| Spirit System https://www.spirit-system.com/phpBB3/ |
|
| μSpirit on a Trex 150 - some observations and suggestions https://www.spirit-system.com/phpBB3/viewtopic.php?f=20&t=2148 |
Page 1 of 1 |
| Author: | horneteer [ Thu 10. Aug 2017 20:36:22 ] |
| Post subject: | μSpirit on a Trex 150 - some observations and suggestions |
Hello, I have a μSpirit on my Trex 150 and am having some tuning problems. Basically there seems to be too much cyclic gain even on the lowest setting. The problem manifests as a persistent nodding/vertical tail bounce that appears to originate from the head as opposed to the tail. I am using the standard mounting tape and have been through the setup process based on default Trex 150 settings. I have adjusted all cyclic parameters independently, including the 6 degree setting, to understand their effect on the flying characteristics of the Trex 150. I have tried a number of different balanced main blades and head speeds. All other hardware has been renewed twice with the exception of the motor, frame and servos. This should rule out bent shafts, bad bearings or worn dampeners. Increasing the elevator filter to 2 from the default setting of 1 caused the heli to violently shake which resulted in a crash. After rebuilding the machine, I went the other way with the elevator filter and found that a setting of 0 made the heli flyable, but still with a slight cyclic wobble that appears to be on the elevator axis but may be on the aileron axis as well. Reducing the cyclic gain down to minimum setting of 20 further reduced the wobble but it is still there. I should mention that I used to see the same rapid nodding when the gain was set too high on the original board. With the μSpirit, it seems that if an elevator filter setting of -1 or -2 was possible, the nodding would disappear and the cyclic performance would be perfect. Is extending the lower range of the elevator filter even possible? If the PID gains were adjustable, I would be inclined to leave the P as it is, but would reduce the 'I' value and maybe add some 'D' gain. It took some time to get the tail dialled in. The gain range for the tail also seems to be on the high side for this little heli. I am using an XP7A Fast ESC flashed with Blheli on the tail. For those that are interested I have tested different bells (6 and 12 magnet) and different tail rotors (different shapes and sizes from 42mm to 65mm - all new and statically balanced). The optimum setup seems to be the original Trex 150 tail motor (now sold by Oversky) and Align's new '23' tail rotor. Tail gain is about 15. Any higher induces wag but the heli is still flyable up to a tail gain of around 65. The tail performance is acceptable. I haven't flown it hard yet but from gentle flying it looks like tail performance is about the same as it was with the original board. It is nowhere near as locked in as a well set up double brushless mcpX though. I would like to know if it might be possible to improve the cyclic performance of the Trex 150. On the other hand the current firmware may have already been extensively tested on Trex 150 models and this problem may not have been seen before. If the former, I'd be happy to test a beta firmware if that would help. If it is likely that the nodding is caused by a problem with the model I will replace the remaining parts (servos, motor and frame). As an aside, it would be good if throttle hold would shut off the tail motor after a few seconds. All the best Al |
|
| Author: | ZeXx86 [ Thu 10. Aug 2017 20:39:54 ] |
| Post subject: | Re: μSpirit on a Trex 150 - some observations and suggestion |
Hello, it looks like Geometry 6° is not configured correctly (value is way too high). I recommend to start with simple test: 1. Enable Subtrim (tuning) 2. fold blades together and check if the ends are in the same level (this is necessary) Can you verify this? Next issue could be, that the unit is not holding tightly (it must be as tight as possible - no wobbles). We were testing with T-Rex 150 with no issues, so I believe we will solve it easily. Nodding is also usually caused if the main shaft is bent or main blades are too tight in the blade holders - you can verify this too. But I think that it will be rather settings problem since no oscillations should occur. Tail performance can be greatly improved with Revomix. You can share your settings so I can check it. |
|
| Author: | horneteer [ Thu 10. Aug 2017 21:15:09 ] |
| Post subject: | Re: μSpirit on a Trex 150 - some observations and suggestion |
