Giant Scale Electric Backup Battery System

[kyleservicetech]

I've got 24 A123 cells and a Hacker A60-16M motor for a giant scale project to be worked on during the winter months. Several issues on the project exist, one of them being a primary/backup battery system for the model. It's true the Castle Creations uBEC's are reliable, mine have been working perfectly for three years. But, this will be a $1000 model, so the primary/backup system is the way to go.

Several considerations were made, including primary/backup Castle Creations uBEC's, Nicad backups, Lipo Backups, and so on. Finally settled on using the same 2300 Milliampere Hour A123 cells for the backup duty.

This system uses the CC uBEC as the primary receiver power system, with the A123 battery system as the backup. The backup system is isolated with two three amp diodes, first to isolate the A123 batteries from the uBEC, and second to drop the A123 battery voltage down to about 6.2 Volts DC or so.

Bench tests show it works very well. Next project is to install this system in my Showtime 50 for trials before next year. Did find though that a diode is required on pin 3 of the regulator. Otherwise the A123 battery pack can feed through the regulator backwards, keep the relay energized, and eventually run the A123 battery pack down in a week or so.

Take a look at the attached JPG's. Most of the components were purchased at Radio Shack. The 10 turn potentiometer was purchased from www.digikey.com. (If anyone should build up this backup system you need to add a small silicon diode (Part number D3 on the schematic) to the input of the voltage regulator. When the power was off, the A123 cells wound up backfeeding through the regulator, keeping the relay energized. That ran the A123 cells down in about a month or so.)

I eventually wound up setting the CC BEC voltage to 6.6 Volts DC, lower than the 7.0 volts shown on the schematic.

Take a look at the attached JPG's!

[kenchiroalpha]

Hi Dennis
Will be following this project with great anticipation
Take care
Hank

[CHELLIE]

Thats interesting, I was just wondering, could you put a diode on the pos feed on the battery and one on the UBEC, they would both feed, but there would not be a feed back between them, just a thought, Take care, Chellie

[kyleservicetech]

Originally Posted by CHELLIE

Thats interesting, I was just wondering, could you put a diode on the pos feed on the battery and one on the UBEC, they would both feed, but there would not be a feed back between them, just a thought, Take care, Chellie

Hi Chellie
Yes you can. I'd recommend a Shottky diode, NOT a silicon diode for this purpose. Shottky diodes have a lower forward voltage drop of about 0.4 volts DC under load, making it better for our purpose. They do cost more however, about $2.00 each, not including shipping. A number of club members have gone to this based on an article I wrote in the club newsletter last year.

You need to adjust the uBEC output voltage high enough so that the backup Nicad battery is not loaded during flight.

Schottky Diode www.digikey.com # 90SQ035-ND (Digikey only has 5000 left, better hurry!)

(http://search.digikey.com/scripts/Dk...5-ND&x=27&y=21).

This is a 9 ampere Shottky diode that will handle any type of receiver load you'd ever require.

[Larry3215]

As usual, nice job!

I do have a couple of questions and a concern though.

If Im reading the schematic right, you have it set up so the back up comes on if the 6S2P A123 pack goes dead?

Seems to me its far more likely the Castle BEC would fail than that A123 pack running dry or failing.

That 6S 2600 mahr pack could probably run your plane for 100 flights between charges.

Or wait - thats one of your main flight packs? Hmmmmm... Even if you hit LVC, it will still have enough reserve capacity to run the rx and servos for a long time.

Id feel safer using the BEC output voltage to trigger the back up kicking in

Second question - why the voltage regulator feeding the back up A123 pack? Is it charging that pack in flight or?

[kyleservicetech]

Originally Posted by Larry3215

As usual, nice job!

I do have a couple of questions and a concern though.

If Im reading the schematic right, you have it set up so the back up comes on if the 6S2P A123 pack goes dead?

Seems to me its far more likely the Castle BEC would fail than that A123 pack running dry or failing.

