Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

Battery Adapter, Typhoon H, DIY

PatR

Premium Pilot
Joined
May 1, 2016
Messages
10,800
Reaction score
6,316
Location
N. California
Due to the advice of the esteemed Mr. Carr this is being posted as a PDF for several reasons. First, it's long-real long. For those that want to download any of it a PDF makes it easier. For me it made it easier to post without having to work around post word counts and posts from people with questions before the primary info was finished posting. So, for those with interest, desire, and ability, is the first part of making a battery adapter for the Typhoon H. Unfortunately it focuses on a temporary adapter, which is the most difficult one to make and, IMO, not the one you really want to do. But it's an option you didn't have before.

This thread will be updated with performance data as it is developed. So have fun with a 13 page read, and think about things before doing them.

The amended PDF contains the results of the first flight test. Link to the amended PDF file in Dropbox: DIY Batt Adapter_2.pdf
 

Attachments

  • DIY Battery Adapter.pdf
    1.7 MB · Views: 396
  • Amended DIY Adapter_2.pdf
    1.7 MB · Views: 302
Last edited:
Due to the advice of the esteemed Mr. Carr this is being posted as a PDF for several reasons. First, it's long-real long. For those that want to download any of it a PDF makes it easier. For me it made it easier to post without having to work around post word counts and posts from people with questions before the primary info was finished posting. So, for those with interest, desire, and ability, is the first part of making a battery adapter for the Tyhoon H. Unfortunately it focuses on a temporary adapter, which is the most difficult one to make and, IMO, not the one you really want to do. But it's an option you didn't have before.

This thread will be updated with performance data as it is developed. So have fun with a 13 page read, and think about things before doing them.
Well done Pat. Lots of really good information.
 
Last edited:
  • Like
Reactions: KellyR
With the underwhelming response to this thread I was sorely tempted to let it lay and go fly, but in the interest of finishing what I started some data was captured from today's first flight of the new battery arrangement.

First and foremost, it works, and works just fine. Secondly, I'm calling B.S. to Yuneec's 5400mA labeling on their factory H battery Ain't no way, not possible, no how, and there's only one three letter word I have for their mA labeling practices. It's simply not possible for a 5400mA, low C rated battery to provide more flight time than a 6000mA, 35C rated battery, even allowing the 6000mA equipped aircraft weighed 70g more than stock. For the record, my usual flight time for a 6300mA factory shell battery runs between 13 and 15 minutes, landing at 14.4V. The first flight of the 6000mA battery, flying aggressively for half the flight and hovering the other half, provided 0:13:37 minutes to the first low V (14.3V) warning, landing at 0:14:09. Total power on time for the battery was 0:17:47

Since most are primarily concerned with "how much flight time?" I can say the first flight with the 6000mA battery was about average for what I normally obtain from a 6300mA battery. I'll be doing a flight with a 7000mA battery shortly just to see the difference, if any.

For those that like numbers, the following.

Aircraft weight for the flight was 2018 grams
Stock aircraft weight is ~1948 grams
Aircraft Test Weight Delta: +70 grams

Amount of battery extending aft of the battery well: 10mm
Power On Battery Voltage: 16.74V
Power On voltage drop: 16.74V down to 16.71V, or 0.03V
Power On Amps: 0.88
Motors On Amps: 3.06 initial spike, falling to 2.91A
Camera Recording Amps: 0.88 climbing to 0.91V, or 0.03V
Hover Amps: 19.5A variable to 20.8A
Voltage at Landing: 14.3V
Voltage Rebound, Motors Off: 14.68V
Gear Swing: ~1A Hover voltage at 19.5A, gear swing increases current to 20.69A
Peak current, entire flight: 44.64A
Peak Watts: 677.1

The aircraft was flown in Angle mode at "Rabbit" speed for the entire flight. GPS was turned off for a portion of the flight to obtain maximum speed and climb rates. Full power climbs combined with maximum forward flight and maximum roll maneuvering were performed to maximize current demand. Absolutely no negative flight performance was experienced in any part of the flight. Hover was stable. maneuvering was normal. Braking was crisp. Take off and landings were both what I would call normal, with zero issues landing in an 8mph cross wind. Landing was "full stop", upright on the ground, without any thought of using the arming button to assist. General weather was clear, dry, and ~82*F.

