Eliminate Voltage Sag / Voltage Stabilisation

Hey everyone,

I think it is an issue, that an half empty battery provides less RPMs to the props,
and that the internal resistance of the battery causes a significant drop of voltage when going full throttle. This has caused a motor-out (or better: a battery-out) on my setup, since my BMS protects the battery by shutting it off, when the minium set voltage has been reached.

What do you think of a circuit that stabilizes the voltage at a set value, such as 50 Volts for example? This would result in always the same Voltage at the input of the ESCs, unrelated to the batterieā€™s actual voltage, and X% Throttle will always give the same result in RPM and thrust.

I was searching online for a step up (or step down) dc-dc converter with a possible output of 60 Volts, but didnā€™t find anything that would support more than 30Aā€¦

Until a colleague found something different:
(itā€™s the german ebay, but the productā€™s description is in english)
https://www.ebay.de/itm/300A-200V-72V-144V-DC-motor-PWM-Drehzahlregler-Kontroller-Speed-Bremsung-RS232/172668879474

It is a DC Motor PWM controller for up to 300A and the ability to program the output voltage.

Any input from you engineers?

1 Like

The size of the inductor needed for such power levels in a boost converter is substantial, and may require cooling.

There is no free lunch, and there are losses involved too. So lots of weight and some power loss.

Further, your cells will still be providing the power! That means they will still be drawn just as low as they had been.

Makes more sense to have higher discharge rated cells, or greater capacity to reduce C rate.

Additionally, a BMS that reduces throttle, rather than cut off power, may be a better solution for this application.

If we wanted to always have the same linear relationship between throttle and RPM it would be possible to do this with the new hub PCB in batch 4 that provides telemetry feedback for RPM. The throttle would then intelligently adjust the PWM in order to maintain a consistent RPM.

You could also do this without RPM data (batch < 4 ) and measure the RPM at different voltages relative to the PWM input. Then create a mapping and save it to the throttle to reference pack voltage as well as throttle position when sending PWM.

1 Like

We have tons of moving air around us!

the problem is the voltage sag.
My Battery is full with 54,6 Volt and empty at 32,5 Volt.
when the battery is at, letā€™s say 38 Volt, I can only go 35% throttle or so, which may not be enough to maintain altitude or even gain altitude.
I would prefer to still be able to go full throttle, to get over an obstacle for example, accepting an even shorter flight time.

the LG HG2 (18650 Li-ion cells) that I have in my system are the ones with the highest discharge rate available (within the 18650 Li-ion models).

I have decided to do just that. I already topped up my capacity from 2,16 kWh to 2,88 kWh
and my plan is to even double that some time towards the end of this year.

that would be good, however the BMS that I have is most likely not compatible with the OpenPPG arduino controller - and still, as mentioned above, Iā€™d prefer to be able to go full throttle at all times during the flight, if needed, rather than only being able to go only 30%throttle towards the end of the battery capacity.

I am definitely in, when it comes to upgrading my equipment to the newest controller, that you guys come up with - however the issue I was referring to is less the linear relationship between RPM and throttle,
but more the ability to still go full throttle when the battery is half empty (the voltage sag prevents me from doing this).

My point is that your BMS will still cut off even with the boost converter as the cells will be drawn just as low as they had been.

Makes more sense to add capacity than add a massive inductor and additional power electronics.

Also, just semantics, but you are still getting ā€œfull throttleā€(duty cycle 100%) but are reduced in voltage. I know what you mean though.

Okay, I think my non-engineer brain now understands your point :slight_smile: Thanks!

If I understand correctly, you have two problems:
1- BMS shuts off batteries at a low voltage
2- At lower voltages the motors wonā€™t spin fast enough

Your suggestion may fix option 2 (if there is enough capacitors in that device to handle the load), but it will just increase the amps coming from the batteries which will lower the voltage further causing problem ā€˜1ā€™ to occur even sooner.

Just my 2 cents.

Yes, but thatā€™s not a ā€œproblemā€ in that sense, itā€™s just a fact.
The BMS is designed to do exactly that to protect the individual cells from being damaged.

Yes, at least not as fast as they did with a full battery,
I would not call it a ā€œproblemā€ either, just an ā€œissueā€

Your conclusion is actually the answer to my question, just like @jhair 's conclusion,
so it make sense to top up capacity, rather than adding electronics (which also add weight and cost, and substract efficiency from the equation).

1 Like

Would 4 of that be useful?

If put 4 in series would they be good for 50v?

Or makes it more sense to put more cells with the same weight and size, that also helps alot to prevent voltagesag.

SmartSelect_20190509-005618_Chrome

Iā€™d like to weigh-inā€¦and take anything I say with a grain of saltā€¦

Here is the problem!

The system is not built for a voltage range from 32.5 v to 54.6 volts.

Your real life voltage range is from 54.6 v to 48.49 v without load! No problem to climb with 48.5 voltsā€¦

A better bms may help by calculating battery capacity better and hence not just turn off on voltage setpoints. Check out texas Instruments bq76940

Cheers, Evan

On their larger units, the MGM controllers have a function to correct for voltage sag, and they can be programmed to reduce power at a set voltage. They are expensive thoā€¦I tried to upload the manual, but itā€™s a bit too largeā€¦