would be a harmonic drive - I could give you a perfectly serviceable 50:1 450W unit out of a robot (too much lash for precision but still a perfectly serviceable unit.) I have a few lying around.

Just a suggestion.

Regards,
Ken

 

[mayhugh1]

Thanks all for your informed comments - Terry

 

[mayhugh1]

The final step in the liner construction was to bring all the bores to a common diameter and then lap the i.d.'s for cast iron piston rings. Since neither the pistons nor the rings have yet been machined, the actual diameter isn't yet important. What is important is that the i.d.'s end up identical so a single diameter piston ring can be machined for use in all the cylinders. The resolution of my dial bore gage is about a tenth or so, and I was able to identically lap all the cylinder bores in my last two engines to within its resolution. So, I set the same goal for the Merlin's liners.

All twelve liners plus the three spares that were initially launched managed to survive their machining and avoid being scrapped. Initial measurements indicated all the bored i.d.'s were within a thousandth of one another, and so I was looking forward to a brief lapping session. However, the chatter marks with which I had struggled on some of the liners were deeper than I had hoped and weren't picked up by the anvils on my internal micrometer. After some localized clean-up on a couple of the worst looking parts, my dial bore gage showed the actual diameter spread was closer to .0025". This was really disappointing because the grooves on these parts ended up setting the final diameter of the entire group of liners, and what I had hoped to be just a single messy afternoon of lapping turned into days of grinding.

Each liner was engraved with a unique number so its progress could be tracked throughout the honing/lapping process. A set of worksheets was set up on which the measured i.d.'s of each liner were recorded at three different depths as the lapping progressed. With up to .0025" to be removed from many of the liners, I would normally have re-bored the whole lot, but I hadn't learned anymore about getting chatter-free bores on Stressproof than I knew when I bored the liners the first time. I zero'd the dial bore gage to the center of the spread so I could focus on deviations instead of trying to keep track of absolute numbers.

I used commercial Acro barrel laps designed for through-holes and Clover silicon carbide grease in three different grits: 280, 600, and 1000. A separate lap was used for each grit. The laps were mounted on an arbor and spun in a battery-powered drill while the liners were simply held in my gloved hand. This method of supporting the liners would have required greater care had I been working with the stock .030" thin-wall liners since a tight grip, especially needed for the coarse compound, would have tended to distort their bores. With the drill running at 300-400 rpm, the liner was slowly oscillated along the length of the lap with about half the lap being exposed at each end. The liners were always slipped onto the laps through their bottom end to avoid accidental damage to the sharp sealing edges at the tops of the liners. These sealing edges are an important reason to avoid using brush or bottle hones on the Merlin's liners.

Since the starting bore was a non-standard 1.198", I modified the stock Acro 1-1/4" brass barrels by turning them down to 1.195". With so much liner metal to be removed, I started with 280 grit grease. This coarse compound was extremely abrasive to the laps, and the grinding process removed about the same amount of material from the laps as it did from the liners. The expander bolts on the ends of the laps had to be continually adjusted as the lap wore, and it was necessary to take small steps in between many measurements in order to avoid overshooting targets. I probably made three to four measurements during the removal of each half thousandth. By the time all the liners had been ground to a common diameter, two 280 grit laps had been fully expanded. Measurements showed their starting diameters had been reduced from 1.195" to 1.180" by the grinding, and this volume of metal was nearly identical to the volume of metal that had been removed from the liners. Even though I was using barrel laps, I wouldn't consider this first step to have been a lapping operation. It was more of a controlled grinding operation used to bring all the liners to a common diameter.

The 280 grit grease produced a surprisingly smooth and beautifully frosted finish that was free of visible machining marks or scratches. The real lapping, though, began with the 600 grit paste after the liners were all at the same diameter. During this operation only a tenth was removed from each of the nearly finished bores, and maybe a thousandth from the single lap. Each tenth took longer to remove than when working with the 280 grit paste, and so it wasn't difficult to get all the liners to precisely the same diameter with fewer measurement checks.

I've learned during lapping that there is some technique involved with using a (my) dial bore gage. My only experience with these gages is with my own imported model, and so I don't know if a more expensive gage behaves in the same way. But mine seems to have one or two tenths backlash. When I rock the anvil back and forth inside a bore to take a measurement, I get a slightly different result depending on the direction that I rock the gage. I've learned to make it a point to approach all measurements from the same direction when making comparisons.

