Real world results with A123 pack.

[wildhare]

OK for all you folks who are interested in the A123 batteries, I have some real world data for you.

This data was collected today, 11/23/2007 flying my 106” 100cc Giles G202. The plane weighs about 27 lbs., has a DA-100, 6 5955s and 1 8711 on the rudder. No matchbox, the ailerons are on a Y cable on 2 channels, 2 servos per channel. The linkages are as close as I could get them mechanically and there seems to be no real binding although there is probably a small amount of mismatch at extreme throws, it’s hard to get it perfect.

In the plane I have a single A123 2300mah pack that goes to two switches through a Y and then into the receiver using two standard JR servo connectors. It’s a JR 921 2.4 GHz. Receiver and I use a JR X9303 transmitter, not that it really matters.

I made two flights of about 10 minutes each, with (apparently) 3-4 minutes of recorded data from the startup period. I did a range check, taxied around a bit, the usual.

When flying I flew a lot of typical IMAC type maneuvers, loops, hammerheads, humpties, lots of rolls and point rolls, and a lot of snaps. I flew several multiple snap sequences of 3 or 4 rolls in an effort to put as much stress on the electrical system as possible.

I used my Eagletree systems elogger V3 and the included software that downloaded it to my PC and analyzed it.

Before doing my flight test I tested the logger itself by running my battery tester through the logger and then recharging. The Eflight power meter said I used about 1430 mah, the elogger said I used almost exactly the same and the Cellpro charger recharged to pack and said it put in 1431 mah. Close enough; all 3 readings agree within a fraction of a percent, I think we can trust the logger’s data.

I collected data samples 10 times per second, or every 100 milliseconds. Any event lasting less than 100 milliseconds could theoretically have been missed. Any voltage spike or current surge lasting greater than 100ms will certainly have been recorded, and a shorter duration event would have had to be positioned exactly between the sample periods to be missed, making hidden events even more unlikely. I think that what we see here is a good representation of exactly what happened.

My only disappointment was the fact that I was unable to save the graphs off as a jpeg file which would have been useful in making this report. The best I could do was to take a photo of the computer screen which is not real clear. In the photos the pink traces indicate current draw, the blue line shows voltage.

So, on to the data. What did I learn?

The freshly charged pack started out at 7.05 volts. It took about 5 minutes of usage/flying to come down to 6.6v, and another 3 or 4 of flying for it to settle in at the 6.5 volts that it stayed at for the rest of the test. Average voltage throughout the test was 6.58 volts, and the lowest voltage recorded was 6.12v. According to the graph voltage dipped below 6.2 volts about 6 times, probably during the extended multiple snaps. The voltage dips coincided with spikes in current usage as one might expect. This A123 pack is not going to allow reboots, you’d need to melt the wires to pull it down to 3.5 volts.

What about current demand? The maximum current drawn in both flights was 11.56 amps. There were about a dozen times when current draw went over 6 amps, and a half dozen times when it exceeded 8 amps. Again, the voltage drops coincide with the current spikes.

The average current demand throughout most of the 2 flights was in the 2 to 3 amps range. It never fell much below 2 amps and rarely exceeded 3 amps. Based on my recollection of the flight, the big current draws were in long snaps and long knife-edges. The rest of the time power usage was fairly modest.

Total power used was 479 mah; this is pretty consistent at around 25mah per minute of flight time. A single 2300 mah A123 pack will support at least 60 minutes of flight time plus all the messing around that goes with it with probably a 25% cushion at the minimum.

So, what do we know? On a plane of this size, especially if it uses multiple rudder servos, you should probably have 15 amps available in bursts, and that burst should not pull the voltage down much if you want your flight performance to be consistent. But you can probably fly IMAC basic with no more than a 1200mah Nimh pack since that type of flying rarely draws more than 3 amps.

I would speculate that a 40% plane with 14 servos could easily require double this amount of power, needing burst current of 30 amps but living between the bursts easily on 5 or 6 amps. I’d use dual 2300 packs on a 40%, if not for the added current then for the redundancy, but I have no problem with using one A123 2300 on a 40% plane since one pack will easily burst over 30 amps.

Tom Fawcett
Wild Hare R/C

I replaced the photos with actuual jpg files and deleted the originals.

[HuckinChikn]

thanks for the great info wildhare! Only one 2300mah pack, huh? very cool!

[Absolut Yak]

Fantastic post, Tom. Makes me completely comfortable with my A123 setup in my 50cc plane. Thanks for the post.

Doug

[Shoe]

Excellent post Tom, thanks much!

[ChrisR]

Great post, and great information.

