Glow to electric conversion: 201 Level Course

Tench745

Master member
If you were to repair the in the fiberglass cowling, I could see duplicating those blisters as a bit of a headache. Probably could make a female mould from part of the cowl, lay up a patch piece in the mould, then trim to fit the hole and glass in place.
If the ABS works long term it sounds like it'll be a lot less effort.
 

willsonman

Builder Extraordinare
Mentor
Well, my experience with smaller ABS parts has been very favorable. Even the cockpit interior pieces I've done have held up well inside their respective cockpits and functional parts have faired perfectly. The intercooler doors on my p-47 still perfectly slide like the day they were installed. I've just had no bad experiences with ABS prints other than bed adhesion and dealing with prints as they cool/shrink.

On that front, Before leaving for work this morning the print did finish. It's not perfect but It's certainly workable and I can smooth it out really easy. If it's the right size then we will be all set. I may have a look at that briefly this afternoon but my time on Thursdays is usually pretty limited.
 

willsonman

Builder Extraordinare
Mentor
UGH, It's been weeks but I've been crazy busy with work and home life. I was hoping to have this one possibly ready for an event coming up but not looking good. Too much going on and not enough build time available. Be on the lookout. I'll update here when I can.
 

bearit

Active member
Question on conversion, I just was given a Breitling silvershark that is mostly balsa build has a os 60 on it and i want to go electric.Wing span is 60" and has the fiberglass molded shark head, What do you suggest for a motor and esc?
 

willsonman

Builder Extraordinare
Mentor
Question on conversion, I just was given a Breitling silvershark that is mostly balsa build has a os 60 on it and i want to go electric.Wing span is 60" and has the fiberglass molded shark head, What do you suggest for a motor and esc?

You think you are asking a simple question that will yield a simple answer and I can give you one suggestion... or I could give you a dozen. In reality, I'll simply quote myself...

If you are curious about some more specifics on my approach to motor and prop selection, feel free to watch this video on my thoughts:
 

willsonman

Builder Extraordinare
Mentor
So here we are about a month later and I've finally made some progress. I was able to get BOTH sets of wiring completed. A pair of 21" long servo extensions were made for the throttle control. A good rule of thumb here is to remove the + (power) wire from the connection of ONE ESC. There are some folks who report instances of interference between ESCs but this was years ago before switching BECs were more common. Still, I've had no issues using just one BEC and frankly on more complicated ships like this one, you probably should be using a dedicated flight battery anyway. It's more safe to provide a dedicated system for the servos in the event that you lose one or both ESCs. As for my preference, I will be using a dedicated flight pack of 2S LiIon cells. The higher voltage will give me the higher torque rating for all of my servos on such a large airplane that I would like to keep that insurance policy.

So then we need to address the motor/ESC situation. No additional bullet connectors came with the motors I selected and nothing with the ESCs either, and I mean nothing. They came with pre-tinned wires for the motor and ESC and no connectors so time with the soldering iron is inevitable. The Motors came with 3mm bullet connectors and for some folks that's fine but since I'm in a situation where I'm trying to reduce and much constraint on the electrical system I opted to swap those out for 4mm bullets. The increase in contact surface area of the larger connection allows for higher current, but also a lower resistance for that current to flow. While it's probably not a requirement at this 65A current level, IMO it is a minimal weight penalty for some peace of mind. New bullets and some heat shrink was done there to complete that task.