That 6S 2600 mahr pack could probably run your plane for 100 flights between charges.

Or wait - thats one of your main flight packs? Hmmmmm... Even if you hit LVC, it will still have enough reserve capacity to run the rx and servos for a long time.

Id feel safer using the BEC output voltage to trigger the back up kicking in

Second question - why the voltage regulator feeding the back up A123 pack? Is it charging that pack in flight or?

Thanks!
Lets go through how this backup works.

First, a pair of A123 cells could be used by them selves, just isolating them from the CC uBEC with diodes D1 and D2 on the battery backup system schematic. So, the receiver will work with EITHER the uBEC, OR the A123 battery. And if both are activated, the supply with the highest voltage will run the receiver, and the two D1, D2 diodes block current from the uBEC backfeeding to the A123 pack. Two diodes are used to lower the voltage from the A123 pack slightly.

And, that's all you'd need. A pair of diodes.

But me, not understanding the Keep It Simple System (KISS), had to build up a charger for those A123 cells so I don't have to haul the giant scale model to my basement workshop to charge them.

So, that LM317 linear voltage regulator has been adjusted to put out exactly 7.20 volts DC to charge the A123 batteries whenever the motor batteries are powered up.

Resistor R4 (now changed to 50 ohms 4 watts) limits the charge current to the A123's at about 150 milliamperes. The 12 Volt DC reed relay was required, since with the motor batteries disconnected, the A123 batteries would drain out through resistors R1 and R2.

And, D3 had to be added since the LM317 regulator would backfeed and keep the relay energized, running the A123 battery dead in several days.

Lot's of work, but after 10 flights, it has worked out very well. Normally, the A123 pack is never used, but if it is, on landing that LED will be lit up indicating something has happened. I've also got a little microcontroller alarm on the uBEC, that will let me know if its output voltage ever sagged below 6.00 Volts DC. And record just how far it sagged if it did. All that for a $1000 model. (I've built up 15 of those alarms, and club members bought up all of them!)

[CHELLIE]

Originally Posted by kyleservicetech

Hi Chellie
Yes you can. I'd recommend a Shottky diode, NOT a silicon diode for this purpose. Shottky diodes have a lower forward voltage drop of about 0.4 volts DC under load, making it better for our purpose. They do cost more however, about $2.00 each, not including shipping. A number of club members have gone to this based on an article I wrote in the club newsletter last year.

You need to adjust the uBEC output voltage high enough so that the backup Nicad battery is not loaded during flight.

Schottky Diode www.digikey.com # 90SQ035-ND (Digikey only has 5000 left, better hurry!)

(http://search.digikey.com/scripts/Dk...5-ND&x=27&y=21).

This is a 9 ampere Shottky diode that will handle any type of receiver load you'd ever require.

Great Info Dennis, Thank You, Chellie

[Larry3215]

Ah! I understand now

Not KISS for sure, but it should work just fine

[kyleservicetech]

Originally Posted by Larry3215

Not KISS for sure, but it should work just fine

Well now.
After a dozen flights, with the Castle Creations 10 Amp BEC set at 6.6 Volts DC, virtually zero milliampere hours were pulled out of the two cell A123 backup battery pack during a days flying.

And after a dozen flights on this model, it was a little hard to land, and several club members indicated it looked tail heavy, even though the cg is 3/4 inch IN FRONT of the rear most suggested cg point.

So, the backup battery with its charger, located 8 inches behind the cg was pulled. Next, the two cell A123 pack was rigged up with two series 9 Amp silicon diodes in its red wire, along with a balance connector. These two silicon diodes drops the voltage by about a volt or so, depending on the load, allowing the CC BEC to carry 99% of the power required by the receiver and servos. Those diodes also prevent interaction between the A123 battery and the CC BEC.

This backup battery was wired with #18 wire, and was placed up by the motor, a foot in front of the wing spar. That moved the cg forward 3/8 inch, which made a big difference.