The test aircraft carried a Watts Up ammeter affixed to the back of the camera gimbal that weighs 71.5 grams. If we deduct for the Watts Up the test weight aircraft and a stock aircraft are within a gram or so of each other, ready to fly. The only performance question I have after the flight is how the test aircraft CG compares to the CG of a stock aircraft.

Fore and aft Center of gravity was checked by balancing the aircraft on balance beams perpendicular to the two arms on each side of the aircraft. Lateral CG was not checked. As someone mentioned in another thread, the Typhoon H has an aft CG when in stock form. As I have yet to check balance of a stock aircraft I do not know if the CG condition is the same or worse with the test aircraft. What I can say for certain is that it required 155 grams of ballast to move the CG forward enough to balance the aircraft on the side to side arms. We might want to note the aircraft body is roughly 195mm long and the side to side arms are positioned ~95mm aft of the front of the aircraft, ~100mm forward of the back of the aircraft. They are not at frame center.
Edit: boom location revised a couple posts later. Correct boom location is 103mm aft of the nose.

The next flight will use a 7000mA battery that weighs just a little more than the 6000mA battery but is 8 grams less in weight than a stock Yuneec "5400mA" battery, and does not extend past the end of the battery well at all. BTW, I'm calling their battery 6300mA from this point forward unless and until they prove to the contrary.

First flight impressions? It works just fine and I could not see or "feel" any difference between the test aircraft and a stock aircraft. I won't ever be buying another factory shell Typhoon H battery. After the 7000mA test flights I'll be changing out the temporary adapter for the permanent version. It's lighter, eliminates an intermediate pair of connectors, and opens up a few mm of space in the battery well. I'm considering obtaining a more forward gimbal mount to better offset the aft CG to see how it affects flight performance and time.
 
Last edited:
I'm considering obtaining a more forward gimbal mount to better offset the aft CG to see how it affects flight performance and time.
If you have a 3d printer you could go to Thingiverse to download and print the gimbal rebalance adapter I uploaded there. My screen name on Thingiverse is shockazulu. If you don't have a 3d printer I'll send you the part no charge. I like the way you collect data. I am most interested in if flight times improve with vs without the drone balanced. With the 1 inch adapter it brings the stock drone configuration into perfect balance and still fits into the back pack case. Yuneec shipped the H's tail heavy and I want to know if it even makes much of a difference if we have it balanced correctly.
 
Last edited:
  • Like
Reactions: Ziggy
Moving an aircraft’s CG aft will always have an impact, most often on flight stability. A given aircraft can have positive stability, neutral stability, or negative stability just by shifting CG. With multirotors an imbalance requires things to work harder and faster to maintain stability, and working harder consumes power. Multirotors are by design unstable but a central CG makes them less so.

I do not have a printer. I have the same question about effect of CG on flight and power system performance as you do.

When building an aircraft or balancing one I prefer not to add ballast to correct CG, and instead would relocate components to achieve a desired CG. Your camera mount would add some weight but I suspect that amount would be small and of little impact in and of itself. Moving the camera forward would have a much greater effect. When I checked CG for the test flight the ballast used to determine how much of an imbalance it had was not used for flight, the weight was removed prior to flight and the aircraft was flown with the aft CG.

We should remember my test aircraft will be carrying almost 3 ounces of additional weight in the form of an ammeter as long as data is being collected, which has some level of influence on the test data. However, I have it affixed as close to airframe center as possible to minimize effect on CG. I have yet to check CG without the meter but that will be done after data collection is complete. It’s effect on performance through total aircraft weight cannot be eliminated without eliminating the collection of performance data.