Moving on to the actual power side of the system. Moving electrons from the battery to the actual ESC is a bit of a challenge here. The intention is to keep the ESCs in the nacelles and the batteries up front in the nose. As mentioned earlier I have opted for the longer battery wires rather than the longer motor wire option. This will save a little bit of weight but there's another factor here too. Voltage drop. The big reason we wdo not use DC voltage in our homes is the voltage drop over long distances. The most basic way to mitigate this is by increasing the diameter of the wire (gauge) to lessen the resistance of the wire. Again we are using Ohm's law wheel here. Now we can go online to a DC voltage drop calculator and the basic idea is that I'm extending the wires 2 feet. I'm using 12 gauge wire and the ESCs are rated for a constant max current of 65A on a 4S (14.8V) system. Using math we determine that over the length of this wire we are only losing about ONE HALF VOLT, or about 4%. Now, in my eyes as a scientist, that's EASILY within the margin of error for even measuring the load on the system. That amount can easily be "measured" just by changing from one prop to another... the same diameter and pitch. I'm really not going to sweat this amount of loss or the amount of power going through the wires. Of course I will do proper post-flight inspection to verify that these factors are all as my experience has suspected but 24" of length will not be a problem with this size wire on this system.

So, to extend the power wires it's a chore of cutting wire to length and soldering on some battery connectors as well as getting these extensions soldered to the ESC wires. I'm using heat shrink butt connectors and my soldering heat gun. I'm slowly rotating this connection until I see the solder flow into the connection all the way around the joint. I did NOT pre-tin the ends but left them clean and dry while pushing the strands of wire into each end before slipping the heat shrink connectors over the joint.

I know this post is long but there's a lot to cover. The last part is running the systems through the airplane. I got the ESC mounted inside the nacelle using industrial double-sided tape. The wires run through the wing and I had to remove the flap servo to do this and then push the wires to the side to replace the servo. All minor stuff and it just took time.

I hope that's clear. I don't have any pictures as there's not much to tell with them. Moving forward I need to address leaks in the pneumatic system and keep working on the cowl situation.
 

willsonman

Builder Extraordinare
Mentor
More work accomplished so let's start with some basics.

The elevator push rod sleeve (guide tube) was previously re-glued and the clevis was not replaced on the elevator before putting the stab back on so I removed the tail gunner canopy, removed the bolt holding the tail on and put the clevis back on.

The pneumatic system had a pretty massive leak in it and I determined that the Robart quick connects were the culprit. FOr those not familier with these systems, there are essentially two lines. One line pressurizes on side of a piston that extends the landing gear and the other line pressurizes the opposite side of the piston to retract the gear. There's a lot more to it but a smart installation will have one color on one line and another color on the other. This is smart for servicing. These systems require periodic maintenance and this really does help trace lines. There are after all THREE retracts on this airplane. There was a pair of the quick connects at the wing joint. This is to help the wings separate for easier transport. I ditched those for some more permanent barbed joiners. That got me a step in the right direction but I was still leaking a lot of air. The other set of connects was between the wings and the fuselage where the air tank and valve reside. The filler port as well as pressure indicator also reside in the fuselage but these disconnects were leaking badly. My go-to for these connections is luer lock fittings. Its the same kind of fitting that's on the end of a syringe and needles use this fitting to create a seal. You can get barbed fittings for tubing connections and these create a perfect seal that will withstand LOADS of pressure.

IMG_1533.jpg

I use a male/female on each side of the connection. I do this for two reasons. 1: when not in use, each end (wing or fuselage) can have the lines connected to each other to keep the system sealed and free of debris, moisture, and even small spiders and other insects. 2: when you go to connect the wing and fuselage there is NO QUESTION on whether or not you've got the right line hooked up. Think through this in advance and it just makes assembly at the field a LOT easier.

IMG_1532.jpg

After changing the disconnects for luer lock fittings I was able to get to 100psi (couldn't before) and I was only losing about 5 psi every 10-15 minutes as a worst-case estimate. My minimum operating pressure is 60 psi and I don't think I'll be flying for the better part of an hour.

Moving along, my second battery pack was due to arrive so I figured I should get the new battery bay prepped. I opened the hole with my razor saw and installed a simple 3/16" ply plate with some balsa bits on each end. Installation with wood glue. I use wood glue here because CA can be brittle and not really expand/contract with the wood here and the joint can fatigue over time. Wood glue is a better choice for this application.