The two cell A123 pack is now simply topped off with my Cellpro Powerlab 8 before flying for the day. And, it has worked out very well.

If anyone is interested, I can provide a schematic, let me know.

Now, it really is KISS, Keep It Simple Stupid!

[ramboman]

Hello
I am working on a similar project...
Taking into account the "float" voltage of a123 cells, I am came to the following concept...
I am using a CC bec pro set to 7.0V
A first diode drops the voltage to 6.6V... enough to keep the 2s a123 pack charged and to avoid back current... a second diode drops voltage to 6.2V and isolate from the second branch.
The second branch gets two diodes to get the appropriate voltage.
That's it... I would appreciate your comments.

[Larry3215]

Thats not a safe way to protect your servos if they are digitals only rated for 5 volts.

The problem is that the amount of voltage drop you get across the diodes depends on the current flow. So when the servos are under heavy loads they will get less voltage and when they are under light loads or no loads they will get too hi a voltage.

Digital circuitry in some servos can be very sensitive to over voltage - even at low loads.

However, that varies from servo to servo. Check the big bird forums on sites like Flying Giants to find out which servos can actually handle A123 voltages direct and which cant.

A much safer option - but more expensive - would be to use a linear regulator between the A123's and the rx. Fromeco, Duralite, Smartfly and MPI are a few manufacturers. Prices vary from the $20 range for low amp models to $50 for 10-15 amp models.

[ramboman]

Originally Posted by Larry3215

Thats not a safe way to protect your servos if they are digitals only rated for 5 volts.

The problem is that the amount of voltage drop you get across the diodes depends on the current flow. So when the servos are under heavy loads they will get less voltage and when they are under light loads or no loads they will get too hi a voltage.

Digital circuitry in some servos can be very sensitive to over voltage - even at low loads.

However, that varies from servo to servo. Check the big bird forums on sites like Flying Giants to find out which servos can actually handle A123 voltages direct and which cant.

A much safer option - but more expensive - would be to use a linear regulator between the A123's and the rx. Fromeco, Duralite, Smartfly and MPI are a few manufacturers. Prices vary from the $20 range for low amp models to $50 for 10-15 amp models.

Larry, 2s a123 is ok, with or without diode, for most radios and servos that accept 5s MiMh. There are of course some exceptions.

Voltage fluctuation on the Schottky is max. 0.1V... much less than fluctuation on RX pack under charge. I plann to make some chart when testing the project.

[kyleservicetech]

Originally Posted by ramboman

Hello
I am working on a similar project...
Taking into account the "float" voltage of a123 cells, I am came to the following concept...
I am using a CC bec pro set to 7.0V
A first diode drops the voltage to 6.6V... enough to keep the 2s a123 pack charged and to avoid back current... a second diode drops voltage to 6.2V and isolate from the second branch.
The second branch gets two diodes to get the appropriate voltage.
That's it... I would appreciate your comments.

I've programmed my CC 10 Amp BEC to 6.6 volts DC, and connect it directly to the receiver's battery input. The backup A123 system is connected through a 10 Amp Silicon diode to an unused servo connector on the receiver through a switch.

So far, this setup has worked very well, and very few milliampere hours (perhaps 25 mah) is pulled out of the two cell A123 pack during a days flying.

The CC BEC has an undervoltage monitor on it, and it has never dropped below 6.0 Volts DC, with a model that has seven Hitec 645MG servos on it.

As far as the voltage limit on the typical servos, I've conducted discharge tests on a brand new 5 cell, sub "C" type 2000 Milliampere Hour Nickel Hydride receiver battery. The discharge curve on the Nih battery at 5 amps was virtually identical to the discharge curve on the two cell 2300 Mah A123 battery pack.

So, if your radio and servos work with a 5 cell Nih pack, it will also work with the two cell A123 pack.

But, some servos don't work well with the 5 cell Nih pack, so do be careful.