EEVblog #751 - How To Debunk A Product (The Batteriser)

[5ky]

I found this quote, from a PHD who actually seems to understand battery chemistry and underlying motivations, interesting:

"...All the marketing, videos and posts by the company in question still show that they do not have the necessary understanding of battery chemistry which is needed. If they interchangeably mistake voltage for power than this might sadly be enough to fool venture capitalists but it will not help their product at all.

Yes it is amazing that you can develop a boost converter with such a small footprint nowadays, but this does really not change the physics and chemistry which is responsible for a hugely nonlinear relationship between battery voltage and remaining energy in the battery. If you take a piece of copper and a piece of zinc and press it into a lemon, you get a battery with quite a high voltage, but there really is not much energy which could be extracted this way. The same is true for the chemistry of a regular AA battery: the voltage is determined by the chemistry and the energy stored is determined by the amount of remaining metal and chemicals.

Also, as Dave Jones brilliantly explained lately, the current you need to extract from the battery to deliver the same power from the boost converter, rises towards the very end of the lifetime of a battery, making the internal series resistance more and more working against you.

There is no significant energy left in an alkaline AA battery, once the voltage has dropped below 1 V, even though this is still 67% of the original voltage - but the anode metal is consumed at this point!

There is only one party here who really wants to make money from this story, and it's neither the battery manufacturers, nor is it Dave Jones.

Uwe Zimmermann,
PhD, Senior lecturer in electronics @ Uppsala University"

- http://www.edn.com/electronics-blogs/brians-brain/4440136/The-Batteriser--scam-or-savior-
Sorry if it's already been posted elsewhere, the comment threads are hard to keep track of.

EXCELLENT find!!

Also, it's only 4 hours of logging since I had to go to bed due to having to get up early for work, but here's a graph of the discharge of fresh AA duracells in the exact same gps unit that batteriser used in their video.  In their video, they show it dying at about 1 hr 50 minutes.  Mine was still full screen brightness (as far as I could tell) at the 4 hour mark, and since it was still at 2.55v, that means it still had another 450mV until it hits voltage cutoff.  Their video is absolutely bullshit

[FrankBuss]

There is an easy way to tell if you can squeeze more juce out of a Battery.
If you know the theoretical Capacity of the Battery, you can easily tell how much juce is left.
For this reason i did a chemical analysis of a brand new DURACELL Battery.
The results are in the attached pdf file.
In short:
The Battery delivers a capacity of about 3,5Ah with a constant Load of 5mA (datasheet) und this is almost the theoretical Limit (calculated from the analysis: 3,554Ah) !

I haven't seen that done before, nice!, thanks for sharing.

OK, so we can now discharge our batteries down to a specific voltage, measuring mAh as we go, and we know exactly how much energy is left in the battery when we reach that voltage. No need to calculate areas under curves.

It might be worthwhile doing that and plotting a graph of battery voltage vs. remaining energy (if anybody has the time/inclination/hardware).

ESR is a problem though - the observed voltage will be different under different loads. You'd have to plot a graph at several different loads for it to be useful.

What would be cool is a programmable device that could vary the load every minute or so, logging the voltage at each load - do it in one pass.

I don't know if this would help. Batteries are strange. If you discharge it with 100 mA constant current, but then measure it with 5 mA current, voltage will rise over time. How high does it rise and how long does it need? As mentioned by Akra, it is the recovery effect.

And there are other effects, like if you discharge it with 25 mA constant current, looks like you get more energy from the battery until it is discharged compared to 500 mA. At least this suggests this datasheet, but specifying the capacity in mAh might be misleading and Wh might be the same. But nevertheless it was interesting to see how to calculate the max theoretical energy of a battery with the chemical reaction formula.

[PeterL]

And there are other effects, like if you discharge it with 25 mA constant current, looks like you get more energy from the battery until it is discharged compared to 500 mA. At least this suggests this datasheet, but specifying the capacity in mAh might be misleading and Wh might be the same. But nevertheless it was interesting to see how to calculate the max theoretical energy of a battery with the chemical reaction formula.