IMG_1537.jpg

I'll again note that this is NOT the stock location for the batteries. They were designed to reside under the wing saddle and then you added lead ballast in the nose compartment to balance. I would have needed about 2 POUNDS of ballast to get this to balance at that point. This location is about 6" forward of that and should be just fine. I'll outfit the nose greenhouse with a frame and magnets to make battery ingress/egress a lot more convenient but it will likely initially fly naked.

I had removed the cockpit canopy to start making a tunnel to fish the battery leads through for every time I'd assemble the airplane but I discovered that there are passages just below the cockpit area that I can pass wires through... more on this that will be simple and more stealthy, leaving me room for possible cockpit details later on.
 

WillL84

Active member
So what creates the pressure? Is there an onboard pump or is it just a tank that you fill with a bicycle pump or something before hand? Never knew there were pneumatic retracts - that's neat!
 

willsonman

Builder Extraordinare
Mentor
So what creates the pressure? Is there an onboard pump or is it just a tank that you fill with a bicycle pump or something before hand? Never knew there were pneumatic retracts - that's neat!
GOOD QUESTION. The onboard tank is filled through a hand pump or an electric pump. Basically just like a bicycle tire but there's a specific fitting for the inlet. I have an electric one that I use that has an onboard gauge. It's fairly accurate.

Pneumatic retracts were the "step up" from mechanical retracts for a long time. Since everything is billet aluminum, they are quite expensive and require a lot of energy to operate so the pneumatic way was a viable solution. There are electric options in this size now that I would recommend over the pneumatic system but I will note that either system is about the same in terms of cost. There are cheaper ones of the electrics but I've been around long enough to know that on retracts you ALWAYS get what you pay for... unless you buy the Robart electric retracts... then you get ripped off. Ask @wilmracer

Larger models require these beefier options because of the weight of the model but also the transfer of energy. The side loads on these gear is quite tremendous so have the metal trunions on the retracts is a big deal. As you work with these things you'll learn more about them. I've got a video on rebuilding them that would be a good reference to get more familiar with these systems...
 

Tench745

Master member
I sometimes help refit boats during the summer. When we're running wires through a hull or up a mast we refer to various tables for acceptable voltage drops on various sizes of wire over various lengths. An important thing I learned when using these tables is that the length of a wire run is the total distance power has to run. Using your 24" extension as an example you're sending power an extra 2' to the load, but the power has to then come back to the battery the same 2' for a total wire run of 4'.
You were non-specific of the lengths you used in your calculations, so I just wanted to mention that for the collective.

Also, I really like those luer lock fittings. I'll have to file them away for the day I need small quick-connect pneumatic fittings.
 

willsonman

Builder Extraordinare
Mentor
An important thing I learned when using these tables is that the length of a wire run is the total distance power has to run. Using your 24" extension as an example you're sending power an extra 2' to the load, but the power has to then come back to the battery the same 2' for a total wire run of 4'.
You were non-specific of the lengths you used in your calculations, so I just wanted to mention that for the collective.
Interesting. In all my years I've not once heard this. :eek: Still, a drop of 1V is OK for me.
 

willsonman

Builder Extraordinare
Mentor
Moving along we have progress on the wiring. To make things MUCH simpler for installing the wing I decided to cut off my wire extensions right where they come out of the wing. This extension then got routed through the fuselage formers under the cockpit but next to the nose gear retract bay. It's a narrow spot that was a bit finicky to get the XT60 connectors through but I managed. The cut end I'm using 4mm bullet connectors that will easily let me pump more than 100A through. The wires come through at the wing saddle and will join up with the wiring in the wings.

