DIY LCDC (Low Cell Detect Circuit) by Bob Kopski

[Ken Myers]

Bob Kopski has designed a simple electronic circuit device that looks at each of a Li-Po cell's discharge through the taps during a flight and adjusts the throttle down, NOT OFF like a LVC, when the voltage in one cell drops below a given voltage. When the voltage rises, the throttle is fully responsive again until that or another cell drops down to the predetermined point, and the process starts again. This process protects the cells in the pack from over discharge. Several Ampeer electric flight newsletter readers have already built and used this circuit very successfully.

To learn more, http://members.aol.com/richardyea/ampjun06.pdf

Has anyone in this forum built and tried one yet?

There will be an update to the circuit that allows for higher "detect" voltages in the upcoming July '06 issue of the Ampeer.

[Gary Warner]

Can you help me understand 'why' this device is used?

If a cell drops too low and the circuit cuts the power back until the voltage returns, allowing full throttle authority should cause the low cell to once again trip the low volt detector when full throttle is applied. I don't see the advantage of repeating the low voltage routine when a cell is already considered exhausted. Am I missing something in the functional logic?

I can see it being useful when over-amping the battery pack causes a drop below 3 volts on a cell, dictating that the pilot needs to either fly at a lower throttle setting or needs to reduce the load by changing gears or props. But excessive loading should be found on the bench before any flight happens with an amp meter. I don't 'get it'. :confused:

Gary
--

[Acetronics]

Hi,Gary

Yes and no ....

Cells have an internal resistance ( I make it short ...do not blame me :D ) ... so, at full throttle, power is cut .... OK ???
after that control comes back ...

If you only use low or mid throttle, batt voltage is not at the lowest permitted, so you can fly your plane gently ( but immediatly ...) to the landing zone ...

You've been alerted !!!

Alain

[Gary Warner]

So this is like the ESC's that I have that allow you to re-arm the ESC by pulling the throttle back to off and then you can advance the throttle again. This made sense since you have to reduce the load to expect it to have 'flyable' power to make the runway for a safe landing. If this LiPo low volt protector reenergizes the power with the throttle stick in the same place, the results will be the appearance of a quickly dieing battery pack. The natural reaction from a pilot would be to add power, not cut it back.

I guess I'd have to try it for myself to decide it it's an improved way of handling low volt situations. Seems to me that getting more power with a reduction on throttle stick movement would have an unnatural feeling about it.

Gary
--

[FenceMagnet]

no, it doesn't give you more power as you throttle back.... it basicly clamps the throttle to a safe setting for the depleting cell(s).... you can reduce the throttle from there, but you can't go up.... if the cell(s) "recovers" a little as it cools down, the throttle clamp point may increase a little, but that's NOT the idea.... protecting the weak cell(s) is primary...

:cool:

[Gary Warner]

Thanks for the explanation. Now it's clear. I 'get it' now. :)

Gary
--

[Acetronics]

no, it doesn't give you more power as you throttle back
:cool:

Hi, Fence

Not more power ... but more energy !!! LOL ;)

Alain

[Ken Myers]

Instead of speculation, please read what Bob wrote in the April 2006 Ampeer, http://members.aol.com/kmyersefo/ampapr06/ampapr06.htm

Also, Bob asked me to post this. It is directly from him.

"The (DIY) Low Cell Detect Circuit (LCDC) is intended to monitor LiPo packs on a cell-by-cell basis and gradually, automatically retard throttle (current drain) as the first cell within a pack drops to a safe "detect" voltage level. As such, it protects against any cell from being discharged too low. Operation is smooth and gradual allowing plenty of go-round opportunity. The LCDC allows full throttle control from zero up to whatever power level is available when a going-low cell is detected and as the declining pack condition permits. The LCDC does not stop the motor and "rearming" is not necessary although power would eventually be reduced to "unflyable" levels. Basically, the LCDC does not permit overdischarge damage of any cell - no matter what! Think of this as the "discharge equivalent" of the "care and feeding" associated with LiPo "balance charging".

[Gary Warner]

Thanks Ken. The original post was confusing to me seeing how I’d never considered this approach. It sounds like a good idea and a good way to protect these expensive packs.

Charging up some packs last night for this weekend's flying, I noticed a TP2100 3 cell that's not as flat as it once was (but not puffed up with air)and it has some green corrosion along the inside and the TP label is faded. Looks like this pack took a bit of over discharge damage in my Switch Blade. I think I do recall having a bit too much fun the last time it was flown. I probably ran the pack down too far.

Gary
--

[Acetronics]

Hi,Ken

"Operation is smooth and gradual " ... yessss, it is supposed to extend somewhat the pulse duration with liPo voltage passing under the 2.5v+Vbe limit - WITH a battery voltage reducing SLOWLY. ;)

Yesss, but, if you have a look to LiPo's discharge curves, you'll see them vertical at that voltage ... means voltage is falling at the speed of light !!! :eek:



AND your esc. quickly cuts the POWER ... :p

I just build one to confirm ... I do not want to dy as an Idiot !!!

