Thread Tools
Aug 16, 2004, 05:25 AM
Registered User

Yet another Lipo lvc to protect from undervoltage on older esc's: schematic

I have seen a few schematics on this board that implement a low voltage cutoff or warning for use with lipo cells.
The problem is that you need to make sure you don't discharge your Lipo cells below 3 volts per cell. There are 2 ways to do this:
- make an audible or visible warning system (LED or Buzzer)
- cut the engine (hard of soft) when batteries get low

I didn't like the first solution, so I looked at schematics for the second one.
After looking at 2 solutions (1 based on nand gates, the other on a voltage reference with a transistor to ground the signal) i desided I had to make my own version:

- only 4 parts (3 resistors, 1 comparator)
- only 6X6 holes on a breadboard
- possible to mount in-line on the servo cord coming from the esc and going to the plug that goes into the receiver

I will post some pictures in the next posts
Sign up now
to remove ads between posts
Aug 16, 2004, 05:41 AM
Registered User


This is a picture of the finished prototype.
A few minutes later the construction would be wrapped in a bit of duct tape.
Aug 16, 2004, 05:43 AM
Registered User

back of breadboard

Here you can see how it is soldered.
It is actually not too difficult to solder on breadboard this compact if you first make a breadboard design (see next post)
I find the breadboards with the round copper surfaces work easier for this then the ones with the square copper surfaces.
Aug 16, 2004, 05:52 AM
Registered User

breadboard design

And this is the breadboard design.

A bit of explanation to my "syntax" here:

red square = LM2903 dual comparator, 8 pins DIL
X in red = hole with a pin in it (either from component or wire)
green rectangles = resistors
lines in green = connections between pads, to be made on the back of the board
curly green lines = cables attached to the breadboard
blue arrow = indicates an error, X should be moved.

Wires to the left and right are red-orange-black from top to bottom,
left wires go to esc, right wires go to receiver.

The wire to the top left goes to the positive of the battery (for ease of connectivity I chose to connect it to the + of the engine connection on the ESC, which *with my esc* is electrically the same as the + of the battery. PLease check, this is only correct if your esc is using N-channel mosfets to switch the negative)

Note the error in component values: the 3K3 resistor actually should be 5K1 for a cut off voltage of 8.75 Volts. I found this calculation mistake whilst testing, so also in the pictures the wrong resistor value is shown.
Aug 16, 2004, 06:06 AM
Registered User
Because I know my notes are hard to decipher, here is a quick schematics drawing:

U1 = LM2903 (only 1 of the 2 comparators in this IC is used)
R1 = resistor 5K1
R2 = resistor 6K8
R3 = resistor 2K2

X1 = signal coming from receiver
X2 = signal going to esc
X3 = regulated +5V (red wire) from ESC (and to receiver)
X4 = negative (black wire) from ESC (and to receiver)
X5 = red wire from battery pack (or: on most ESC's positive motor connection can be used)

circuit explanation.

The LM2903 is an open collector output comparator.
The - input of the comparator always sees a regulated 5V as coming from the BEC of the ESC.
If the voltage in the battery pack is higher then 8.75V, the resistor divider R1/R2 is going to show a voltage higher then 5V on the + input of the comparator.
So since the voltage on the + input is higher then on the - input, the open collector output of the LM2903 will be high, meaning that the output transistor will be off. The signal can travel freely from receiver to esc through R3.
Should the voltage of the pack drop below 8.75V, the + input from the comparator will see a voltage lower then 5V.
So since now the voltage on the + input is lower then on the - input, the open collector output of the LM2903 will be low, meaning that the output transistor will be in saturation. The output will be pulled to ground (0 volts) by the transistor and the input signal will fall over the R3 resistor, ensuring the output current the receiver needs to deliver is kept small. The input of the esc no longer sees a signal and will do whatever it does when signal from the receiver is lost (mostly: shut down the engine until signal is restored).

WARNING: do not use an opamp or comparator with a TTL/totem pole / driver on the output stage. Only use an open collector comparator (like LM2903). Otherwise 1) it will not work and 2) you might fry your receiver.
Aug 27, 2004, 06:24 AM
Registered User
I can now give you a full report on actual in-plane testing:

- when battery gets low, the circuit begins by letting your engine stutter (+/- 0.3 sec interruptions in your trottle, starting each +/-5 seconds and occuring more and more often until you have real stuttering at a pace of about 1 interruption/second

- when you don't obey this warning (by reducing trottle and landing), after about 30sec the interruptions become a little longer. At that moment the esc is thinking "my God, what are all these signal losses here? Let's shut down the engine and keep it off until I see an idle trottle setting on a stable signal. Once we are there, I'll cooperate again.)

- this results in the very same behaviour then the esc's own cut-off: engine is stopped and you can only resume by first going to zero trottle. If you give too much trottle, voltage will drop again and the same history repeats itself.

So I'm happy with my work, now let's hear who else likes to build this circuit.

I think with only 4 components and total cost < 1 Euro it's the smallest and cheapest Lipo-saver circuit around.