Hello, Many thanks for the very quick response. It is possible that the 6 degree setting is not perfect as it was done with a visual pitch gauge rather than a micro digital one which I don't have. In troubleshooting this I tried going up and down for the 6 degree setting. My setting is 93 and I tried 88 to 98 but didn't see any difference in the nodding. Blades touch at the tips when folded back - this is how I set zero pitch. The μSpirit is stuck down firmly with the stock tape. It is not in contact with the frame at the edge. There is however a small tail ESC (XP7A) taped to one side and a small orange sat taped to the other, both attached with the same stock tape. While not ideal, the wires have some slack and vibration levels are extremely low. I doubt this is the problem but would appreciate your thoughts. I have tried different blade tightness levels as well as many different new main shafts and two new heads. I was suspicious of the standard main shaft because it is made of aluminium and perhaps the new ones I had were slightly bent at the factory. Before asking for support I wanted to rule this out by trying the new Align '150 Main Rotor Holder Upgrade Set H15H013XXT'. This includes a steel main shaft and longer control arms for better resolution: https://www.align-trex.co.uk/150-main-r ... 13xxt.html This has helped slightly but the heli still nods. Settings are as follows: Geometry 6 deg: 93 Elevator filter: 0 Cyclic feed forward: 5 Rudder delay: 0 Rudder - Revomix: 0 Rudder dynamic: 7 Pirouette consistency: 167 Let me know if you need any further info. Thanks again, Al |
|
| Author: | ZeXx86 [ Thu 10. Aug 2017 22:43:35 ] |
| Post subject: | Re: μSpirit on a Trex 150 - some observations and suggestion |
Perfect, please make sure that the zero degree setup apply also when Subtrim (tuning) is enabled. Some pilots are doing it so that they set mid stick, but this is not a proper way. (can be affected by transmitter configuration). While Subtrim (tuning) will always move the servos exactly to the center. If you will find it correct, you can try just to decrease the Geometry 6° value to around 65 - 70. I believe it will help. What is frequency of the cyclic servos? This can make a big difference too. Pirouette Consistency value is probably too high. I recommend to decrease it to 150 and instead try to increase Gyro Gain. Then increase Revomix to 3. I believe it will make the rudder more solid. |
|
| Author: | horneteer [ Fri 11. Aug 2017 15:37:27 ] |
| Post subject: | Re: μSpirit on a Trex 150 - some observations and suggestion |
Ok I can confirm the zero degree setup applies when Subtrim (tuning) is enabled. I have reduced Geometry 6° value to 70 and this does help. The heli is flyable with the cyclic gain set to 30. The frequency of the Trex 150 servos does not appear to be specified anywhere so I have set it to 200Hz. 333Hz works and the servos don't get hot but thought this might be unnecessarily high. Pirouette Consistency value has been decreased to 150 and Gyro Gain increased to 25. Revomix set to 3. Rudder is ok but wags a lot in a light breeze of about 5mph. I can probably further improve cyclic performance by finding the optimum balance of Geometry 6° value and cyclic gain in the software. I presume that the Geometry 6° value is effectively a cyclic gain multiplier. In which case should I set this to the minimum of 65 and try to the maximise cyclic gain dial? Or should I minimise the cyclic gain dial to 20 and try to increase Geometry 6° value instead? |
|
| Author: | ZeXx86 [ Sat 12. Aug 2017 19:34:14 ] |
| Post subject: | Re: μSpirit on a Trex 150 - some observations and suggestion |
The goal is to increase Cyclic Gain and Gyro Gain as high as possible (with no oscillations). With higher values the performance should be good. You can also try to decrease the Pirouette Consistency even more (to the lowest value) - and increase Gyro Gain again. Playing with Rudder Delay could help a lot too (to prevent from oscillations even with high gain), but this is very dependent on rudder response (speed). There we have to do more tests to give you precise values. Geometry 6° is working as a multiplier, but not just for Cyclic Gain. Cyclic Gain is only one part. With higher value it should be more stable, but some oscillations/nodding could occur. It is ideal that the unit is mounted as hard as rock to the frame of the model. So the unit itself can't flex and cause possible oscillations itself. Have a nice day! |
|
| Page 1 of 1 | All times are UTC + 1 hour |
| Powered by phpBB® Forum Software © phpBB Group http://www.phpbb.com/ |
|