Looking forward to the 7 inch specs and a BUILD log here...plus videos     

(08-Apr-2021, 05:48 PM)iFly4rotors Wrote: Looking forward to the 7 inch specs and a BUILD log here...plus videos     

There you go :-)

https://intofpv.com/t-sub250g-auw-7-dji-...-endurance

Hi Felias,

I have posted a reply on the sub250g 7 inch build thread.

Bit of a necro, but this is clearly not a universal cutoff. If the 3.5v cutoff was real, none of our 1S whoops would fly for more than a minute.

I have an unmodified CrazyBeeX in my NanoLR. It flies down to 2.4v, at which point video cuts out and i need to land blind. I guess you are running into the issue identified by others as the voltage dropout caused by the voltage regulators in the ESC. I wasnt even sure that ESCs had access to the voltage that they are switching

(03-Jun-2021, 03:13 AM)Tokyo_Dom Wrote: Bit of a necro, but this is clearly not a universal cutoff. If the 3.5v cutoff was real, none of our 1S whoops would fly for more than a minute.

I have an unmodified CrazyBeeX in my NanoLR. It flies down to 2.4v, at which point video cuts out and i need to land blind. I guess you are running into the issue identified by others as the voltage dropout caused by the voltage regulators in the ESC. I wasnt even sure that ESCs had access to the voltage that they are switching

Hi Tokyo Dom,

The boards are Not all the same. I have some 1S boards that will drain below 3.5 volts and the some that don't.  Not all boards are created equal. A few of the 2S + boards don't have this regulator, but most do. It all depends upon how the manufacturer implemented the board circuitry. For instance, some of the Flywoo boards have the ability to run the battery down farther than 3.5 volts.  On the other hand, many of the boards will NOT drain below 3.5 volts and DO, in fact, have a Low Voltage Cutoff regulator. 

Plus, there is software configuration controls in BLHeli which provide some control. 

This is all been proven by many others. Much research has been done and Much experimentation has been performed to determine where and why the cutoff was happening. Up front, there were many theories, however, in the end it seems that some limiting is done in software, but most is done by a regulator on the FC that feeds the ESC circuitry. Simply {I say that with a grain of salt} bypassing this regulator allows the batteries to be drained further. 

Specifically, the HIFIONRC F7 AIO boards DO have a Low Voltage Cutoff regulator that cuts power when the per cell voltage is at or below 3.5 volts, however, there is a hardware hack to bypass this regulator which allows draining the batteries lower; link below:

[ESC] 2s li-ion discharge trick

I think there are two separate issues being confused here. One is the low voltage cutoff enabled in firmware of some ESCs to protect the battery by cutting motors if the voltage gets two low. The other is the driving circuit for the FETs.

FETs are basically just very fast, high power switches. By applying a voltage to the "gate" pin, current is allowed to flow from the "source" to the "drain". There is a minimum voltage required to switch the FET on, and this voltage is supplied by a driver circuit. As the FET drivers operate at much lower voltages than the ESCs typically operate at, the driver circuit is powered by a voltage regulator and diode. This takes Vbat and provides a constant lower voltage to operate the FETs. The issue arises when the input voltage is too low, and the voltage regulator cannot supply the desired output voltage. The minimum voltage required depends on the FETs and FET driver used, and these differ between 1S ESCs, 2S-4S ESCs, 3S-6S ESCs, etc. Further, different ESCs likely use different components, so the minimum voltage required by one 2S-4S ESC may be different than another 2S-4S ESC with the same specs. 

In the latter case, only the total input voltage matters. The FET driver has no way of knowing what the average cell voltage is, it only cares whether the total voltage is greater than the voltage required to trigger the FETs plus the voltage drop through the diode and regulator.

Joshua_A has a nice writeup on identifying and bypassing the voltage regulator and diode in order to allow the ESC to operate at a lower voltage (linked above). Edit: Felias has a great video on this too!

(07-Apr-2021, 02:39 PM)iFly4rotors Wrote: The follow is a sampling from actual flight log data at the end of the flight.
Various quads, but only the 3S and 4S runs and I didn't duplicate the information.