I finished the liners with 1000 grit compound which didn't remove any measurable amount of metal but did seem to improve the surface finishes a little. I'm not sure if the 1000 grit pass is needed or even desirable for cast iron rings, but since I used it on the cylinders in my last radial I also included it in the Merlin's liners.

My process for getting all the liners to a common diameter was not to grind on each one continuously until it was at that final diameter. Instead, I worked in small steps across the whole group of liners to finesse them all, as a group, across the finish line. This approach took more time and involved many small steps and measurements, but it helped to avoid accidents, and it better handled surprises such as undiscovered grooves or scratches. It was also easier to spread the work out over several less tiring sessions over several days when not grinding incessantly for long periods of time on the same part. Changes in the lap (and the lapper) were slowly spread out over a number of bores for a more consistent final result.

The 280 grit compound typically required about 5 minutes of actual lapping time to remove a couple tenths from a single liner, but when the times for cleaning and measurement were added, the total time was closer to ten minutes. It was important to thoroughly clean and dry the liner before each measurement to not only get consistent measurements but to also avoid damaging the gage's anvils. I used kerosene for removing the lapping grease, and it left me smelling so bad that neither the wife nor dog would come around me for days.

After the 280 grit grinding was completed, I let the parts sit for a day so I could rest up for the final 600 grit lapping step. For consistency, I used the 600 grit lapping operation to remove the last tenth from all the parts in one session. The final result was that 14 of the 15 liners finished out to better than a tenth with beautiful smooth finishes. There were no measurable circularity errors after (or even before) the lapping. The 15th liner had the deepest chatter marks, but It would have likely cleaned up with another two or three tenths pass. Since the other fourteen liners were already at the finish line, I decided that having a third spare wasn't worth the effort or risk to the other liners. - Terry

 

[rdefrei1]

Running engine from Dynamotive castings.

[ame]https://www.youtube.com/watch?v=KTPf3kZz9u0[/ame]

 

[ddmckee54]

Terry:

I check in every day or two to get my Merlin fix. It's been an immensely enjoyable ride so far. If you are talking about starters it's gotta be close to the time when this one is gonna be goin vroom-vroom. Or whatever the sound is that you've been hearing in your head when you look at this beast.

Whatchya gonna build next?

Don

 

[petertha]

I used kerosene for removing the lapping grease, and it left me smelling so bad...

Great write-up as usual Terry. Liners look very nice.

Kerosene is probably the only solvent I don't have in my arsenal. But for whatever reason I found WD-40 removed a healthy amount of lapping compound, leaving the harsher solvents reserved for final clean. WD-40 almost seemed to disperse the grit off the metal, but maybe I was hallucinating. I found thinners I usually prefer like acetone dilutes the grease but the grit is still kind of free floating & clingy.

 

[petertha]

Another thought for a compact gearbox would be a harmonic drive Ken

I've heard of them, but assumed they were brutally expensive. Do you have links for ones in/around this size?

Also, for model engine starter motor applications like this, would they be typically on/off or is there some (ideal) requirement to ramp up to rpm for whatever reason?

 

[Ken I]

I've heard of them, but assumed they were brutally expensive.

Correct: They are fiendishly expensive as are their cousins cycloidal drives.

However look up your nearest robot supplier (Yaskawa, Fanuc, ABB, Kuka etc.) and if you speak to them nicely they might give you an "old" unit.

I would.

These are often changed out because even minute amounts of lash become problematical on a robot - but they are otherwise still perfectly serviceable as a "gearbox".

Unfortunately they typically toss them in the bin as they are no use as a "spare" so you might have to wait.

http://makeagif.com/9jh2He

Above link to a harmonic drive in action - note these things can give 50 to 100 to one reductions at 70% efficiency (mind boggling). An efficiency of better than 50% is needed in robotic & N.C. applications to maintain control over deceleration of large inertial masses (do that with a 50:1 worm drive and it will self lock because of its less than 50% efficiency (and therefore non-reversible / self locking) and the inertial load will then proceed to tear off all the teeth).

https://goo.gl/images/65ZlNB

Above link for a GIF of a cycloid in action

and what they typically look like -

https://goo.gl/images/71F5H0

A complete and very compact gearbox / radial /axial thrust bearing assembly combined - when you see an industrial robot, each axis turns on one of these (including the main "S" axis on which the entire unit stands) with typical reduction ratios from 90 to 190:1 at 75% efficiency and zero backlash - really cool.

At first glance both of these types of drive look like they would be inefficient - just the opposite they are even more efficient than ganged planetary gear sets and can be built to interference fits to be backlash free (no really - zero lash) - but they do have to made to very exacting standards - which makes them very expensive.