If things go as planned one of the tests will be establishing the effects of making the aircraft progressively more tail heavy. To do that will require affixing a given amount of weight to the back of the battery and extending weight progressively further aft to simulate longer and heavier batteries.

Just curious, what is the furthest distance your battery adapter can extend aft of the aircraft, in mm if you can, and how much did the largest capacity battery used weigh? Within 1/10g would be as accurate as my scale for consistency.
 
Last edited:
  • Like
Reactions: Andrew Grant
I needs make a dimension correction. The body is 195mm in length. The booms perpendicular to the body (side to side at 90*) are located 103mm aft of the nose, not 95mm. The error was from trying to “eyeball” a measurement instead of doing it correctly with vertical stops and reference marks.
 
The Tattu 7000mA battery testing did not go well. No fault of the adapter or additional weight, but the pair of Tattu 7000's I have are just over 4 years old and have not been used for a couple years. We might say there was a significant maintenance error with failure to periodically cycle batteries between a charged and storage state.

Take off performance was excellent, initial hover was totally stable. While descending to eye level and hovering to obtain data from the Watts Up meter the battery voltage was observed to be descending rapidly. In the span of only a second or two battery voltage fell from ~16.2V through 15.2V and the aircraft started descending. During that period of time, while applying full power throttle, the aircraft increased the rate of descent and clipped the edge of the plastic 55 gal barrel (breaking prop 1) it had launched from, causing it to flip inverted, impacting the ground and rolling down a 4' slope with the motors turning, breaking 3 additional props. Had it missed the barrel it would have just landed hard, but it didn't. Only damage was the props so it could have been a lot worse. Blades were flying everywhere and looked like someone was tossing a prop salad.

Data collected from the doomed flight was minimal, but did provide some useful info.

Aircraft weight: 2100.6 grams
Initial battery voltage: 16.67V
Post flight battery voltage 16.24V
Initial cell voltages displayed less than 1/10V delta with highest cell at 4.19V
Post flight cell voltages: 4.065, 4.105, 4.124, 4.129
Initial IR: 5-5-3-3
Post flight IR: 5-5-3-4
Total power on time: 0:08:12
Total flight time: 0:01:46
Peak current: 43.63A
Peak Watts: 673.6W

Post crash analysis suggests one battery cell failed under load, creating a voltage imbalance the remaining cells could not compensate for. In essence the battery divided power delivery between the aircraft and attempting to balance the failing cell, which decreased power supply to the aircraft below what was necessary to maintain flight. This is something I've experienced before when one or more cells falls more than 1/10V below the other cells in a pack. When you're observant, and lucky, the voltage drop allows enough for you to land before voltage can no longer meet the demand. If you're not observant a crash is assured. In this case I was very observant, and almost close enough to the aircraft to reach out and grab it, but not quite close enough. Such is the way it goes in flight testing.

I have a second 7000mA battery but it is of the same age as the failed battery, so I'm a bit leery of giving that one a go. Perhaps a full charge/discharge cycle on the charger to check cell state prior to flight ops will tell me what I need to know.
 
Last edited:
Great info, @PatR.

Please do not take a lack of replies as a sign for lack of interest. I believe many are following along without feeling the need to clutter your progress reports with jabber.

Looking forward to more tests and tests flights.

By the way, just wondering if motor age is being taken into consideration with these flight times testing as well as reports of diminishing flight times we keep seeing? Was it @AH-1G who replaced aged motors and found significant increase in flight durations?

Hopefully my last comments will be more for rhetorical thought rather than fodder for clutter in this thread. Apologies in advance if it happens.

Will be reading!

Jeff
 
  • Like
Reactions: Mike Irish
Motor age has not been a consideration thus far as the flight time averages are pretty much the same today as they were when new three years ago. As I fly several different aircraft my H’s don’t get flown as much as some do. In fact, I’ve only put a couple flights on each of them since November ‘17. This one flew twice in that time before these tests. Prior to starting the tests it got checked out pretty good and the motors are still smooth with no grit or chatter.