Alain

PS : Note it is just a question of trip point setting ... ;)

re PS : explanation here : http://members.aol.com/dublinel/lcdc2brd1.PDF

[FenceMagnet]

I've got two Apex 15c 3S 2100 packs that are "not as flat" as new.... seems to have happened when I charged them up in Florida at sea level and flew them in my MiniUltraStick in Denver at 5000'+..... they seem to be working just fine... :confused:

[Dan Baldwin]

Hi,Ken

"Operation is smooth and gradual " ... yessss, it is supposed to extend somewhat the pulse duration with liPo voltage passing under the 2.5v+Vbe limit - WITH a battery voltage reducing SLOWLY. ;)

Yesss, but, if you have a look to LiPo's discharge curves, you'll see them vertical at that voltage ... means voltage is falling at the speed of light !!! :eek:



AND your esc. quickly cuts the POWER ... :p

I just build one to confirm ... I do not want to dy as an Idiot !!!

Alain

PS : Note it is just a question of trip point setting ... ;)

re PS : explanation here : http://members.aol.com/dublinel/lcdc2brd1.PDF

I haven't built one, but I believe the cut out point for each cell is about 3.15 volts (2.5 volts for the TL431 plus .65 volts emitter-base voltage on the sense transistor). Alain, can you confirm or refute that?

Dan

[Acetronics]

hi,Dan,

you're absolutely right ;) ...3.1v !!! see re PS.

Here joined LiPo curves ...

Alain

[BEC]

It's really simple - the LCDC monitors each cell individually rather than the pack as a whole, and does a "soft cutoff" not unlike the soft cutoffs of, say, Jeti controllers. It acts when the FIRST cell reaches the set threshhold instead of waiting until the whole pack gets there. Since it's not the speed control, it does this by commanding the speed control to a lower throttle setting such that the the lowest cell stays above the desired point.

The FMA Direct "discharge protection modules" are another implementation of the same basic idea albeit with a lower trip point . At least in the case of the Cellpro 4s version there is a rougher cutoff that is intended, like the not-so-smooth "soft" cutoff of Castle Creations controllers, to let you know it's acting. Unfortunately the Cellpro unit trip point is so low that the situation Alain talks about does happen - the voltage is falling off very steeply and there's nothing really left to land with.

I believe Bob has or is going to offer up a slight component change to the LCDC to raise the set point to 3.3V per cell which will help a bit more.

[arx_n_sparx]

I believe Bob has or is going to offer up a slight component change to the LCDC to raise the set point to 3.3V per cell which will help a bit more.
This would be a really good thing IMO. Hopefully he will post it here - seeing as this thread *is* about his circuit and all... ;) It would be nice to thank him personally.

Brad

[Acetronics]

Hi, All

I think he already has done that ...

see: http://members.aol.com/dublinel/lcdc2brd1.PDF ...

for calculation : Ra = 4700/2.5 x ( UT( Trip point ... ) - 3.1 ) in Ohms.

But,
1) one have to remember the esc. MUST reduce power for input pulse increasing ...

2) What is the circuit behaviour with 5 "radio" Cells instead of 4 ??? have to re-compute R12 and R13 ???

3) What is the sensitivy to RC Batt voltage changes ???

Alain

[Ken Myers]

Bob's "upgrade" to a higher trip point of your choice is in the upcoming July 2006 Ampeer, which will be posted in about 10 days.

If you have a specific question, please PM me, and I will see that Bob gets it and gets back with you. Please include your email address to me in the PM so that Bob can email the answer back to you.

[pmackenzie]

But,
1) one have to remember the esc. MUST reduce power for input pulse increasing ...
I think it is the other way around. Shorter pulse = lower throttle:

This loads the collector
signal of Q4 resulting in the gradual narrowing of
throttle control pulses to effect throttle retardation.
This in turn reduces current demand on the pack and
preempts further decline of the cell that caused this
circuit action in the first place.

I have yet to encounter an esc that didn't work this way.
Pat MacKenzie

[Acetronics]

Hi,Pat

Interesting point of view ... I didn't read the notice, but tried to understand the scheme !!!

... I must see that on my scope ... for me C1,C2,or C3 low voltage locks the pulse trailing edge to allow Q5 conduction , just for a while ...

May be I'm wrong ...

Alain

[BEC]

Every currently available ESC (that I know of, anyway) has increasing throttle with increasing pulse width. That's the way all transmitters except Futaba work as well - and why Futaba owners have to reverse the throttle channel at the transmitter to use an ESC.

Now I can name two ESCs that are out of production - one that worked the other way to be Futaba compatible (early Viper Models units) and one that could be set up EITHER way (MEC MX-50/80) - but you're not likely to encounter either anymore.

[Ken Myers]

I have posted the July Ampeer today, June 20, 2006. Version 2 of Bob Kopski's LCDC is in this issue. Version 2 allows for the voltage "trip point" to be set to a higher voltage. The previous version can be easily modified to do so.

The July 2006 Ampeer, electric flight newsletter of the Electric Flyers Only of southeastern Michigan:
http://members.aol.com/kmyersefo/ampjul06/ampjul06.htm

[arx_n_sparx]

Thanx, Ken!

