Source:
http://www.rcmodelreviews.com/coronafhssissues.shtml
As soon as I hooked up the Turnigy 9X transmitter module and receivers I purchased from HobbyCity (aka Hobby King) I knew there were problems.
As soon as I hooked up the Turnigy 9X transmitter module and receivers I purchased from HobbyCity (aka Hobby King) I knew there were problems.
Slow stick deflections produced jerky and jerky movement of the servos
I checked with several different types and brands of servos but the results remained unchanged.
On several occasions I also observed uncommanded movements of one or more servos.
Closer investigation pointed to an issue with the amount of processor power available to perform the various tasks required in such an application and the way the system was hopping.
To cut a long story short, I sent my observations and suspicions as to their cause to Corona and they confirmed them.
Corona has now had a couple of tries at producing a frequency-hopping spread-spectrum system and they've still got work to do.
The current version is close -- but still has some critical issues that really need to be remedied before I could recommend it for safe flying.
The Issues
1. Random/high Latency
A key cause of the erratic and jerky servo-movements associated with slow stick deflections is the very slow data-refresh rate between transmitter and receiver.
A key cause of the erratic and jerky servo-movements associated with slow stick deflections is the very slow data-refresh rate between transmitter and receiver.
In a normal FM radio, the servo position data is sent 50 times per second but in the case of the Corona, even though the transmitter still produces information at this rate, it appears only to be sent to the receiver at a lower rate. In effect, the system is dropping some of the data.
What causes this random/high latency?
It appears that the microcontroller (MCU) chosen by Corona to perform the various tasks required simply doesn't have enough steam to keep up with all those jobs. As a result, it processes only some of the data-frames.
What is the effect to your flying?
If you fly relatively slow sport-type planes then you may never notice the jerkiness and latency this problem creates. However, if you fly very fast models or airplanes with large surfaces/throws (such as used for 3D) then you'll almost certainly find the results undesirable.
2. Uncommanded servo movements
I've not discovered the exact cause of these but they may well also be linked to the overstressed MCU. Because there is only so much processing power available and so much of the work is time-critical, there's a very good chance that every now and then the processor is asked to do two things at once. Clearly it can't do two things at *exactly* the same time and this manifests itself as a slightly stretched servo-pulse or two -- resulting in an uncommanded servo movement.
I've not discovered the exact cause of these but they may well also be linked to the overstressed MCU. Because there is only so much processing power available and so much of the work is time-critical, there's a very good chance that every now and then the processor is asked to do two things at once. Clearly it can't do two things at *exactly* the same time and this manifests itself as a slightly stretched servo-pulse or two -- resulting in an uncommanded servo movement.
Corona themselves have confirmed the existence of the bug.
3. Failure to relink
When testing for the lowest voltage at which the receiver would continue to operate, I discovered that sometimes the receiver will simply "go to sleep" and not relink with the transmitter, even once full operating voltage was restored. This is clearly not a good thing.
When testing for the lowest voltage at which the receiver would continue to operate, I discovered that sometimes the receiver will simply "go to sleep" and not relink with the transmitter, even once full operating voltage was restored. This is clearly not a good thing.
If you are flying a model with a weak BEC or battery, it's quite possible that the receiver could enter this state when heavy servo loads are experienced.
4. Loss of RF link
It very much appears as if the Corona's FHSS receiver can be confused to the point of losing all contact with the transmitter if/when only a relatively small portion of the band is subjected to high noise levels.
It very much appears as if the Corona's FHSS receiver can be confused to the point of losing all contact with the transmitter if/when only a relatively small portion of the band is subjected to high noise levels.
Although there may be as much as 75% of the band unaffected, and although the transmitter continues to hop across that unaffected part of the band, the receiver is unable to extract valid data and all control is lost.
One of the key benefits of a frequency-hopping system is that it should be able to make use of even the smallest interference-free window within the band, the Corona can not. This may be attributed to the receiver processor being overloaded as it attempts to decode the noise as a valid signal -- thus leaving no time for hopping or other essential tasks.
The Bottom Line
People, including myself, have successfully flown the Corona/Turnigy FHSS system so you might ask "are these issues really important?"
The bottom line is that the system does work under average to ideal conditions.
If you fly slow/sport-type models in a noise-free environment with well-charged batteries and/or a decent BEC you may never encounter the effects of these issues.
It's worth remembering that we're talking about a module and receivers that cost only $20 or so. This is not a Futaba FASST system costing hundreds, it is a very bottom-end product with (in its current form) significant limitations. If you are prepared to accept that then you may find the Coronal/Turnigy FHSS system adequate and acceptable value for money.
However, the 2.4GHz band is becoming increasingly busy, not only with RC systems but with all manner of other consumer electronics. A park or flying-site that is relatively quiet today may become noisy enough to cause problems for the Corona/Turnigy FHSS system tomorrow.
Corona are therefore to be highly commended for their decision to withdraw this system from sale until such time as they have remedied the problems. They responded promptly to the issues I uncovered and forwarded to them and that's a responsible move on their part.
Likewise, HobbyCity's offer to refund (less postage) any modules/receivers returned is a reasonable one.
I certainly hope that Corona receiver do sort their issues, the hobby really does need as many good, reliable, robust and well-priced 2.4GHz systems as it can get.
沒有留言:
張貼留言