Tomorrow gets a battery cycle to see what develops. If the battery fails I’ve still got the 6000 to work with and there are some tests with weight and fore/aft CG it can be helpful with. It would be fun to establish the relative effects of weight, time, and higher capacity batteries but if the 7000 fails the higher capacity battery thought won’t come to pass. That would be a shame as it fits so well.

Worst case, we find out what we can get away with. Best case we determine if a forward CG shift improves flight performance. I now have a theory about how CG might be the cause of that “twitchy” response coming off a hover. We’ll see.

In the mean time we’ve established Yuneec or other manufacturers can’t hold us hostage with a proprietary shell battery. Good for us H owners but I have a hunch the Plus and 520 owners are going to be looking for alternatives soon. Their batteries seem to cost almost twice as much and last half as long.

Strange as it might sound, but despite today’s crash this kind of stuff is smack in the middle of my comfort zone. I got seriously burned out with telemetry analysis, design reviews, and flight testing a few years ago and it feels good to get back in the saddle. It’s work, but it’s fun work and having some fun is what it’s all about.
 
If things go as planned one of the tests will be establishing the effects of making the aircraft progressively more tail heavy. To do that will require affixing a given amount of weight to the back of the battery and extending weight progressively further aft to simulate longer and heavier batteries.
The major issue with shifting the weight further aft is the possibility of, should you go full power forward and suddenly change to full power reverse, it will flip backwards. And if close to the ground it will not have enough time to regain balance and it will cause a crash.
This off balance issue is how I got started with the Typhoon H.
I was making battery adapters for 2 other drones and was requested to make one for the Typhoon H.
The person who requested the adapter for his H had used one of the earlier adapter versions with a large 8000mah battery. The adapter he used was a shell where the battery fit inside and you slid the hole thing in like a regular shell battery. His issue was those adapter shells used large amounts of space in front of the battery for the wiring and connectors, causing more of the battery to need to hang out the back. He had just experienced the type of crash I described above.
This is where to help him in his quest, I purchased my first Typhoon H so i could see if i could design something better than the adapter he was using. The answer to the problem that I found was routing the wire down the side and out the back so that now the batteries can fit all the way into the drone giving it the best chance for balance, with the only drawback now being that we will be inserting the battery backwards and making the connection on the outside.
Just curious, what is the furthest distance your battery adapter can extend aft of the aircraft, in mm if you can, and how much did the largest capacity battery used weigh? Within 1/10g would be as accurate as my scale for consistency.
So far I have only located 3 Production LiPo batteries that are 6300mah or higher and also fit into the battery hole. All 3 of those batteries are the same length and able to lock in using the thumb release outer locking mechanism on the adapter. For a exact measurement the top edge of the batteries lines up exactly with the top edge or the battery door opening. So if you measured straight down from the top edge of the opening that would be the distance that the bottom corner of the battery sticks out. The adapter is designed that a longer battery would fit, only the lock to hold the battery in will not be able to actually lock. Another method to prevent the battery from sliding out would be needed for longer batteries.
 
Last edited:
Great info, @PatR.

Please do not take a lack of replies as a sign for lack of interest. I believe many are following along without feeling the need to clutter your progress reports with jabber.

Looking forward to more tests and tests flights.

By the way, just wondering if motor age is being taken into consideration with these flight times testing as well as reports of diminishing flight times we keep seeing? Was it @AH-1G who replaced aged motors and found significant increase in flight durations?

Hopefully my last comments will be more for rhetorical thought rather than fodder for clutter in this thread. Apologies in advance if it happens.

Will be reading!

Jeff
I notice some of my motors sounding a bit odd after gezzzzz 75 hours of flight time.
Also notice several of the motors didn'tt feel the same while hand rotating, so I decided to purchase all new motors.
I can't really say "significant" flight time, what was significant, landing at 14.5 V rather than 14.7.:p
I would assume some of my motors were heating up "working harder", thus the more energy was being consumed, but how much?
Also notice a bit less vibration while still using the same blades.
 