They are also lubrication critical - pump in the wrong grease and the gearbox self destructs as many robot users have found out the hard way.

Sorry this post kinda got well away from the original question - a simple full on switch would only present problems if there is sufficient lash in the system to let it get a head start and result in an impact - so a soft start would be better - a simple two step would be fine - a resistive switch on at low torque to take up the lash followed a split second later by full on would be quite sufficient.

The only manufacturers I know are Harmonic Drive Systems (invented by an American CW Musser in 1957 but perfected by the Japanese).

https://en.wikipedia.org/wiki/Harmonic_drive

http://www.harmonicdrive.net/

The cycloidals I work with are made by Sumitomo

http://www.sumitomodrive.com/modules.php?name=Product&op=productBrand&brand_id=13


Regards,
Ken

 

[petertha]

Terry, re starter motor, just happened to stumble on this Conley V8 vid. Some views of starter motor around 20:20. First I thought it looked like an RC motor, but I don't recognize the open can so maybe brushed? Its circa 2011 so not sure if that's what he still uses.
Youtube search Conley Factory Tour Model V8 Working 1/4 Scale Engine

You mentioned avoiding brushless RC side if possible & I don't blame you. OTOH, they are getting quite inexpensive & available in multitude of KV & diameter flavors. You likely could utilize your same 12v drive battery as its close to 3-cell lipo ~11-12v nominal & amps are low. So the shopping list would be brushless motor, ESC rated for amp duty & cheap servo tester to throttle the ESC. Probably still $100 combined. But my electronics hits a wall with how to turn on/off though. The tester is meant to mimic TX signal & throttle motor between 0-100%. I suspect trying to insert switch [SW] between tester & ESC would cause ESC arming issues & generally be bad. Now whether the turn knob pot thingy could be modified or tricked.. above my pay grade Anyway, food for thought if push comes to shove. Some hobby CNC-ers are making spindle motors with this configuration but its continuous vs. on/off like starter motor. This vid shows larger diameter outrunner, but same principle.

http://www.raynerd.co.uk/brushless-dc-motor-cnc-high-speed-spindle/

 

[kiwi2]

Hi,
It's probably not on topic as far as constructing a Merlin engine goes, but I snapped the attached picture this afternoon.
A local guy has made a set of formers to construct plywood Mosquito fuselages. This is his second one and the twin Merlins came roaring overhead this afternoon. A very impressive noise. It's amazing these engines are still operating after 70 odd years.
Do you have any plans to put your quarter scale engine into (for example) a quarter scale scale Spitfire fuselage or doesn't the scaling work like that?
Regards,
Alan C.

 

[mayhugh1]

Peter,
Thanks for the post. I didn't realize there were such things as servo testers. I was expecting I would have to design something to drive the ESC if I went brushless, but one of those looks exactly what I would need.

Alan,
I really don't have any plans for the Merlin beyond a display shelf.

Terry

 

[ddmckee54]

If you aren't familiar with ESC's and servo testers, there are a couple of things I'd recommend.
1 - I'd go with an aircraft ESC, not a car or boat ESC. That way you only get one direction out of the ESC, instead of a bi-directional output with full stop at mid-range.
2 - I'd replace the pot in the servo tester's input circuit with a switched fixed resistor, you're building the equivalent of a retract switch. That way the output of the ESC will either be on or off,

Don

 

[mayhugh1]

With the cylinder liners completed, the next step was to machine the pistons. As shown on an earlier assembly drawing, I had to make several modifications to the stock piston design in order to adapt it to my modified liners. Despite the fact that in the process I scaled them down by some 10%, at 1.197" diameter by .830" long, they're still the largest pistons I've machined to date.

The Quarter Scale's pistons are similar to the 'short length' pistons used in some versions of the full-size engine. They use two rings and an oil control groove located immediately below the second ring. During the piston downstroke, oil is scraped from the cylinder wall and into this groove by the lower ring and allowed to escape into the interior of the piston through a series of radially drilled holes. Both radials that I've built had a non-trivial third ring designed specifically for oil control. Hopefully the Merlin's approach to oil control, which is much easier to implement, will perform as well. The lack of a third ring also reduces the friction that the starting system will have to deal with.

