What was your switch printed with? I use PETG and find it has pretty good layer adhesion. Also, are you using my latest file which has 3 tabs like this:
Yeah, that is the version I have. I have no idea what it was printed with. A friend of mine printed it for me for free, so I can’t really complain.
This is looking awesome! Keep it going!
I’m using the smaller version of the APDs on my custom OpenPPG, they’re very nice ESCs. Best is all the telemetry you get off them.
My reversed throttle is fixed. That Micro Maestro servo controller board I mentioned earlier is definitely an awesome solution for a DIY throttle control. It has six inputs/outputs that can be set to analog inputs, digital inputs, or PWM outputs to drive a servo or ESC. It took me about 15 minutes to write and tweak a script to use the analog input from a hall-effect thumb throttle (something like 1.5-5 VDC) to drive a PWM output. It would be easy to include a safe/arm switch or other safety devices to lock out the PWM output. On top of all that, it’s not much bigger than a quarter, takes a pretty wide range of input voltages, and includes a small regulated 5VDC output so you could probably run the whole system without an extra BEC.
I also finished up the power wiring for the second battery bank. All that is left before I can completely power this thing up is to balance the prop!
1 Like
I wired my “BMS” today. It’s just six of those little lipo tester/low voltage screamers, three on each battery bank. Trying to keep track of the cell voltages by looking at the screens is pretty useless, but I’m hoping that by putting the screamers ahead of me I’ll be able to hear them go off, and the displays flash when a battery gets below low voltage threshold. It is not an elegant solution, but at least it gives me a chance of catching a cell that is over-discharging. Right now they are set to a conservative 3.7VDC, but depending on how much voltage sag I see at full throttle I’ll probably lower that a bit. eCalc says I should be seeing around 3.3VDC per cell when the batteries are almost done.
I also balanced the prop. For something that is supposed to be “pre-balanced” I had to sand away a HUGE amount of wood from the hub to get it right. You can see how much I took off on the lower-right lobe of the hub. That was a bit disappointing, but it is done now and ready to mount. In fact, there really isn’t anything left to do before I can start power-on testing other than bolt the prop to the motor.
The Iron Thermal is spinning a prop under her own power! I did some testing today at 2/3 voltage (12S instead of the full 18S) just to verify the numbers coming from eCalc. What I found was interesting… motor temp was within 3 degrees (30.9 vs 28 predicted) after 6 minutes of run time (2 at 50% throttle, 2 at 75%, and 2 at 100%). RPM was 95% of predicted (~2750 measured to 2900 predicted), but the wattage was 70% of predicted (2600 measured, 3700 predicted). It could just be an error in the data logging, or it could be voltage loss over my 2+ meters of battery cables. The measured voltage while running at full throttle was 42 vs. my expected 43.
Either way I’m happy. The prop seems to be running smoothly, the motor temps are close to predicted, and everything else was staying super cool. My thrust angle looks good and the frame is staying level as the power increases. My next task is to drill and through-bolt all the couplings, and then I’ll be ready for full voltage (18S) testing.
I can’t believe this thing is coming together. It felt a bit like a pipe dream when I started this four months ago!
1 Like
You can easily determine the loss of voltage due to the cable. simply measure the voltage directly on the battery with the voltmeter. then directly at the entrance to the esc. simply use thin needles that you insert into the cable. if the dimension of the cable is selected correctly, the voltage drop for a 2 meter cable can be measured at a maximum of 0.1 volt (for copper). how did you log the data? directly from the esc and display or manually with a-meter and voltmeter?
The data is being logged by the ESC in a log file, which it output in graphical format. It is logging:
Input voltage
Bus current
Phase current
Power
ESC internal temperature
eRPM
Throttle input
Motor duty cycle
The log interval is variable, but I have it set to 25ms right now.