Aug 31, 2004, 03:40 PM
Registered User
KODEL---Very interesting.
I think it would help most people if you showed a complete (in the plane) schematic showing the Motor, Batt.,Recr. ESC,your Batt. alarm circuit,etc.
How about it.
Arent there any single comparator ICs that could make the circuit even smaller.
G Pearce
Sep 01, 2004, 05:54 AM
Registered User

connection diagram

Here is a diagram that shows how to connect things.

Regarding use of a singlecomparator IC:
I don't think there exists a single comparator IC in something like a DIP6 package, the smallest I'm aware of is DIP8. There are some TOxx "can" type single comparators, but I doubt you would be able to solder those more compact. Only way to go smaller is using an SMD IC, SMD resistors and a dual layer PC. A lot of added complexity to gain a couple of millimeters (you still need space to attach the wires with some mechanical strength...)


Sep 01, 2004, 08:00 AM
Crash master...
Lazy's Avatar
Great man...
But - if I use diff. packs - 2S/3S/4S...May be use DIP switch ?
Sep 05, 2004, 08:32 PM
Registered User
hul's Avatar
I built the circuit on strip board. It's about 13x18mm and 4 grams with connectors.

It didn't make the motor stutter in my case but cut it without warning. The speed controller could then be reset (throttle to zero). That gave me enough power to fly around the field to land on the strip.

Resting voltage turned out to be about 10.7V on my 3s pack.

Good work, Kodel.

Sep 12, 2004, 01:02 AM
Registered User
HELModels's Avatar
I might agree that your design is the best for what I have been looking for in Lipo saver circuit. If anyone can answer my question about the operation of this circuit than It might confirm to me ,and possibly other electronically challenged, this circuits usefulness.

X1 is the signal from the receiver?
What is the purpose of R3? To stop IC output from going into Receiver?
would the +5V from the comparator alter the throttle signal coming from the receiver, or I should say what is the non-cutout output from the IC? Is R3 to stop IC output from going into receiver?

It looks cool I just need to understand it better before trying it.
Sep 13, 2004, 08:55 PM
Registered User
hul's Avatar
X1 is the signal from the receiver. The comparator grounds its output pin (where X3 connects) when the voltage is low. R3 limits the current from the receiver to ground through the comparator to about 2mA.

With voltage high enough, signal from receiver goes to speed control unchanged. The added resistance in the signal path doesn't matter because no current flows.

Sep 15, 2004, 03:16 AM
Registered User
Thanks Hans for taking care of the reply for me.

The X1 -X5 are where you cut/solder/connect a wire, it's not a component by itself. In Spice I used this "connector"component to be able to simply label the wires.

FYI, the device is still doing perfect in my Formosa.
Sep 15, 2004, 10:30 AM
On patrol.
Cary Reeves's Avatar
I have successfully built one of the buzzer type lv warning devices but have a hard time hearing it. I would like to try your design. I can follow someone elses circuits but know nothing about design and substitution. Also, I am not familiar with your notation.

Do I read your notation of 5k1 as a 5.1k resistor? Also, are the tolerances critical? I.e. 1% required, 5% o.k.?

This is for a 3s battery pack? What would change for a 2s battery pack?

Thanks for your contribution.
Sep 15, 2004, 11:55 AM
On patrol.
Cary Reeves's Avatar
Two more questions:

1. X3 and X4 each represent two connections? The red wire from the ESC and the red wire to the REC both connect to X3 and similarly with the black wire on X4. So, X3 and X4 tap into the BEC power in parallel, X5 is the battery monitor, X1 and X2 are the signal?

2. I assume if X5 is open (0 volts) that that is the same as low voltage and the ESC with shut down the motor?

I ask #2. b/c if opening X5 for some reason allowed the ESC to work normally (which I doubt) then it would be an easy matter to connect X5 to the positive side of a double ended battery plug and plug it in between the battery and ESC when using lipos and unplug it when using NiMh. This is how I did the buzzer circuit which only needs to be parallelled to the battery terminals. Then I can choose to use the LV monitor as needed.

Another option would be to use a servo extension to make modular leads for X1-X4 and a pair of battery plugs at X5 so the whole unit can be plugged in or removed from the plane as needed. This adds $3.60 in connectors to your <$1 circuit . But it would also allow for 2s and 3s versions of the circuit to be swapped out on the same plane.

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Using a FMA Lipo LVC with a GP C-10 micro ESC andyg Power Systems 0 May 27, 2005 05:48 PM
Yet another thing that can go wrong on a Slow Stick bradm Aerial Photography 31 Oct 14, 2004 10:26 AM
yet another lipo question nose_first Batteries and Chargers 4 Dec 13, 2003 03:39 PM
Yet another LiPo question please! new2heli Batteries and Chargers 4 Sep 22, 2003 09:01 PM
Yet another GWS Zero question - rolls left on launch Peter Khor Parkflyers 5 Jul 14, 2002 11:09 PM