Brad

[jcyr1]

Ken & Everyone,

[jcyr1]

Hi All,
Srewed up again and sent prematurely. Anyway, I've got the 5cell LCDC finished and working(with an active source). I'd like to include photos but wife left with the camera(gone to Europe). It works just like the 3cell and the 4cell I've built previously. This one used a "Piggy back" board. The boards stack upon each other with Buss Bar Wire as connections. Only had one problem, when connecting the boards, I forgot to mirror the connecting points so when I went to solder nothing matched(looked good on paper!!!!!). Problem solved and everything works great. No more 5cell packs overdischarged!!!!! If anyone has the capability to make a smaller board with SMT's let me know.
John

[Ken Myers]

Bob Kopski caught an error in the July 2006 issue of the Ampeer for the LCDC two that allows for the higher detect voltage for the Li-Po cell. The corrections have been made to the online versions. Anyone wishing to build this version should use the corrected versions at:

http://members.aol.com/kmyersefo/ampjul06/ampjul06.htm or
http://members.aol.com/kmyersefo/ampjul06.pdf

Sorry about the error.

[flieslikeabeagle]

Now I can name two ESCs that are out of production - one that worked the other way to be Futaba compatible (early Viper Models units) and one that could be set up EITHER way (MEC MX-50/80) - but you're not likely to encounter either anymore.
About two years ago I had a Wattage 10 amp brushed-motor speed control (called the IC-10A, if memory serves) in my aileron trainer, a GWS E-Starter. This ESC automatically set the throttle endpoints, based on assuming zero throttle was whatever pulse width it saw when first powered up, and full throttle was whatever pulse width it saw during the arming sequence. For the ESC to arm, it had to see three or more seconds of zero throttle, followed by three or more seconds of full throttle, followed by three or more seconds of zero throttle.

One day I bumped the throttle stick as I took the Tx out of the car, and then compounded that by making the very common beginners error of not double checking the throttle joystick position before powering up the model. So I powered up the model, set it down on the grass facing into the wind, then noticed the joystick was at nearly full throttle, and pulled it back to zero. I then pushed it all the way forward, waited a few seconds for the ESC to arm, and pulled it back to zero. If you've already figured out what happened next, give yourself a gold star - as I pulled the stick back to zero throttle, the motor zoomed up to full speed, and the model took off across the field, shocking the wits out of me!

Since my hands were already on the transmitter sticks, I was able to immediately take control and fly the model up to altitude, where I started to do circuits (still at full throttle, with the throttle joystick pulled all the way back to zero!) while trying to figure out what on earth had just happened. After wondering whether I'd be forced to fly for 15 minutes or more at full throttle before I could deadstick the thing in, I finally realized the ESC had managed to get its endpoint settings reversed, because I had not properly zeroed the throttle joystick before powering up the ESC!

Sure enough, pushing forward on the throttle stick slowed down the motor, and pushing it all the way forwards turned the motor completely off. Okay, I thought, this will be easy, I'll just deadstick the landing, and my troubles will be over. But I had not realized the extent to which my RC piloting reflexes were automatic - as I came in for the landing and started the flare, my left hand, seemingly on its own, reached out and zeroed the throttle, as I had trained it to do on hundreds of previous landings. Of course that now fed full throttle into the model, and it zoomed off again, this time at waist height and heading for the perimeter road around the field immediately to my right, where a young lady was just jogging by. No time to think about the throttle - I had to make sure I didn't hit the jogger - so I pointed the plane into the nearest tree, and let it crash to a halt.

Believe me, an ESC that is able to alter itself to believe that zero throttle corresponds to maximum pulse width is an evil, pulse raising, plane destroying device. :D

The E-Starter came out of the tree with minimal damage, and flew hundreds more flights. That Wattage ESC, however, had to go - I pulled it out of the E-starter and clipped off the motor leads emerging from it with wire cutters. Afterwards it served duty as a BEC in a slope glider, but I was not about to trust it to run a motor ever again!

-Flieslikeabeagle

[arx_n_sparx]

I had a heckuva time trying to find the PCB layout; it's here: http://members.aol.com/dublinel/LCDCBrd2.pdf

Brad

[arx_n_sparx]

Ken: I was wondering if you could have a chat with Bob Kopski about a lower voltage cut-off for the A123 cells - something in the 2.0-2.2V range. I dunno if this is do-able with this circuit, but I figure asking is free... :p

TIA,

Brad Kacer

[xtal]

Did anyone consider using a PIC with ADC's and a few resistors to
accomplish the same function....

[Acetronics]

Hi, Xtal

There's a PbP Basic example on Melab's site ( 12F675 ...) see support section.
" 12F675_Voltage_Monitor.bas "

Just the pulse modify section to add ...and the LCD section to remove.

There is a very interesting calibration method in this example ...

Alain

[xtal]

Hi Acetronics

Do you have a link for MeLab ,,, I've never been there.....

Never mind... found it...

http://archive.devx.com/sourcebank/download.asp?resID=1032112&url=http://www.melabs.com/resources/samples/submitted/12F675_Voltage_Monitor.bas