The reason I’m a bit skeptical of the power logging is that there are moments at lower throttle settings where the motor current jumps lower, with a corresponding increase in RPM. At one point, running at 50% the average current is slightly negative. Maybe it has something to do with active freewheeling? At any rate, it’s strange, but I’m not that surprised that a controller that’s capable of 300+ amps isn’t accurately logging current at ~14 amps.
the difference in speed will occur if you have programmed the erpm too low. that means you have a 12 pole motor with 6 pairs of magnets. as an example: if you want 5000 rpm on the propeller, the max. erpm on the esc to 6 times 5000 = 30000. i don’t know your esc but i guess that is because of the mistake. regarding freeweel. can you choose between freeweel and freeweel-no-syncro? if so try freeweel-no-syncro. the control is then more dynamic and does not react to the mass and acceleration of the porpeller. if you can set the run-up time take at least 2 seconds or more. (we fly eppg with mostly 2.5 sec run-up time) regards
When the logs come out of the controller, they are usually accurate to 0.1 A. we always log at 60 Hz / 1 sec. that’s sufficient and ideal if you combine the data in Excel with other CSVs, for example. from the vario or other devices.
I ran another test today, this time at 18S. I did 2 minutes at 50% throttle, and then 4 minutes at 75% throttle, checking temps at 2 minute intervals. After six minutes of running, the windings were at 45 C, and everything else was unremarkable. I couldn’t resist doing a brief full-throttle run to see how it felt. This thing has some power! Even though it’s hanging from a simulator for testing, it almost pushed me off my feet at 75%, and I had to brace myself better to get to 100%. Full throttle numbers (on low batteries): 3750 RPM, 140A, 8800W. That’s all in line with what eCalc was predicting on 18S with low charge.
Next up is to do full-throttle testing with temperature measurements every 30 seconds to check motor heat buildup.
yes that is already around 32 kg thrust (3.63 g / w). so you already have a climb of around 0.2-0.4 m / sec if the wing is good and big enough. as soon as you go flying always be careful with thermals. since you have a very high pitch speed, the reaction is very fast, which means that when you reach thermals and the wing brakes, you swing forward faster than with a standard paramotor with half the pitch speed.
I’ve been doing thermal testing to see how hot my motor is going to get with extended full-throttle runs. After 2.5 minutes of full throttle, the peak winding temperature was 93C, and the external rotor temperature is around 45C. Everything else (ESC internal, wiring juctions, cap pack, ect) is 30C or less. Running this system flat out is definitely pushing the limits of what the motor can handle. I had a discussion with Neumotor by email, and this is what they had to say on a temperature limit for the motor:
“The insulation is rated to 180C but as a general rule the magnets will be be damaged if they get over 150C. To be safe we prefer that the motor outer case stay below 100C as some internal parts will be higher.”
One of these motors has been used on a high powered ebike (@bratwurst linked to it above) with winding temps up to 130C, and that guy has been running for well over a year now with no issues, so I’m going to continue my testing to see how long I can run before the windings approach 130C. I’m also playing with the timing to see if I can improve the motor efficiency a bit to generate less waste heat, but so far that approach is inconclusive. I’m also more motivated to build an aluminum motor mount so there is more surface area to sink heat from the motor. I think that there is a chance I may end up running just cool enough to make it work, but it is going to be close. I’m already starting to think about larger motor options.
important to consider: if the windings can withstand 130 degrees and the magnets “supposedly also 150 degrees” then always remember that the temperature radiates as soon as the motor stops rotating. there are no magnets that like more than 100 degrees for a long time. the performance of the magnets decreases over time. Professional engine manufacturers have special markings on the inside of the engine to see whether the engine has been overloaded. Manufacturers of engines for aviation allow a maximum temperature of 90 degrees. the esc software for flying also specifies a limit of 90 degrees celsius. this has been introduced due to many years of experience and security awareness. neu motors has no experience with air sports, so it is understandable that the company has no thought that the pilot must also fly safely.