At higher currents you lose more energy in the ESR. Also at higher currents the terminal voltage is lower, so the battery has to be discarded as empty sooner, because of the larger voltage drop across the ESR.

[5ky]

Battery chemist is a tricky thing.  Even using Peukert's constant, you still have to account for battery age and temperature.  Tricky business 

[Fungus]

I don't know if this would help. Batteries are strange. If you discharge it with 100 mA constant current, but then measure it with 5 mA current, voltage will rise over time. How high does it rise and how long does it need?

That's all part of the learning process... 

[dr_frost_dk]

Just did a Google Search for "Batteriser" from a computer that have never searched for that, also not at home so the INET provider is also different.
Same results as "private browsing", batteriser.com and various pages, at 4-5 place the debunking pages/videos begins

And 5ky nice work, looking forward to see the results, i used to test Li-Po/ion battery's so i know it can be time consuming...

If you have any kind of time-lapse camera you could also log voltage by having an "always on" multimeter in frame with the GPS and just having it take picture every 1 sec, i have a lot of tiny "spycams" that does this and makes it into a 30fps video (1 sec video is then 30 sec real time), this is for a very simple setup anybody can throw together, will not be the most precise but will be MUCH better then anything i have seen so far from batteriser.

[5ky]

Just did a Google Search for "Batteriser" from a computer that have never searched for that, also not at home so the INET provider is also different.
Same results as "private browsing", batteriser.com and various pages, at 4-5 place the debunking pages/videos begins

And 5ky nice work, looking forward to see the results, i used to test Li-Po/ion battery's so i know it can be time consuming...

If you have any kind of time-lapse camera you could also log voltage by having an "always on" multimeter in frame with the GPS and just having it take picture every 1 sec, i have a lot of tiny "spycams" that does this and makes it into a 30fps video (1 sec video is then 30 sec real time), this is for a very simple setup anybody can throw together, will not be the most precise but will be MUCH better then anything i have seen so far from batteriser.

funny enough, I used my old smartphone to do a timelapse last night.  i had an analog clock in view, as well as the computer monitor where you can see the voltage, and in the foreground you can see the gps unit and the servo setup

[FrankBuss]

If you have any kind of time-lapse camera you could also log voltage by having an "always on" multimeter in frame with the GPS and just having it take picture every 1 sec, i have a lot of tiny "spycams" that does this and makes it into a 30fps video (1 sec video is then 30 sec real time), this is for a very simple setup anybody can throw together, will not be the most precise but will be MUCH better then anything i have seen so far from batteriser.

It is better to use an Arduino, as I did, because then you have the voltage reading data:
https://www.eevblog.com/forum/blog/eevblog-751-how-to-debunk-a-product-%28the-batteriser%29/msg743529/#msg743529
Testing if the GPS is still running could be done with an additional photodiode, which would result in another curve with probably three levels: full brightness, dimmed screen after the warning message and device off. Bonus points if you use a uCurrent to measure the current as a third channel

[5ky]

If you have any kind of time-lapse camera you could also log voltage by having an "always on" multimeter in frame with the GPS and just having it take picture every 1 sec, i have a lot of tiny "spycams" that does this and makes it into a 30fps video (1 sec video is then 30 sec real time), this is for a very simple setup anybody can throw together, will not be the most precise but will be MUCH better then anything i have seen so far from batteriser.

It is better to use an Arduino, as I did, because then you have the voltage reading data:
https://www.eevblog.com/forum/blog/eevblog-751-how-to-debunk-a-product-%28the-batteriser%29/msg743529/#msg743529
Testing if the GPS is still running could be done with an additional photodiode, which would result in another curve with probably three levels: full brightness, dimmed screen after the warning message and device off. Bonus points if you use a uCurrent to measure the current as a third channel

Yeah I probably should have gone that route.  Would be more robust than free version of benchvue.