[Dr. Yankem]

Great post Tom..Confirms my gut feeling that I need 2 packs for my 40% Godfrey. If the weather warms up here in the NE I'll supply the data on usage with my 2 packs, therwise it wont be until I get down to Florida for the winter in 3 weeks
Jerry

[Greyhoundman]

Tom, great info I have been looking for some real world numbers on these packs this helps all the skeptics out there like me.

[psk560]

Great work Tom. I've been using two a123's on my 2.6 extra and they last all day. I could loose the weight by going to one cell but I don't see the need as it pulls nicely out of a hover now. I'm considering putting these in my 37.5% yak and your data sure helps.

Craig

[dmcquinn]

Good work, Tom. I also use the Eagletree telemetry / logger system. Here's how to capture one of the graphs on your PC: Get the graph that you want on the screen. Hold down the Alt-key and while holding this key down, press the "Prnt screen" key. This is a standard windows sequence that puts the current window as a bitmap into the clipborad. Then launch "paint" (start programs accessories paint) and do "edit paste". The save this image as a JPEG file. Here's an example (this one is showing cylinder temps)


Well --- I'm having trouble uploading the jpg file. It is within the size limitations and looks OK to me, but I get "upload failed".. suspect this is a server issue now. will try later.

[wildhare]

A little research turned up the ability to export the graphs as .jpg files. I have replaced the pics in my first post.

TF

[camss69]

Nice post Tom, your A123 info lately has really been helpful.

Thanks.

[Pale Rider]

Tom--thanks for the detailed info. So from your results it seems that the standard JR connectors can handle the loads pretty well as long as they are split into two paths. Ive always thought that people got a bit carried away with heavy duty plugs, no switches etc on 35%'er's and down. As long as you are not running all the power through one standard connection they seem to hold up just fine.
I really cant see why anyone would want to setup a new plane with a regulated setup after seeing how the A123's perform.

[dick hanson]

Skeptics ?
we have skeptics?
Chee--- My setup on my departed EDGE was similiar -I likely flew faster - (piped 80 setup clipped wings etc ) and did knife edge loops etc and in 12 minutes used close to 400ma - I use data loggers -and take off votage fresh - 7 volts .
really the loads from flying the packs as servo power are quite low loads for the 2300ma A123.
My electric models see FIFTY (50)amps all the time and voltage drop is only on the order of one to eventually 2 volts ---- from 18 volts at takeoff.
skeptics ,these cells ain't lipos or the LiIons you typically see.

[wildhare]

Yes, I have tested the standard servo connectors holding them in my hand, they are fine for up to about 8 amps in short bursts and can easily handle 5 amps continuous. Two of them are perfectly adequate for a 35% plane in this application.

For a 40% 14 servo plane you would need to use 2 receivers with 2 connections each, but that will work fine also. To use a single receiver you will either need something like the AR9100, and that actually does not have enough places to plug in 14 servos. Assuming you use a matchbox on the rudder and ailerons the single AR9100 should work fine.

But isn't the 14 servo plane pretty much obsolete? Most (if not all) of the 41% Extras I have sold are being flown with 2 servos on each aileron and 1 on each elevator half, just like the 35% planes. They do have 2 or 3 rudder servos, but the 5955 and 8711 have pretty much done away with the need to have 14 servos in a plane. 8x8711 or 5955 seems to be the norm these days, so a single AR9100 ought to be the cat's meow for a big plane.

I'll have more info later on the Liion setup.

TF

[wildhare]

Here's a little follow up on how other batteries perform. I was a little surprised at the results.

The first graph shows an A123 pack that sI used as a baseline. I ran it on my load tester and tried to simulate a flight. I tried to keep it in the 2 to 4 amp range most of the time with occasional bursts of a few seconds up to 11 amps. It appears from the graph that I got a little carried away, there were some bursts to around 20 amps. Oops, the A123 will put out way more than I expected andmy inline power meter did not reactquickly enough to tell me I had gone beyond the target load.

The second graph shows two sessions, the first is a freshly charged 5 cell 2700 Nimh pack, these are Sanyo 2700 mah AA size cells. The second part is a freshly charged 2400mah Liion pack similar to what everybody uses with regulators. In this test I did not use a regulator preferring to see just what the pack itself will produce.

Predictably the Liion pack has higher initial voltages. The voltage swing during peaks on the Nimh pack was about .7 volts. The voltage swing on the Liion was more like 2 volts. Voltage swing with the baseline A123 was only about .6 volts and that was going up to 20 amps.

In both the Nimh and Liion tests notice that at the end of the test I just opened it wide open to see what would happen. You can see the results for yourself.

TF