Wing saddle and connection:
IMG_1539.jpg

Nose battery bay connection:
IMG_1540.jpg

Mock fitting of the nose bay cover and batteries:
IMG_1541.jpg

As you can see, there will be ample room for the batteries and their connection. The greenhouse cover will also have plenty of surface area for magnet action to go on. I need to add the bullets to the wing connections now and that should wrap things up. I should then be able to get the airplane fully assembled, powered up, calibrate the ESC throttle positions (you HAVE to do this on BOTH ESCs at the same time), program the throws, and check the CG. Due to the gear retract position I will check the CG with the gear UP, since the mass will be shifted aft in that position.
 

WillL84

Active member
Interesting. In all my years I've not once heard this. :eek: Still, a drop of 1V is OK for me.

I wire industrial machines and I've never heard of that. The voltage drop tables and calculators are calculated for the single run length and don't include the return trip. The load doesn't care about the return trip, it only cares about the voltage it's actually seeing.

What if the load (A light for instance) has a 30-foot power wire but is grounded where it's mounted? Do you ever calculate the resistance of the frame components for grounds like that?
 

willsonman

Builder Extraordinare
Mentor
Ok, so there's a list of things to update on. All small things but they add up.

The bullet connectors on the wing power wires were done. Pretty quick job. This allowed me to finally install the wing to the fuselage and start checking systems. First up was to AGAIN check the landing gear. I noticed that the centering spring for the nose gear had come out of place so I adjusted that. It's small but it keeps the nose gear from turning to the side once the pull-pull lines slacken during retraction, preventing the gear from binding. There's some additional binding that I need to investigate there too that I've not yet addressed but likely is just needing some lubricant.

Since the wing was on and everything plugged in it was time to also review the wing control surfaces. Ailerons were wor... no that's not right. One aileron was working and then it wasn't. Odd. Flipped the airplane over only to discover that one mount for the servo in the hatch had broken free. Folks, I preach it over and over. CA glues are for the IMPATIENT. They are brittle and just need constant inspection. A quick dab of epoxy and we are back rolling. While the epoxy cured, I looked into the flap movement. The flaps were not evenly level with the trailing edge of the wing so I set my basic programming in my radio and went to investigate. One side was really off my one tooth of the servo arm. To access the servos you have to remove the nacelle. It's just two bolts but it's clumsy with it all just dangling there by pneumatic tubing and power wires. The one side was then fine. To address the other side I approached it the same but the adjustment of the clevis was the order... but no... I went to adjust the clevis out and then the clevis came out. So, I adjusted the servo arm by one tooth and then started REALLY adjusting the clevis to get the trailing edge to match. The odd thing here is that this side had the two flaps uneven with each other, not just the wing. Sometimes other's make their own problems. All minor stuff but it adds up. Sound familiar? Aileron servo replaced in the wing and working fine.

So with the wing now sorted it's time to check the landing gear AGAIN... POP. Oh for crying out loud. One of the clevises from the pull-pull lines had popped off the servo arm because I'm a dummy and forgot to replace the fuel tubing over the clevis to secure it. The hatch pops off and I replace the clevis only to find a triangle of wood on the bench. Another popped piece of wood held on by CA glue. UGH. Wood glue this time and sand the surface before application.

While that dries, It's time to sync the ESCs. This is CRITICAL on any multi-engine airplane in the electric world as well as the fuel world. It's just a LOT easier with electrics and does not get effected by the weather. Transmitter on, power on the receiver, props off, throttle cut off, move throttle to full. Plug in BOTH batteries for the motors and let them sing. Next is to lower the throttle to the minimum and the ESCs know the range of the stick. That's it. I check the bare minimum of stick movement to see if the motors engage at the same time and they do.

With the motors synced, I put the original cowls on just for the maiden, they WILL be replaced. Props on and 3D printed spinners go on with the prop nut. Finally, I get to do a throttle test. A full run-up on the floor with storage charge packs gives GOBS of thrust. She will not have a hard time flying. CG check is now up with just a touch nose-heavy (1cm forward of the mark) with the gear down and bang on with the gear up.