I do have a uCurrent so maybe I'll have to work that into the mix when I retest this weekend

[FrankBuss]

Looks like your nice 34461A can save the logged data to an USB memory stick, if the free PC software is too limited. But is it true that it can't measure voltage and current at the same time? Then an Arduino would be indeed better for this application. See this post for the latest Arduino sketch, because my first version could crash because of unsafe functions within the interrupt. One sample per second is sufficient, so you can oversample it for higher resoultion than the 10 bits ADC. And if you supply a reference voltage to the AREF pin (and enabling it with analogReference(EXTERNAL)) it gets even better.

[5ky]

Looks like your nice 34461A can save the logged data to an USB memory stick, if the free PC software is too limited. But is it true that it can't measure voltage and current at the same time? Then an Arduino would be indeed better for this application. See this post for the latest Arduino sketch, because my first version could crash because of unsafe functions within the interrupt. One sample per second is sufficient, so you can oversample it for higher resoultion than the 10 bits ADC. And if you supply a reference voltage to the AREF pin (and enabling it with analogReference(EXTERNAL)) it gets even better.

As silly as it sounds, my model can't data log without a computer. I think the 70 model adds that menu option. It can measure voltage and current simultaneously, but I'm not sure if that works while data logging. I might have to try. I'm curious what the burden is on that thing

[edy]

I wrote a blog post highlighting some big IndieGogo and Kickstarter "Scampaigns" (similar to Batteriser) and one of the ads on my page pointed to this company:

https://www.krowdster.co/media-lists

It's no wonder "media" and "journalists" on the web don't think twice about disseminating and promoting dubious companies and products. All of these so-called writers are simply trying to make a living on the internet and will feature just about anything for the right price. The "Krowdster" site is another example of how PR firms/"Journalist"-Bloggers and such are paid to flood the internet with whatever a company wants you to hear. Sadly, most people don't know how rampant it is, and tend to believe articles.

I'm glad to see the PC-WORLD article that Batteriser keeps linking to getting a flood-storm of negative comments. At least they didn't turn that off, although one person said their negative post was removed. But there is a lot of chatter on the PC-World article:

http://www.pcworld.com/article/2928997/batteriser-is-a-250-gadget-that-extends-disposable-battery-life-by-800-percent.html

[TheAmmoniacal]

I wrote a blog post highlighting some big IndieGogo and Kickstarter "Scampaigns" (similar to Batteriser) and one of the ads on my page pointed to this company:

https://www.krowdster.co/media-lists

It's no wonder "media" and "journalists" on the web don't think twice about disseminating and promoting dubious companies and products. All of these so-called writers are simply trying to make a living on the internet and will feature just about anything for the right price. The "Krowdster" site is another example of how PR firms/"Journalist"-Bloggers and such are paid to flood the internet with whatever a company wants you to hear. Sadly, most people don't know how rampant it is, and tend to believe articles.

I'm glad to see the PC-WORLD article that Batteriser keeps linking to getting a flood-storm of negative comments. At least they didn't turn that off, although one person said their negative post was removed. But there is a lot of chatter on the PC-World article:

http://www.pcworld.com/article/2928997/batteriser-is-a-250-gadget-that-extends-disposable-battery-life-by-800-percent.html

That stuff is scary, almost makes me want to mention GamerGate 

[ccs46]

There is an easy way to tell if you can squeeze more juce out of a Battery.
If you know the theoretical Capacity of the Battery, you can easily tell how much juce is left.
For this reason i did a chemical analysis of a brand new DURACELL Battery.
The results are in the attached pdf file.
In short:
The Battery delivers a capacity of about 3,5Ah with a constant Load of 5mA (datasheet) und this is almost the theoretical Limit (calculated from the analysis: 3,554Ah) !

I haven't seen that done before, nice!, thanks for sharing.