Total Pack Voltage /  per cell voltage ( VPC )
                    
GNB 3S 1100 mAh:       9.72  /   3.24  
GNB 3S 1100 mAh:     11.4    /   3.8
GNB 3S 1100 mAh:     10.4    /   3.46
GNB 3S 1100 mAh:     10.2    /   3.4

GoldBat 4S 650 mAh:  13.8   /   3.45   << Even the 4S dies at 3.45 VPC, but the Pack voltage is still 13.8 !!!
 
The quad drops from the sky before the per cell voltage reaches 3.0 volts regardless of the pack voltage.

Now, tell me how the cell count or pack voltage makes any difference. 
It simply Does NOT, it is clearly the Per Cell Voltage. 

Increasing the Cell Count did NOT make a difference in the per cell voltage nor the battery draw down.
That is NOT the answer and does not resolve or change the situation. 

So, that regulator is working on a per cell voltage, definitely NOT a per pack voltage. 
If you think about it, that is the only logical way since you can run a wide variety of cells per pack with varying voltages.
Trying to set the regular for various pack voltages is simply not logical, but per cell is.

It does, however, make sense to bypass that regulator; which I WILL do. 

I also believe that there might be some boards out there with the regulator set to a lower value; I just haven't found any.

If these are resting voltages, you are easily drawing these batteries down to ~3V in flight. The GNB 3S 1100mAh is a relatively low C battery, and the 4s 850mAh is quite small. Have you measured the battery resistance of any of these packs? If you regularly fly them down to such low voltages, I would guess the resistance is quite high and would cause significant sag, especially near the end of the pack. All of this points to your batteries giving out, not the ESCs. 

What ESCs are you running on this build?

Edit: I overlooked the most important factor with these results, which is battery chemistry. We're talking about 18650 Li-ion batteries here, but your data is from LiPos. If you look at a LiPo discharge curve, 3.4v is pretty much dead as voltage drops rapidly after that point. Li-ions, on the other hand, can continue providing significant current well below 3V. I suspect if you repeat your test with a 3S Li-ion pack with quality 18650 cells, you'll be able to operate well below 3.5V/cell.

First, the sampling of data above is from a variety of quads having different FCs and configurations. The point was NOT to show that the battery drain was being cut at approximately 3.5 volts per cell, BUT Rather to show that any limiter is on the per cell voltage and NOT on the pack voltage as was suggested by some. Yes, this data comes from LiPo batteries, however, the Li-Ion results are quite similar and only draw down to about 3.2 volts per cell when used on the same quad. 

Yes, if you are only running LiPo batteries, then it really doesn't matter since good battery management suggests that you don't drain a LiPo battery below 3.5 volts per cell. No worries, No issues, Not even noticed. The Low Voltage Cutoff doesn't even come into play unless you want to drain Li-Ion batteries down to 2.5 volts per cell for maximum Fly Time. 

The whole idea is to be able to drain a Li-Ion pack down to 2.5 volts per cell to extend the flight time as was the original concern from a variety of members including myself. WHY is there any Low Voltage Cutoff?  Where was it being limited? How? How to Bypass it? Some suggested software limits in BLHeli and maybe even in Betaflight. However, this was not a complete picture. The remainder was hardware. Regardless of the specific circuitry or reasons, the question was how to get around it. Yes, Joshua_A's research and fix did produce the desired results. 

Thanks go out to Joshua_A.

The goal is to be able to drain a Li-Ion pack down to 2.5 volts per cell.  Some FCs appear to lack this hardware restriction and allow the pack to drain down. On the other hand, some FCs {maybe many}  have hardware limits which must be bypassed via physical modification.

For me, I know all that I need to know. When I want to be able to drain an 18650 pack down to 2.5v per cell , I will use a HIFIONRC F7 AIO {the one Joshua_A used to perform the hardware regulator bypass}, perform the necessary hardware modification, and then check the software for any Low Voltage Cutoff limits; both BLHeli and Betaflight.  Since I can get over 20 minutes of Fly Time without the mod, I only need to perform this hardware modification for the one build that I need to get 30 minutes of Fly Time from. That is my Challenge build the Quest250-LR. Beyond that, I will likely Not do it again and will not be much concerned about it.