IMG_1547.jpg IMG_1548.jpg

Last thing I noticed was with all this testing, I realized that my drawer find battery was really quite fatigued. More so than I thought. I have ordered some NiMh batteries and I'll make my own 5-cell pack since it's SO much cheaper to do. I need to finalize control surface throws and rates, throw in some expo for good measure and do some taxi testing. I taped the cockpit canopy on just for the maiden since it'll eventually get replaced. Yeah, I got that hatch re-installed already so that's off the list.

So, set throws and rates, address the minor nose gear binding, build a custom battery pack. That's really all that's left.
 

Bo123

Elite member
IMG_1057.jpg

Couldn't be bothered waiting and I chucked a 5055 motor in there. Yes this used to be glow. I need a way of cutting the old firewall out to make space for a battery. Any ideas? Ignore the covering it is better now.... and it was my first time.

All I need now is an 80A esc, 5000mah 6s Batt. I have a friend who is sending me some Spekky receivers (I have a TX but no RX) and other gear. Thanks Andy.
 

willsonman

Builder Extraordinare
Mentor
Cutting out an old firewall can be tricky stuff but my best recommendation is a razor saw. Usually about $20 or less but can make pretty quick work to cut along the side cheeks there. Use some triangle stock to reinforce the back side of the new firewall and then add some balsa to close up the fuselage from there.

More and more I am recommending the Gecko line of ESCs. I've had great results with them and they run very cool for their rating.
ZTW Gecko 85A ESC

More economically the ZTW Beatles line is also good but lacks some features and robustness from the Gecko Line but still really good ESCs.
ZTW Beatles 80A ESC
 

Bo123

Elite member
Cutting out an old firewall can be tricky stuff but my best recommendation is a razor saw. Usually about $20 or less but can make pretty quick work to cut along the side cheeks there. Use some triangle stock to reinforce the back side of the new firewall and then add some balsa to close up the fuselage from there.

More and more I am recommending the Gecko line of ESCs. I've had great results with them and they run very cool for their rating.
ZTW Gecko 85A ESC

More economically the ZTW Beatles line is also good but lacks some features and robustness from the Gecko Line but still really good ESCs.
ZTW Beatles 80A ESC
I’ve already been ofered an esc. Thanks for the advise however!
 

Tench745

Master member
I wire industrial machines and I've never heard of that. The voltage drop tables and calculators are calculated for the single run length and don't include the return trip. The load doesn't care about the return trip, it only cares about the voltage it's actually seeing.

What if the load (A light for instance) has a 30-foot power wire but is grounded where it's mounted? Do you ever calculate the resistance of the frame components for grounds like that?
I can't claim to be any kind of expert; I only know what I've been told, so my information is only as good as its sources.
In the marine wiring I've done we ignore the return circuit when using bus bars or chassis grounds because there is enough cross-sectional area in the path that resistance and thus voltage drop is negligible. But, when we're running wires 15ft from the battery to the mast and then another 30ft up the mast that adds up to some significant resistance if you under-sized your wire. Now that I think about it, we're typically more concerned about sizing the wire adequately for the the entire circuit, not figuring the voltage at a given load. Here's an example of the sort of table we would use for that.
Marine-Wire-1.jpg


Drawing from what I remember of my high-school physics class, if you want to calculate the voltage drop in a series circuit you factor in the resistances of the whole circuit. This would include the the wire on either side of the load. But I see what you're saying, that if you only care about the voltage drop at the load and not across the whole circuit, you would only need to figure in the resistances to that point. Out of curiosity, do you use the the positive or negative side of the circuit when figuring your wire run length? I ask because in most of my experience DC circuits are treated like the power comes from the positive side, but electrons are negatively charged and therefore flowing from the negative to the positive side. I'm curious what someone in the field has to say about it.

Willsonman, if you feel like I'm hijacking this thread feel free to give a slap on the wrist. Do you have an estimated maiden day yet?