OK, so we can now discharge our batteries down to a specific voltage, measuring mAh as we go, and we know exactly how much energy is left in the battery when we reach that voltage. No need to calculate areas under curves.

It might be worthwhile doing that and plotting a graph of battery voltage vs. remaining energy (if anybody has the time/inclination/hardware).

ESR is a problem though - the observed voltage will be different under different loads. You'd have to plot a graph at several different loads for it to be useful.

What would be cool is a programmable device that could vary the load every minute or so, logging the voltage at each load - do it in one pass.

I don't know if this would help. Batteries are strange. If you discharge it with 100 mA constant current, but then measure it with 5 mA current, voltage will rise over time. How high does it rise and how long does it need? As mentioned by Akra, it is the recovery effect.

And there are other effects, like if you discharge it with 25 mA constant current, looks like you get more energy from the battery until it is discharged compared to 500 mA. At least this suggests this datasheet, but specifying the capacity in mAh might be misleading and Wh might be the same. But nevertheless it was interesting to see how to calculate the max theoretical energy of a battery with the chemical reaction formula.

Would you like the source code for my Arduino that's designed to do this?


Sent from my iPhone using Tapatalk

[robby1995]

it looks like the Red curve is WITH bat :bullshit:teriser and the geen without.
in the red curve the current is getting bigger while the voltage drops.
maybe they mixed up the results 

[Fungus]

it looks like the Red curve is WITH bat :bullshit:teriser and the geen without.
in the red curve the current is getting bigger while the voltage drops.
maybe they mixed up the results 

Those graphs are exactly what you'd expect to see if the GPS already has a built-in DC booster. I wonder if it does...? 

[photon]

A discussion has started on the website of the VC company backing Batteriser:
http://sktainnopartners.com/welcome-to-batteroo/

I would not want to be the VC who approved this investment. Talk about career shortening decisions!

[Fungus]

So they are claiming that the GPS device only used 300-400mAh out of a brand new pair of duracells before becoming unusable? Nonsense.

You might have indirectly found the first Batteriser efficiency measurement!   

They managed ten hours at 150mA. That means they got about 1500 mAh from their batteries, not 300-400.

But that's on the GPS side of the Batteriser... on the battery side it's constant-power so we need to use mWh:  1.5V @ 150mA is 225mW of power.

Here's the battery datasheet:

http://www.professional.duracell.com/downloads/datasheets/product/Ultra%20Power/Ultra-Power_AA_MX1500.pdf

On page 2 there's graphs for constant power. According to the second graph a drain of 225mW will last for approx. 11 hours. They got 10.

So...either:
a) Batteriser is 90% efficient at 225mW,
or
b) They cheated.

Thoughts?

(nb. This post is a correction of a previous one where I didn't check his math)

[PeterL]

So they are claiming that the GPS device only used 300-400mAh out of a brand new pair of duracells before becoming unusable? Nonsense.

You might have indirectly found the first Batteriser efficiency measurement!   

They managed ten hours at 150mA. That means they got about 1500 mAh from their batteries, not 300-400.

But that's on the GPS side of the Batteriser... on the battery side the current isn't constant so we need to use mWh:  1.5V @ 150mA is 225mW of power.

Here's the battery datasheet:

http://www.professional.duracell.com/downloads/datasheets/product/Ultra%20Power/Ultra-Power_AA_MX1500.pdf

On page 2 there's graphs for constant power. According to the second graph a drain of 225mW will last for approx. 11 hours. They got 10.

So...either:
a) Batteriser is 90% efficient at 225mW,
or
b) They cheated.

Thoughts?

(nb. This post is a correction of a previous one where I didn't check his math)

I think the average current draw is closer to 135 mA, the with of the spikes is not to scale here so you cannot really tell. But 135mA gives a power closer to 200mW per cell. And with that they should get some 13 hours, since they go down to 0.5V cell voltage.
And then you get closer to 77% effecient.