1 Like
for heating up time: to heat 2.5 kilograms of mass it takes a power x with factor Y. if this is achieved after usually 2-3 minutes, the current heat will increase the overall temperature. full load or near full load tests are therefore meaningful if you do them for at least 5 or more minutes, of course with constant monitoring. with thermocams this is very easily possible in addition to the internal sensor. if you go far beyond the motor’s limmit, thermal damage can also occur in a few seconds if the windings lose their insulation inside. I’ve already seen motors that were destroyed after 20 seconds even though the housing was only 50 degrees. the rotomax is a motor that is destroyed very quickly due to the thin wire from 10 kilowatts.
At this point, I’m accepting that I won’t be able to run this motor flat out out (with this prop) for an entire battery pack. My new goals are to minimize the motor heating as much as is practical, establish a maximum full-throttle run time that won’t make the motor dangerously hot, and then fly the system to see how it handles. At this point, I’m pretty sure that it will fly, but I’d like to get a feel for it in the air before I commit to buying a new motor. Moving up from this Neu 8057/75 is probably going to cost me $1000+, so I’d like to at least be confident that I like how this thing flies before I spend that much more on the project.
The other side effect of going to a larger motor is that I won’t be able to fly it with my pod harness and tail fairing, since it just makes the whole thing too long. The batteries will have to be cantilevered way out to balance the extra 1+ kg of motor weight on the tail. So moving up to a larger motor is also going to require some significant re-thinking on the frame design.
Hi,
you should try to swap the plywood motor mount to aluminium for better heat dissipation. And a small cooling fan can help a little .
Regards,
Roland
if a system is to be used for thermal flight, the motor must be directly on the body with its weight. the battery must be precisely in the center of gravity or it must be suitable on the scale. everything else is life-threatening and no one does this with flying experience. be careful. I won’t say more about it. it is only important to me that I have warned of this.
1 Like
The way my rig is set up, it is definitely not going to be used for real thermal flying. At most, it will be for launching into ridge lift or glass-off (restitution). I agree, there is way too much angular momentum with the motor in the back and the batteries in the front. @bratwurst I appreciate your warning. Realistically my goals for this project have changed from a self-launch system for thermal flying, to more of a traditional motor system that I can use with my existing wing and harness. It is going to have some serious limitations, but I do think it can fly (within those limitations) so I’m still going to give it a shot.
I did some testing with fully charged batteries today and I was able to run for 180 seconds before the windings got to 120C (measured with a thermocam). The maximum rotor temp was 66C, and everything else was negligable. Peak power numbers were 4000 RPM, 165A, 11500W, dropping to 3843 RPM, 160A, 10500W after 180 seconds. Those numbers are +/-5%, as my ESC log only outputs in graph format so it can be difficult to interpolate exact numbers. All those RPM numbers are with a partial prop stall, so I expect an unstalled prop to be +50 RPM or so, with a 5A drop in current. I plan to repeat the test at 15 and 25 degree timing to see how it affects the winding temperature.
Quick question for @bratwurst or any other data nerds: what % power drop do you normally see in static vs. in flight full power testing?
If the system is designed so that the pitch speed is approximately in the factor 1.4 to 1.6 to the tas / flightspeed, there is no change in the speed or in the power consumption. only the efficiency of the propeller increases a little in part. you can often hear that very well. you can also measure it if you do a thrust test in a laminar wind of 35 to 40 km / h and the test stand is positioned high. i used to do this with an electric stacker. Where you can see a change is when you fly from the spiral into a ground spiral with a full throttle and a small full reflex wing and reach a speed of over 60 km / h, the current drops by up to 15%. There is no change in the speed of the propeller because I work with fixed limmits, which means that when I set the throttle to exactly 2200 rpm, the motor turns it very precisely. in the professional area of eppg you set an rpm limit that is about 10 -15% below the maximum achievable speed. this compensates for the voltage drop and you have a pretty much the same power curve for almost the entire flight. Then there are some other important things for a highly efficient eppg that are relevant but do not fit the topic here.
1 Like