Eagletree graphs may not be accurate

[BrianG]

This post is for anyone using an Eagletree device to gauge their amp draw in their setup.

I was playing around with my XT8 truggy (Medusa 36-80-2200, MMM, 5s 5Ah RCM Lipos) and decided to get an ET graph just for giggles. Once I downloaded the results, I was overall pleased, but a little disappointed at one area of the results. At 161A, the voltage fell to 15.74 (3.148v/cell). The puzzling part was that the rest of the graph didn't support this reading.

So, I plotted a number of different voltage and current points at various places and the total pack resistance varies from 0.0131 to 0.0133, at currents spanning from 53A up to 121.5A. That's a pretty consistent resistance number.

So, something is amiss here. I figure one of two possibities:

1) The pack resistance increased dramatically somewhere between 121A and 161A. Well, I don't believe this to be true. I highly doubt the pack resistance was pretty constant all the way up to 121A, and then rose dramatically at just 40A more. Possible I suppose, but unlikely.

2) Eagletree errors. If I assume the pack voltage was correct (15.74v) and use my extrapolated resistance number of 0.135 (slightly higher to compensate for the higher current), I get a calculated actual current closer to 200A (40C).

Option number 2 makes more sense. And to support this reasoning, I took a real close look at the graph at that 161A spike and noticed the spike flattens abruptly at around 155A, but the pack voltage spike was pretty peaked. This tells me the ET device drastically loses accuracy much beyond its 150A rating. A look at the Hall-Effect datasheet for the sensor the ET uses, I see that it is rated up to 150A, and that's it. It doesn't even give a little extra margin.

So, long story short; if you are using an Eagletree device to gauge your setup, be aware that the currents you are reading may be a lot less than they actually are beyond the rating. I'm thinking a 300A version would be better to use for larger vehicles.

which et did you use: the 100a or 150a?

[Unsullied_Spy]

Originally Posted by drkdgglr

which et did you use: the 100a or 150a?

2nd from last paragraph makes it sound like he's got the 150A version.

Sucks that it isn't as perfect as we all seem to believe, but at least it's accurate to it's rated specs.

[pinkpanda3310]

Would the sudden voltage drop and/or ripple have any effect on the ET's accuracy?

[BrianG]

Sorry, yeah, it's the 150A version.

Not sure about what effects ripple and/or voltage drop would have. It is essentially just a voltmeter with a seperate hall-effect sensor for current. Voltmeters typically have high input impedance, so the ripple would just look like changing DC. And the hall-effect sensor is an isolated circuit. The current measuring part is electrically seperate from the part that measures it, and looks like a straight piece of wire as far as resistance goes. So, I wouldn't think ripple is a big deal at all. I just think the current exceeded the sensor's spec.

Allegro Microsystems (the company that makes the actual hall-effect sensor IC) also makes a 200A version. I emailed ET and asked if this IC could be used in place. But, I'm afraid the answer will be "no". Even though the IC is pin-compatible, I assume firmware changes would be necessary for the rest of the circuit to recognize the higher current. These hall-effect sensors are pretty simple really: They output a linear voltage in proportion to the current. If the circuit expects to see 0v @ 0A and 5v @ 150A, the new sensor will ouput that 5v at 200A instead making the readings all wrong. I'd have to re-interpret the results for every test I take.

[nieles]

in the software you can "manipulate" the voltage an current the eagle tree is reading.

under: advanced > calibrate pack voltage or amperage readings.

so you put 150A for the recorder and 200A for the other meter

[BrianG]

I know what you are talking about, but I don't think that'll work. As far as I can tell, that's just adds in a static offset current. So, if you say the ET is reading 10A when the current is actually 12A, that's just a 2A offset. Using the 200A HE sensor would require a current multiplier of 1.33. So, at 0A, it would read 0A, but at 13.3A, it would read 10A, and so on.

[fastbaja5b]

Might not be 100% but a rough idea is better than no idea.

[nieles]

i tested it with the voltage setting
recorder :10v
meter: 20v

then i plugged in an 5s pack and it reads ~40 volts

[pinkpanda3310]

It's probably less versatile and comes with no warranty but medusa make a 200A 'Oracle'.

[BrianG]

Originally Posted by nieles

i tested it with the voltage setting
recorder :10v
meter: 20v

then i plugged in an 5s pack and it reads ~40 volts

That's odd. That would mean that it applies that calibration amount as a multiplier of sorts, and not a static offset. I got different results. I'll have to re-test I guess.

I still want to wait until I hear from ET before I go ordering the new part and modding my current ET just to find out it doesn't work for some reason. I hope the ET engineers will see the email so I can get confirmation that my hunch is accurate (about it not reading accurately above the hall-effect sensor rating). It would be nice to see those actual peak values without trying to extrapolate.

Yeah, I could go with the Medusa solution, but I read they use a shunt, which has insertion losses (and therefore heat). The Hall-effect sensor is a much better solution IMO.

[glassdoctor]

BG, I saw the same thing when using my 100A Eagletree. It reads pretty good up to about 115-120A but anything beyond that just reads 120A.

So sometimes you will see a big difference between a 110A spike and a "120A" spike, because the latter wasn't 120A but was higher....

[BrianG]

What sucks about this is that what some would deem a "gentle" setup can pull bursts over 150A depending on the setup. I've logged a burst around 145A on a CRT.5 on 3s. I can just imagine what larger setups geared high will pull...

[glassdoctor]

[QUOTE=BrianG;309615]I've logged a burst around 145A on a CRT.5 on 3s. /QUOTE]

Impressive

I think the 150a unit would be plenty for a racer's setup... but probably not for bashing where it's more about speed and wheelies, etc

The only time I would see more than 120A is when I ran my truggy with a 4s 1515 setup... which was only geared for mid 30's but could still trip 100 pretty easy on a full throttle launch

[nieles]

just did the test with an ~1 ohm resistor an an 5s pack
recorder: 1A
meter 2A

it reads ~30A and without the calibrate thing it shows ~15A

test 2 to make sure my test1 was right

recorder: 10A
meter 13.3A

and it reads 20A