[EEVblog]

Also, it's only 4 hours of logging since I had to go to bed due to having to get up early for work, but here's a graph of the discharge of fresh AA duracells in the exact same gps unit that batteriser used in their video.  In their video, they show it dying at about 1 hr 50 minutes.  Mine was still full screen brightness (as far as I could tell) at the 4 hour mark, and since it was still at 2.55v, that means it still had another 450mV until it hits voltage cutoff.  Their video is absolutely bullshit

The big question is did you get the firmware message and the backlight turning off?
You should have, as it's supposed to be a firmware feature if the setting is for Alkaline batteries.
There is a report that the backlight eventually turns back on though.
But in either case though, sounds like their GPS video is thoroughly busted, nice work!

[EEVblog]

As silly as it sounds, my model can't data log without a computer. I think the 70 model adds that menu option.

Correct, the 70 model has the data logging. But the 34461A has the trend lot that can be extracted to USB stick after you stop it.

[AmmoJammo]

I was thinking about this last night... possibly in my sleep....

If a device NEEDS, 1.1volts per cell, or 2.2volts total for something using two AA cells in series, then logically, if there was energy below the 1.1volt per cell cutoff, then wouldn't the manufacturer design the device to run off three AA (or AAA) cells in series, so the 2.2volt cut off occurs at 0.75volts per cell...

But then I also thought, it's easy enough to see how much longer you can run from depleted cells, by taking a device that's meant to run off two, and wiring three in series, once they're already depleted of course (or in a device you don't care about)

No boost circuit required!

[LabSpokane]

And it's pretty well game over for Batteroo. The retailers have obviously walked away or at least will hold off until they see the product perform in the marketplace. 2016 could not have been the plan. They will be just another retailer on Amazon (which anyone can do) and selling on their own e-commerce site. And the suspicions of not having anywhere near enough cash to pull this off are correct. It sounds like the actual retail price will be more than $2.50 each - which would have put their BOM cost around $.80-85 per unit.  That price would require some very aggressive component deals which only could have occurred in volumes much greater than what they have commitments for. The only question left is whether they will be taping money to each package of Batterisers if and when the IGG campaign ships.

It's an ironic end, since after it has all shook out, the business development plan is back to bootstrapping.

[5ky]

Also, it's only 4 hours of logging since I had to go to bed due to having to get up early for work, but here's a graph of the discharge of fresh AA duracells in the exact same gps unit that batteriser used in their video.  In their video, they show it dying at about 1 hr 50 minutes.  Mine was still full screen brightness (as far as I could tell) at the 4 hour mark, and since it was still at 2.55v, that means it still had another 450mV until it hits voltage cutoff.  Their video is absolutely bullshit

The big question is did you get the firmware message and the backlight turning off?
You should have, as it's supposed to be a firmware feature if the setting is for Alkaline batteries.
There is a report that the backlight eventually turns back on though.
But in either case though, sounds like their GPS video is thoroughly busted, nice work!

I had it set to Alkaline and I never saw the brightness message even after the four hours, but since I have the servo hit the screen right where that "ok" button is (intentionally), I never saw it.  It didn't appear to have dimmed yet, however.  When testing with my bench power supply it dimmed at 1.3v per cell, and when I stopped last night, it was at 1.275v per cell, so I'm not sure if perhaps it just hadn't triggered the warning yet or not, but the message matches the one on their video (which you can somewhat read if you pause it at the right time) when it does appear.  I just got a fresh pair started and since it's early and I don't work tomorrow, this test will go until the GPS shuts itself off.  (and I installed an older version of the Agilent DMM software that doesn't have a time limit so I shouldn't have to hit start every hour this time.  Will post video/results when I'm done so people can see how BS their marketing videos really are.

[edy]

And it's pretty well game over for Batteroo.

They may do well on those late night infomercials. And also bundling with toys or cheap dollar-store batteries. People will buy 1 or 2 and then realize it has little practical use. But meanwhile they will sell them by the millions... because nobody will care if they spent a few bucks on a "gamble".