  • Like
Reactions: NorWiscPilot
I do not have a printer. I have the same question about effect of CG on flight and power system performance as you do.

Your camera mount would add some weight but I suspect that amount would be small and of little impact in and of itself. Moving the camera forward would have a much greater effect.

PatR let me know if you want it and I will send you a 3D printed camera balance adapter.
 
Having one would certainly benefit CG experimentation. I’ll send a PM with shipping and contact info. Thanks.
 
Thank you Steve.

Electing to work with the second 7000mA battery this morning led me to a couple different ways to check the battery without flying it. The first was to charge the battery and then perform a discharge cycle on the battery. Although the charger's maximum discharge rate of 5A was selected the charger would not discharge at more than a 0.6A rate. Not enough load at that rate to determine cell health.

Post charge cell states indicated balanced voltage at 4.19V/cell with IR's of 2/3/2/3, which is a minor increase over the IR's of 2/2/2/2 that were obtained after digging the battery out of the storage case. This was concerning as the failed battery also generated an increase in IR in two cells after charging.

The second method entailed clamping the H to a work bench and performing an extended full power run. If you haven't done that before you might become startled when they power up. They act a bit "angry" when their feet are held on the floor. But it does provide a means to assure everything is working properly after a crash without risking the aircraft in flight. The downside of this type of a test is the props don't get to "unload" and generate full thrust. They encounter "blade stall" by remaining anchored in a stationary air mass so full power thrust is never achieved. Making that worse is the fact a multirotor prop is drawing air through the prop on one side while it is being reflected back into the prop from the other side if the aircraft can't leave the ground. Think of two garden hoses being aimed at each other with a rotating fan in the middle.

So after letting the H sit for a few minutes to self calibrate and running the motors at various speeds, mostly full throttle, for several minutes while observing the ammeter there were no indications of rapidly decreasing voltage. The only unusual aspect of the test was the H did not throttle down as much as usual when holding the stick full aft for an extended period of time.

Starting with a battery voltage of 16.78V the H was run for a total of about 7 minutes, with at least 4 of those minutes at full throttle. Ending battery voltage was ~15.2V, which rebounded to 15.97V after the battery cooled down. A post run IR and cell voltage check was performed and IR had returned to the earlier levels of 2/2/2/2, and cell voltages were dead even at 3.99V. The only delta was in main voltages at 3.996V and 3.988V, which at 0.008V is minimal. It looks like this battery will get flown.

For anyone interested, the power values obtained from the bench run maxed at;
52.48A
816.0W

C/G
As the 7000mA battery is heavier than the 6000mA, a quick check was performed that established the C/G became a little more rearward with the 7000mA battery. Using a set of scale calibration weights established it would require 160 grams of ballast be added to the front of the airframe to offset the heavier battery, but if the ballast was added to the forward booms just aft of the motors, balance could be corrected with only 50 grams of ballast. As always, moment arm is everything.

I rigged up a crossbeam between the two forward motors in order to affix a 50 gram weight for flight testing If the battery flight test goes well. The crossbeam consists of a 1/16" carbon fiber rod zip tied across the forward motor booms. The entire contraption, less ballast weight, does not trigger a response on my 20,000g scale, which is accurate to 1/10 gram.
 
7000mA Tattu Battery #2

Despite being a little intimidated from the previous crash, the battery ground test performance indicated it should be safe for flight testing. Despite ground testing OK my gut feelings suggest the flight test was flawed for the following reasons; aircraft yaw response in one direction was extremely slow, some minor "toilet bowling" occurred during the second half of the flight and this aircraft has never in three years toilet bowled before, the first landing warning occurred at 0:12:44 minutes, earlier than with the 6000mA battery. Total battery amp hours were only 4.459, which is only 63.7% of rated capacity. Despite the previous the majority of the flight went just fine, with good take off power and great hover stability during the first half of the flight. The aircraft was not "twitchy" as I thought it might be with a more aft CG, but then again, the toilet bowling might well have been induced by the CG. IMHO, a single post crash flight using a 4 year old battery is not enough to form an objective opinion. Testing needs to be done with a new battery after a full aircraft system check up.

