View Single Post
Old 09-10-2017, 02:54 AM   #71 (permalink)
Registered Users
Posts: 2,028

Join Date: Jan 2015

Originally Posted by kevin75 View Post
Please, if it's possible, insert a Delay time on voltage allarm, because it's works but too much fast.
Example: 12 cells, good final end voltage could be 3,7x12=44,4 V
If i put 44,4 voltage alarm on Graupner radio, it starts to give too much often false alarm because when you make 3D it's normal that for very small time voltage drops under it. So if it will be possible to delay the voltage alarm, for example i can put 2 second of delay , so i cut all false alarm that now appear.
The purpose of alarms is not to report normal conditions but to report critical conditions before they become dangerous.
The purpose of Voltage alarm is to avoid that lipo voltage go below 3,0V because if the lipo voltage go below 3,0V, lipo is permanently damaged.
Voltage alarms for lipo must be set to 3,1 / 3,2 / 3,3V.

Why you set an alarm at the "nominal" lipo voltage (3,7V)?
3,7V is the normal and nominal voltage WITHOUT ANY LOAD.
Is possible that at the end of a flight you can read this value, but on a discharged Lipo you can read this value only without any load applied (no motor, no servos, no receiver, no flybarless unit).
Lipo voltage cannot be used to measure the flight time. To measure the flight time the only valid parameter is the “Used mAh” value. Otherwise your lipos goes below 3,0V threshold.

During flight under the high load of motor is normal that, also due to the IR (Lipo internal resistance) and also ER (cables/connectors external resistances), the voltage drops.
The discharge curve of a lipo can change by brand, by model, by nominal values, by "C" rating, etc. but are all similar.
A typical lipo discharge curve with a fixed load is like that I already put in this post:
As you can see, the ∆V/∆t between 10% of discharge and 90% of discharge is very very small, but the load variation due to the Pitch, Cyclic and tail variations during a flight generate a very higher voltage variation compared to the normal voltage discharge. So, the voltage “noisy” due to the load variation is too high compared to the nominal “signal” variations. Signal/noisy become a value less than one.
Only at the last 10% of power in the lipo the ∆V/∆t become high. But at this time can be too late to have the time to land because if the lipo voltage go below 3,0V the lipo is permanently damaged, and this can easily happen with the variable load of the motor.

I suggest you check the voltage trend of your Lipo during flight with Brain2 Logs function. You can go in DIAGNOSTIC -> Parameters setting and set 50Hz for Sample Rate and assign the “ESC Battery Voltage” parameter to any one of your 10 Log channels. At the end of your flight, after cycling power you can download the log and look at MIN, MAX, AVG values.
With these info's you can easily understand if the nominal value of your lipo is too low, if the "C" rate of your lipo is too low, if the wires must be shortening, if connector must changed with higher ampere rating connectors.

Remember also that the percentage you want left in to the lipo at the land time is correlated to the Max Burst Discharge "C" value.
EG: if you have a 5000mA lipo with a Max Burst Discharge C value of 120, if you land at 20% this mean that 1000mA (is 5000mA*20%) * 120 (Max Burst Discharge C value) = 120A.
This is the max allowable power peak needed by your power train at the land time.
Otherwise your lipos can easily go below the 3,0V limit threshold (damages and "puff").
So, better if you check also in your logs "Esc Battery Current" to have an idea of current peak values immediately before the land.

Last edited by BrainDev; 09-10-2017 at 03:09 AM..
BrainDev is online now        Reply With Quote