(03-Jun-2021, 11:30 PM)iFly4rotors Wrote: First, the sampling of data above is from a variety of quads having different FCs and configurations. The point was NOT to show that the battery drain was being cut at approximately 3.5 volts per cell, BUT Rather to show that any limiter is on the per cell voltage and NOT on the pack voltage as was suggested by some. Yes, this data comes from LiPo batteries, however, the Li-Ion results are quite similar and only draw down to about 3.2 volts per cell when used on the same quad. 

Yes, if you are only running LiPo batteries, then it really doesn't matter since good battery management suggests that you don't drain a LiPo battery below 3.5 volts per cell. No worries, No issues, Not even noticed. The Low Voltage Cutoff doesn't even come into play unless you want to drain Li-Ion batteries down to 2.5 volts per cell for maximum Fly Time. 

The whole idea is to be able to drain a Li-Ion pack down to 2.5 volts per cell to extend the flight time as was the original concern from a variety of members including myself. WHY is there any Low Voltage Cutoff?  Where was it being limited? How? How to Bypass it? Some suggested software limits in BLHeli and maybe even in Betaflight. However, this was not a complete picture. The remainder was hardware. Regardless of the specific circuitry or reasons, the question was how to get around it. Yes, Joshua_A's research and fix did produce the desired results. 

Thanks go out to Joshua_A.

The goal is to be able to drain a Li-Ion pack down to 2.5 volts per cell.  Some FCs appear to lack this hardware restriction and allow the pack to drain down. On the other hand, some FCs {maybe many}  have hardware limits which must be bypassed via physical modification.

For me, I know all that I need to know. When I want to be able to drain an 18650 pack down to 2.5v per cell , I will use a HIFIONRC F7 AIO {the one Joshua_A used to perform the hardware regulator bypass}, perform the necessary hardware modification, and then check the software for any Low Voltage Cutoff limits; both BLHeli and Betaflight.  Since I can get over 20 minutes of Fly Time without the mod, I only need to perform this hardware modification for the one build that I need to get 30 minutes of Fly Time from. That is my Challenge build the Quest250-LR. Beyond that, I will likely Not do it again and will not be much concerned about it.

I believe much of it is hardware limited, as per Joshua_A's post. A guy called Kenji Lin was doing this back over a year ago (if you search for his videos on Youtube he has done some amazing stuff - 1 HOUR flight time on 2S 18650, well under the 250g limit). He even has some videos on how to mod your ESCs to go below 5v. This is of interest to me because i am looking to go even further - to take a 2S FC and use it on 1S (and all the way below 2.5v). 

Current 1S FCs are not limited by cell voltage as i said, most are happy to fly to 2.5v because instead of stepdown regulators, they use step-up regulators to bring the battery voltage to 5v. The biggest issue is the current capabilities of that step-up converter as the input voltage drops, and the VTX usually draws too much when the voltage is below 3v.

But 1S FCs are currently very limited. None have a baro, none (that are available now) have I2C access (to add a baro/compass), some dont even have two full UARTs. All of these things start happening on 2S FCs. So for a fully featured 1S nano longrange, i will be using a 2S FC and just supplying a regulated 5v (via stepup) to the output of the 5v regulator on the board.

It seems very strange that there would be software limitations on per cell voltages on those higher S rated FCs. Its a bunch of processing that really doesnt need to be done and limits the usage of the FC. What happens if cell voltage goes below 3.5v during a punchout? cut power to motors and let the quad drop? A little extreme IMO

Hi Tokyo_Dom,

I do not know if the voltage regulator is tied to a timer and whether it would cut immediately or allow a "time" for battery voltage recovery before cutting the power.

I fly very easy on the sticks...most of the time just enough to maintain altitude and flight....generally about 20 percent throttle. Consequently, the battery drain is smooth and steady...as the voltage drops it has little recovery because it is truly the battery's continual draw and nothing to recover to... unlike a sudden power draw and release.