Become a Patron!

The odd DNA40 battery meter - is it learning?

Giraut

Gold Contributor
Member For 4 Years
Okay so like most DNA40 owners, one of the first things I noticed is how totally unlinear the battery meter is: it'd show the battery full for hours, and then almost before they'd be totally spent, the meter would suddenly drop down to zero.

Then after some experimenting, I noticed it seemed to try and be clever by showing not simply the battery's charge level, but some sort of "remaining vape time" based on the user's usage pattern. It seems to integrate the power actually delivered to the coil by the board over time. Or in other words, the board seems to keep track of the electrical energy spent, and shows what it reckons is left in the batteries based on the amount of energy it's delivered to the coil during the last toke.

But, I think it's even more complex than that: I reckon it's slowly learning the battery's actual capacity, and fine-tunes its calculation over time. Unlike the eVic which has a menu in which you can enter the battery's capacity, the DNA40 doesn't know anything about the battery it's connected to. So if it wants to figure out the capacity by itself, it stands to reason that it needs to "see" several charge/discharge cycles, and that the more cycles it sees and the deeper the discharges, the more accurately the capacity is measured.

With that assumption in mind, I've tried to vape running down the battery to the ground several times in a row (essentially vaping until the board refuses to vape anymore, before recharging), and lo, charge after charge, the meter seemed to become more and more linear!

It's now quite accurate: with two 2900 mAh batteries, vaping at 25W the way I vape, I usually get over 36 hours of actual vape time. About 18 hours since the last charge, if the board doesn't hit TP, the meter shows half-full after each typical toke I take. If I jack up the power all the way to 40W, it shows a little over a quarter full. If I lower it to 12W, or if I let the wick go dry and it hits TP, it shows 3/4 full. If I take quicker tokes, the meter goes up. But as long as I vape "normally" (normally for me, that is), it correctly shows half-full.

The meter still isn't very obvious, but at least it makes sense now. I wish Evolv would explain how it works exactly, but they're totally uncommunicative when you ask them technical questions. So that's my guess, and I think it's correct. Has anybody else noticed the meter becomes more accurate after many charge cycles?
 

Bow2King

Bronze Contributor
Member For 4 Years
i havent noticed it yet!! but are you constantly using the same atty or are you using different ones??
 

Midniteoyl

Gold Contributor
Member For 4 Years
Member For 3 Years
Member For 2 Years
Member For 1 Year
Member For 5 Years
Didnt Brandon say that thats the way the batt meter works when talking to Phil? Thought he did..
 

Zamazam

Evil Vulcan's do it with Logic
VU Donator
Platinum Contributor
Member For 5 Years
From my understanding, the board measures the resistance of the battery in it's energy calculations. Take a series of quick high power hits and the battery needs to rebound, which is what the meter is showing.
 

Giraut

Gold Contributor
Member For 4 Years
I spoke too fast: it's not learning, it's going stupid. My batteries are now fully charged, and the damn meter shows 3/4 full.

This this is officially messed up...
 

Giraut

Gold Contributor
Member For 4 Years
Apparently, the battery meter is highly dependent on the battery's internal resistance: the more the voltage drops when the chip pulls current, the more the meter becomes idiotic.

Worse: even if the battery is perfectly capable of supplying enough current for 40W, if the voltage under load drops too low, the chip incorrectly decides the battery is spent and quits working. This has just happened to me with high capacity, low-amp batteries that are barely a month old: they still read 4.0V unloaded, but the DNA40 won't use them. Yet they are rated for 7A (so 14A with two in parallel), which is more than plenty good enough for 40W.

So that's the reason why Evolv recommends high-amp batteries: it's not that the DNA40 chip performs better with them, it's just to compensate for an extremely clunky and dysfunctional battery metering algorithm. I now use a pair of 35A batts in my flask, and the meter is back to its old stay-at-100%-all-the-time behavior - and I get 3x more battery life, even though they're only 2500mAh.

Stupidy stupidy...
 

VU Sponsors

Top