I used .003" (diametral) piston-to-cylinder clearance which is about the same clearance I used in my air-cooled radials. The deep finned heads and cylinders on those engines are efficiently cooled by the prop wash, and .003" is nearly optimum for an oil film and its protection against scuffing and piston noise. Assuming a temperature expansion of 13E-6 inches/inch/F, the piston temperature rises in the radials were most likely less than 275F. I don't expect the Merlin's heads, even with their liquid cooling, to be as well-behaved though. The prop wash won't be nearly as effective on the Merlin's heads, except for its influence on the radiator that will likely be required in the coolant loop. The full-size Merlins actually used drop-down radiators located beneath plane's airframe in order to prevent overheating during prolonged runway idling periods. For additional margin against piston expansion, the Quarter Scale's piston crowns were reduced to obtain a .008" (diametral) clearance. The piston's highest temperature rise will be at it's crown, and with only the rings providing the primary heat loss path, a little extra clearance is probably a good idea. This extra clearance extends .300" down from the top of the piston and includes both rings.

A second temperature consideration involves the clearances behind the rings. This clearance is determined by the cylinder bore, the radial thickness of the ring, and the ring's groove depth. The ring will be in full 360 degree contact with the cylinder wall, and because its diameter is constrained by the cylinder, only its running gap can decrease as the temperature rises. The piston's diameter is free to grow, however, and its clearance to the rear of the ring will decrease. Clearance must be maintained behind the rings, and especially behind the top one, since combustion gasses pressurize this space and force the ring against the cylinder wall to create ring's seal. I machined the piston groove depths to create a radial clearance of .006" behind the rings. Even at an unlikely 500F, a couple thousandths clearance would still remain behind the rings.

The (axial) widths specified for the Quarter Scale's stock ring grooves may have been a result of the scaling since they were spec'd at .040" for use with a .039" width piston ring. This width is much greater than the published recommendations for model IC engines by Trimble, Chaddock, and Walshaw. I've successfully used their recommendations on my other engines, and so I reduced the axial groove width to .027" for use with a .026" ring.

The construction of each piston began by turning its o.d.'s on the end of a 1-1/4" 6061 round before cutting the ring grooves. Particular care was taken in choosing the feeds and speeds (1500 rpm, 0.5 ipm for a carbide grooving insert on my 9x20 lathe) to achieve good surface finishes on the walls of the grooves. The lower wall of the upper ring is also a sealing surface for the upper ring during combustion. After cutting the grooves, the pistons were parted off in the bandsaw.

Unfortunately, I had once again recorded an incorrect dimension while setting up the process sheet that I created to machine the pistons. Before realizing that I had been turning them .002" undersize, I had machined all twelve. They're sitting in the 'scrap' pile in one of the photos. I gave myself a couple days off without pay after that one.

In order to machine the pistons' interiors, a fixture was created to support them under the mill's spindle. This fixture was designed to be turned 90 degrees so it could also support the drilling and reaming of the wrist pin holes. These holes must be precisely square to the pistons' axes to prevent connecting rod binds, and so a lot of care was put into the fixture's machining. It was finish-machined with a piston blank clamped in place after the two relief slots were cut. I probably spent as much time on that little fixture as I did on both sets of pistons up to that point. The wrist pin holes were drilled and reamed before the piston's interiors were machined, and the fixture was vertically indicated on the mill before each hole was drilled.

The final two operations included machining of shallow flats on each side of the pistons for the wrist pin retainers, and drilling the escape holes in the oil grooves. The full-size Merlins used c-clip retainers for the wrist pins, but the Quarter Scale will use hardened floating pins with soft tipped ends. Both operations were done on a rotary under the mill's spindle with the piston gripped in a machinable 5C collet. The collet was machined so it could grip the piston on either of its diameters since each operation required opposite end access to the piston. At the last minute, I increased the number of oil holes from six to ten. - Terry

 

[mayhugh1]

If you aren't familiar with ESC's and servo testers, there are a couple of things I'd recommend.
1 - I'd go with an aircraft ESC, not a car or boat ESC. That way you only get one direction out of the ESC, instead of a bi-directional output with full stop at mid-range.
2 - I'd replace the pot in the servo tester's input circuit with a switched fixed resistor, you're building the equivalent of a retract switch. That way the output of the ESC will either be on or off,

Don

Don,
Thanks alot for the tips. Sounds like good advice.
Terry

 

[mu38&Bg#]

I worked for a company which supplied Conley the starter motors. They were brushed cobalt motors, but also are out of production a few years ago now. The particular one in the video was marketed under a few labels. Before that, he used Astroflight cobalt motors, until they became too costly. Last I knew, he switched to "modified" (serviceable brush type) brush rc motors with gear reduction.

Unless you need a very small motor, I'd just as soon use a brush motor.

Greg