Regardless, the data collected from the flight is as follows;

Aircraft weight 2098.6 grams
Power on voltage: 16.78V
Power up voltage sag: 16.76V
Power up Amps: 0.43A
Average power on (motors off) current: 0.84A
Motor on Voltage: 16.68V
Motor on peak Amps: 4.73A
Hover Amps: 19.55A
Landing Voltage: 14.3V
Motor shut down voltage rebound: 15.02V
Total power on time: missed data point, roughly 14-15 minutes
Total motor on time: 0:12:44
Maximum Amps: 52.15A
Maximum Watts: 791.1W

An interesting observation was when the system signalled it was ready to fly with the double beep. Current increased from 0.84A to 1.27A during the tones.
 
7000mA Battery #2, Post Flight Analysis

After experiencing what i feel were disappointing and false results from the 7000mA battery test flight, further investigation was performed to determine what might have influenced battery performance.

The battery was permitted to sit overnight to cool down and recover from discharge loads. The pack voltage had increased overnight from end of flight, power off voltage of 15.02V to 15.16V. Having landed at 14.3V there was no surprise battery voltage increased after motor and FC loads were removed from the battery, but an increase of more than 1/2V was a bit of a surprise. Also common was the post cooling increase. However, observing that cell IR's had started at 2-2-2-2 and fluctuated from 2-3-2-3 for the ground run test, dropped to 2-2-1-2 for the flight test, and increased to 3-3-3-3 afterwards was troubling. Prior to the flight cell IR's had been 2-2-1-2. Also troubling was post flight voltage balance between the cells. After the flight and rest period the cells measured 3.781V, 3.792V, 3.791V, and 3.796V. Although only a 0.015V delta between the high and low,it was surprising to find the one cell in the lower end of the 3.78V range instead of low to medium 3.79V.

Ultimately, I've come to the conclusion the 4 year old 7,000mA batteries were not in the best of health and performed as well as their condition permitted, which was not as good as they could have been. Unfortunately I wasted some time and a set of props in the process.

When I started out on this path it was not my intent to spend a bunch of money on batteries as I didn't need any. I've altered that position a little($100.00 little) and just bought another Tattu 7,000mA, 25C battery since GetFPV sent a note they have them back in stock. Very limited stock as the three they had has been reduced to two. Perhaps of interest, but Amazon has some Tattu 6,700mA packs but they are being sold for $4.00 more than the 7,000's at GetFPV, but Amazon has free Prime shipping... what’s 300mA, right?

Had I not started playing around with a 7,000mA battery, sticking with the 6,000mA would have been no big deal, but having involved the 7,000 I just can't walk away from a couple flawed flight tests without learning how much better they might have gone. I've just gotta know...
 
Last edited:
After a brief wait the new 7000mA battery arrived. Shipped voltage was 15.24v with IR’s of 3-3-3-3. Post charge cooled voltage was 16.8V with IR’s of 2-2-2-1. Cell balance is 4.18V for 3 cells and 4.179 for 1 cell.

Dimensionally it’s the same length as the 4 year old 7000 packs but the new one is 2-3mm less in width and height. It’s also 16mm shorter than the battery compartment portion of a stock battery.

Surprisingly, the new 7000mA has significant weight advantages over the stock, 6000 and old 7000mA batteries. I left the numbers in the shop but I’ll post them tomorrow.

Also received was the 3D printed gimbal location adapter from Shockazulu. I’ll be doing some flying with both the 6000 and 7000mA batteries to condition them before moving to flight testing with the gimbal adapter. Lots to do so time to get back to work. Just as soon as the grandkids go home as 7 year old twins leave little time for anything but them.
 

New Posts

Members online

Forum statistics

Threads
20,952
Messages
241,577
Members
27,284
Latest member
csandoval